Microbial oxidation as a methane sink 
beneath the West Antarctic Ice Sheet
Authors: Alexander B. Michaud, John E. Dore, Amanda M. 
Achberger, Brent C. Christner, Andrew C. Mitchell, Mark L. 
Skidmore, Trista J. Vick-Majors, and John C. Priscu
This is a postprint of an article that originally appeared in Nature Geoscience on July 31, 2017. 
The final version can be found at https://dx.doi.org/10.1038/ngeo2992.
Michaud, Alexander B., John E. Dore, Amanda M. Achberger, Brent C. Christner, Andrew C. 
Mitchell, Mark L. Skidmore, Trista J. Vick-Majors, and John C. Priscu. "Microbial oxidation as a 
methane sink beneath the West Antarctic Ice Sheet." Nature Geoscience 10 (July 2017): 
582-586. DOI: 10.1038/ngeo2992 .
Made available through Montana State University’s ScholarWorks 
scholarworks.montana.edu 
Microbial oxidation as a methane sink beneath the
West Antarctic Ice Sheet
Alexander B. Michaud1*†, John E. Dore1, Amanda M. Achberger2†, Brent C. Christner2,3,
Andrew C. Mitchell4, Mark L. Skidmore5, Trista J. Vick-Majors1† and John C. Priscu1*
Aquatic habitats beneath ice masses contain active microbial ecosystems capable of cycling important greenhouse gases,
such as methane (CH4). A large methane reservoir is thought to exist beneath the West Antarctic Ice Sheet, but its
quantity, source and ultimate fate are poorly understood. For instance, O2 supplied by basal melting should result in
conditions favourable for aerobic methane oxidation. Here we use measurements of methane concentrations and stable
isotope compositions along with genomic analyses to assess the sources and cycling of methane in Subglacial Lake Whillans
(SLW) in West Antarctica. We show that sub-ice-sheet methane is produced through the biological reduction of CO2 using
H2. This methane pool is subsequently consumed by aerobic, bacterial methane oxidation at the SLW sediment–water
interface. Bacterial oxidation consumes >99% of the methane and represents a significant methane sink, and source
of biomass carbon and metabolic energy to the surficial SLW sediments. We conclude that aerobic methanotrophy may
mitigate the release of methane to the atmosphere upon subglacial water drainage to ice sheet margins and during periods
of deglaciation.
Methane (CH4) is an important greenhouse gas that affectsatmospheric chemistry and the radiative balance of Earth.Consequently, understanding its global sources, sinks, and
feedbacks within the climate system is of considerable importance1.
The primary pathway for biological CH4 production in carbon-
rich habitats (for example, bogs, wetlands) is the anaerobic
fermentation of simple organic compounds by certain archaea
(acetoclastic or methylotrophic methanogenesis2). An alternative
microbial pathway to CH4 production is the reduction of CO2
coupled to the oxidation of H2 (hydrogenotrophicmethanogenesis),
which is common in anoxic, low-sulfate environments such as
the methanogenic zone within marine sediments2. Conversely,
bacterial and archaeal oxidation of CH4 (aerobic and anaerobic,
respectively) to CO2 is amajor pathway that reduces net CH4 release
to the atmosphere3.
Anoxic habitats in sediments beneath the Antarctic ice sheetmay
be globally important sites of biological CH4 production that could
potentially add significant CH4 to the atmosphere upon subglacial
water drainage to the ice sheet margins or deglaciation4–6. However,
due to release of oxygen into the subglacial environment from the
overlying ice sheet through geothermal heat-induced melting7–9,
aerobicmethanotrophic activity can ultimatelymitigate CH4 release
to the atmosphere. We present data on CH4 concentration and
stable isotopic composition, along with genomic data collected from
Subglacial Lake Whillans (SLW), which lies ∼800m beneath the
West Antarctic Ice Sheet (WAIS). Collectively, these data reveal the
presence of an ecosystem supported, in part, by active microbial
transformations of CH4.
Quantity and source of sub-ice-sheet CH4
CH4 concentration in SLW ranged from 0.024 µM in the lake wa-
ter to 300 µM in the deepest (39 cm) sediment porewater sample
(Fig. 1). Fick’s first law was used to compute a flux of 6.8±1.8
(mean ± SE) mmol CH4 m−2 yr−1 into the surficial sediment
(0–2 cm) of SLW using the concentration gradient in the top 15 cm
of sediment and the associated error of the concentration gra-
dient, which includes any potential sampling artefacts. CH4 in
the SLW sediment had an average δ13C–CH4 value of −74.7h
(range: −77.1 to −70.1h) (Fig. 1) and, together with δD–CH4
values (range:−247.6 to−239.3h), reveals that SLW CH4 is prob-
ably produced by hydrogenotrophic methanogenesis10 (Fig. 2). This
conclusion contrasts with previous models suggesting that potential
CH4 reservoirs beneath the WAIS would be largely formed through
acetoclastic methanogenesis4. Hydrogenotrophic methanogenesis is
common in marine sediments and other environments with low
concentrations of old organic carbon, supporting our results from
SLW, which also has low organic carbon and acetate (2–14 µM) rel-
ative to environments with active acetoclastic methanogenesis10–13
(Supplementary Fig. 1). CO2 for hydrogenotrophic methanogenesis
can be supplied from microbial respiration or bicarbonate in sedi-
ment porewater (2–6mM; ref. 14), and hydrogen can be generated
abiotically from glacially crushed siliceous bedrock, radiolysis of
water, hydrothermal input, or biologically via fermentation2,8,15,16.
Attempts to amplify a marker gene for methanogenic archaea
(mcrA)17,18 from the 0–2, 4–6, 18–20 and 34–36 cm depth intervals
within the SLW sediment core were unsuccessful, implying that the
abundance of methanogenic archaea was low or below detection.

−5 −4
log (Ar
kg) (J kg H2O
−1)
Ar
e− (kJ mol e_−1)
−3
0.5 O2
0.1 O2
0.5 O2
0.1 O2
−2 −1 0 1 2
FeS2 + 3.5O2 + H2O → Fe
2+ + 2H+ + 2SO4
2−
CH3COO
− + 2O2 → 2HCO3
− + H+
NH4
+ + 2O2 → NO3
− + 2H+ + H2O
NO3
− + CH3COO
− + H+ + H2O → NH4
+ + 2HCO3
−
2CHOO− + O2 → 2HCO3
−
NO3
− + 4CHOO− + 2H+ + H2O → NH4
+ + 4HCO3
−
2Fe2+ + 0.5O2 + 2H
+ → 2Fe3+ + H2O
CH4 + 2O2 → HCO3
− + H+ + H2O
0 20 40 60 80 100 120 140
Figure 4 | Chemical anity calculations for the SLW surficial (0–2 cm) sediment. Results are presented in energy density of joules per kg of water (J kg
H2O−1; top axis in log scale) and kilojoules per mole of electron transferred (kJmol e−
−1
; bottom axis) at 50% (0.5) and 10% (0.1) of the SLW lake water
O2 concentration for eight environmentally relevant biochemical reactions.
(Fig. 3)5. Although the pmoA primer set we used was designed
to detect a wide diversity of methanotrophs24, additional putative
methanotrophic genera were detected in the 16S rDNA and rRNA
community analysis (Supplementary Fig. 2), but these generawere at
least one order of magnitude less abundant thanM. tundripaludum.
Aerobic CH4 oxidizing bacteria are typically members of the
Gammaproteobacteria and Alphaproteobacteria25 and further
classified into different types based on the substrate affinity of their
methane monooxygenase enzyme25. Type Ia Gammaproteobacteria
methanotrophs have methane monooxygenase enzymes with low
affinity for CH4 while type II Alphaproteobacteria have enzymes
with a high affinity for CH4 (ref. 26). These type Ia Gammapro-
teobacteria methanotrophs, particularly Methylobacter sp., domi-
nate the active fraction of methanotroph populations in freshwater
environments that have high CH4 (µM–mM) concentrations and
strong CH4 sources25,26. M. tundripaludum possesses a low-affinity
(type Ia) methane monooxygenase enzyme, is known to be cold-
adapted24,26, has been shown to be active at the Greenland Ice Sheet
margin5 and is responsible for significant CH4 consumption in
a variety of other Arctic habitats27–29. Both the low CH4 affinity
and temperature adaptation of the type Ia Gammaproteobacteria
particulate methane monooxygenase enzyme reflect the conditions
measured in SLW surficial sediments (−0.5 ◦C and 0.1 to 0.3mM
CH4; Fig. 1)9. Indeed, a community analysis of 16S rRNAmolecules
showed M. tundripaludum and other methanotrophic taxa were
abundant (≥1.0%) in the water column and upper sediments
(0–6 cm), with their greatest relative abundance in the surficial
sediments (16%; Fig. 1b; Supplementary Fig. 2)21. These molecular
data, based on pmoA gene sequences and 16S rRNAmolecules, indi-
cate that methanotrophs related toM. tundripaludum are abundant
and potentially metabolically active near the SLW sediment–water
interface, where geochemical data indicate peakmethane oxidation.
The role of CH4 in the subglacial ecosystem
We computed chemical affinity (Ar) for the surficial (0–2 cm)
sediment layer to estimate the available biochemical energy from
CH4 oxidation compared to other potential metabolic reactions30,31
(Fig. 4). O2 concentration data in the surficial sediment layer are not
available, so biochemical reactions were modelled at half (36.5 µM)
and one-tenth (7.3 µM) of the average SLW water column O2
concentration. These modelled O2 concentrations are reasonable
given the evidence for O2 penetration to∼16 cm (ref. 14). Although
pyrite and ammonium oxidation are predicted to yield the greatest
metabolic energy in the water column32, aerobic CH4 oxidation is
the most exergonic biochemical pathway in the surficial sediment
despite the modelled tenfold reduction in O2 concentration relative
to lake water (Ae−r : 99.9 kJmol e
−−1 ; Akgr : 2.84 J kg H2O
−1) (Fig. 4).
Themicrobial community composition reflects the chemical affinity
calculations such that iron, sulfide and ammonium oxidizing taxa
are abundant in thewater column21,32 and aerobicmethane oxidizing
taxa are abundant and active in the surficial sediment (Fig. 1).
These chemical affinity calculations corroborate the molecular
and geochemical data by showing sufficient biochemical energy
is present in the SLW surficial sediment to support the abundant
methanotroph population (Fig. 4).
We modelled the rate of biological CH4 consumption in SLW as:
dC
dt
= (Fdiff×A)− (R×V ) (1)
where dC/dt is the change in CH4 concentration over time, Fdiff
is the diffusional flux into the 0–2 cm surficial sediment, A is the
area of SLW, R is the rate of CH4 consumption, and V is volume of
SLW plus the porewater surficial sediment. Assuming steady-state
conditions (that is, dC/dt=0), equation (1) can be rewritten as:
R= Fdiff
HL+ (HSS×ϕ) (2)
where HL and HSS are the height of the lake and surficial (0–2 cm)
sediments, respectively, and ϕ is the sediment porosity. R equates
to 3.0± 0.8mmol CH4 m−3 yr−1. The rate of CH4 removal (R) is
the sum of both CH4 oxidation (Rox) and incorporation of CH4 as
a carbon source (Rincorp) for microbial biomass synthesis. Using the
total CH4 removal rate (R), togetherwith the average fraction of CH4
(∼0.5) partitioned to biomass formation for type Imethanotrophs33,
reveals that methanotrophs may oxidize 1.5mmol CH4 m−3 yr−1
to CO2 (Rox) and assimilate 1.5mmol CH4 m−3 yr−1 (Rincorp) as a
biosynthetic carbon source (Supplementary Table 1). Given 0.5 as a
biomass partitioning factor, the rate of aerobic CH4 oxidation would
be 10- to 100-fold lower than aerobic CH4 oxidation measured
in cold (∼4 ◦C), surficial marine sediments and deep sea, CH4
seeps34,35. The biomass partitioning factor can vary from 0.06 to
0.7 in lakes with active methanotrophy36. When we account for
this potential variability in the biomass partitioning factor and
the uncertainty in the CH4 flux, Rox and Rincorp vary by an order
of magnitude; the range of Rincorp is 0.14–3.0mmol CH4 m−3 yr−1
and Rox is 0.52–3.6mmol CH4 m−3 yr−1 (Supplementary Table 1).
It is important to note that Rox and Rincorp are inversely related
(Supplementary Table 1). Although the overall rate of oxidation
may be low compared to marine sediment methanotrophy, if the
formation of biomass due to CH4 oxidation occurred solely in the
surficial SLW sediment porewaters, where molecular data indicate
peak active methanotroph abundance (Fig. 1b), the biosynthetic
rate would be 26.2 ngC (L porewater)−1 d−1 (range: 2.3–51 ngC
(L porewater)−1 d−1; Supplementary Table 1). This modelled
biomass C production rate via sedimentary methanotrophy is
nearly equivalent (80%; range: 7–155%) to measured rates of
chemoautotrophic biomass C production (32.9 ngC L−1 d−1) within
the SLW water column7. These results indicate that CH4, as
modelled, is an important carbon and energy source for the SLW
sediment microbial community.
The O2 demand derived from the modelled CH4 removal rate
(R) is 6.1× 105 mol O2 yr−1, using 0.5 as the biomass partitioning
factor. Methanotrophy in SLW is responsible for consuming∼16%
(range: 10–24%; Supplementary Table 1) of the O2 supply to the
SLW ecosystem32. Thus, the impact of oxygen demand due to
CH4 oxidation in the SLW ecosystem depends on the balance
between methanotroph growth and energy requirements. Despite
a potentially large range in the biomass partitioning factor, these
calculations show that O2 released from basal melting of the
overlying ice sheet fuels an abundant and active population of
methanotrophs in the lake. Saturated sediments at SLW are similar
in nature to those found beneath other ice streams of the Siple coast
region (for example, ref. 8) and basal ice melt is extensive beneath
the WAIS37,38, which may produce extensive oxic subglacial aquatic
habitats, conducive to cosmopolitan populations of methanotrophs
that convert CH4 to CO2 and biomass.
Our data reveal that hydrogenotrophic methanogenesis is
the main pathway of CH4 formation beneath SLW and that
CH4 is utilized by aerobic methanotrophic bacteria. Contrary to
previous predictions which suggested the potential significance
of subglacial CH4 fluxes to the atmosphere (for example, ref. 4),
our CH4 measurements and flux calculations show that aerobic
methanotrophic bacteria in SLW convert most (>99%) of the
sedimentary CH4 efflux to CO2 and biomass. The bacterial
conversion of CH4 to CO2 beneath the WAIS reduces the warming
potential of subglacial gases39 that may be released to downstream
ice sheet margin environments and to the atmosphere during
episodes of ice sheet retreat. Given the potential for widespread
hydrogenotrophic CH4 production in sediments beneath ice sheets,
such as the WAIS, and the release of O2 due to melting at the ice
sheet base9,37,38, biological transformations ofCH4 maybe significant
for the functioning and persistence of deep microbial life and
biogeochemical processes in Antarctic sub-ice environments.
Methods
Methods, including statements of data availability and any
associated accession codes and references, are available in the
online version of this paper.
Received 17 March 2017; accepted 26 June 2017;
published online 31 July 2017
References
1. Kirschke, S. et al . Three decades of global methane sources and sinks.
Nat. Geosci. 6, 813–823 (2013).
2. Thauer, R. K., Kaster, A-K., Seedorf, H., Buckel, W. & Hedderich, R.
Methanogenic archaea: ecologically relevant differences in energy
conservation. Nat. Rev. Microbiol. 6, 579–591 (2008).
3. Conrad, R. The global methane cycle: recent advances in understanding the
microbial processes involved. Environ. Microbiol. Rep. 1, 285–292 (2009).
4. Wadham, J. L. et al . Potential methane reservoirs beneath Antarctica. Nature
488, 633–637 (2012).
5. Dieser, M. et al . Molecular and biogeochemical evidence for methane cycling
beneath the western margin of the Greenland Ice Sheet. ISME J. 8,
2305–2316 (2014).
6. Wadham, J. L. et al . The potential role of the Antarctic Ice Sheet in global
biogeochemical cycles. Earth Environ. Sci. Trans. R. Soc. Edinburgh 104,
55–67 (2013).
7. Christner, B. C. et al . A microbial ecosystem beneath the West Antarctic Ice
Sheet. Nature 512, 310–313 (2014).
8. Skidmore, M. in Antarctic Subglacial Environments, Geophysical Monograph
Series (eds Siegert, M. J., Kennicutt II, M. C. & Bindschandler, R. A.)
61–81 (American Geophysical Union, 2011).
9. Fisher, A. T., Mankoff, K. D., Tulaczyk, S. M. & Tyler, S. W. High
geothermal heat flux measured below the West Antarctic Ice Sheet. Sci. Rep. 1,
e1500093 (2015).
10. Whiticar, M. J. Carbon and hydrogen isotope systematics of bacterial formation
and oxidation of methane. Chem. Geol. 161, 291–314 (1999).
11. Coleman, D. D., Liu, C.-L. & Riley, K. M. Microbial methane in the shallow
Paleozoic sediments and glacial deposits of Illinois, USA. Chem. Geol. 71,
23–40 (1988).
12. Conrad, R. in Advances in Agronomy (ed. Sparks, S.) 1–63 (Elsevier, 2007).
13. Schoell, M. Multiple origins of methane in the Earth. Chem. Geol. 71,
1–10 (1988).
14. Michaud, A. B. et al . Solute sources and geochemical processes in Subglacial
Lake Whillans, West Antarctica. Geology 44, 347–350 (2016).
15. Telling, J. et al . Rock comminution as a source of hydrogen for subglacial
ecosystems. Nat. Geosci. 8, 851–855 (2015).
16. Lin, L.-H., Slater, G. F., Sherwood Lollar, B., Lacrampe-Couloume, G. &
Onstott, T. C. The yield and isotopic composition of radiolytic H2, a potential
energy source for the deep subsurface biosphere. Geochim. Cosmochim. Acta
69, 893–903 (2005).
17. Matheus Carnevali, P. B. et al . Methane sources in arctic thermokarst lake
sediments on the North Slope of Alaska. Geobiology 13, 181–197 (2015).
18. Lever, M. A. et al . Evidence for microbial carbon and sulfur cycling in deeply
buried ridge flank basalt. Science 339, 1305–1308 (2013).
19. Blazewicz, S. J., Barnard, R. L., Daly, R. A. & Firestone, M. K. Evaluating rRNA
as an indicator of microbial activity in environmental communities: limitations
and uses. ISME J. 7, 2061–2068 (2013).
20. Jones, S. E. & Lennon, J. T. Dormancy contributes to the maintenance of
microbial diversity. Proc. Natl Acad. Sci. USA 107, 5881–5886 (2010).
21. Achberger, A. M. et al . Microbial community structure of Subglacial Lake
Whillans, West Antarctica. Front. Microbiol. 7, 1457 (2016).
22. Grant, N. J. & Whiticar, M. J. Stable carbon isotopic evidence for methane
oxidation in plumes above Hydrate Ridge, Cascadia Oregon Margin. Global
Biogeochem. Cycles 16, 1124 (2002).
23. Coleman, D. D., Risatti, J. B. & Schoell, M. Fractionation of carbon and
hydrogen isotopes by methane-oxidizing bacteria. Geochim. Cosmochim. Acta
45, 1033–1037 (1981).
24. McDonald, I. R., Bodrossy, L., Chen, Y. & Murrell, J. C. Molecular ecology
techniques for the study of aerobic methanotrophs. Appl. Environ. Microbiol.
74, 1305–1315 (2008).
25. Knief, C. Diversity and habitat preferences of cultivated and uncultivated
aerobic methanotrophic bacteria evaluated based on pmoA as molecular
marker. Front. Microbiol. 6, 1346 (2015).
26. Ho, A. et al . Conceptualizing functional traits and ecological characteristics of
methane-oxidizing bacteria as life strategies. Environ. Microbiol. Rep. 5,
335–345 (2013).
27. Martineau, C., Whyte, L. G. & Greer, C. W. Stable isotope probing
analysis of the diversity and activity of methanotrophic bacteria in soils
from the Canadian high Arctic. Appl. Environ. Microbiol. 76,
5773–5784 (2010).
28. Graef, C., Hestnes, A. G., Svenning, M. M. & Frenzel, P. The active
methanotrophic community in a wetland from the High Arctic. Environ.
Microbiol. Rep. 3, 466–472 (2011).
29. He, R. et al . Shifts in identity and activity of methanotrophs in Arctic lake
sediments in response to temperature changes. Appl. Environ. Microbiol. 78,
4715–4723 (2012).
30. Shock, E. L. et al . Quantifying inorganic sources of geochemical energy in
hydrothermal ecosystems, Yellowstone National Park, USA. Geochim.
Cosmochim. Acta 74, 4005–4043 (2010).
31. Amend, J. P. & Shock, E. L. Energetics of overall metabolic reactions of
thermophilic and hyperthermophilic Archaea and Bacteria. FEMS Microbiol.
Rev. 25, 175–243 (2001).
32. Vick-Majors, T. J. et al . Physiological ecology of microorganisms in Subglacial
Lake Whillans. Front. Microbiol. 7, 1705 (2016).
33. Trimmer, M. et al . Riverbed methanotrophy sustained by high carbon
conversion efficiency. ISME J. 9, 2304–2314 (2015).
34. Iversen, N. & Blackburn, T. H. Seasonal rates of methane oxidation in anoxic
marine sediments. Appl. Environ. Microbiol. 41, 1295–1300 (1981).
35. Marlow, J. J. et al . Carbonate-hosted methanotrophy represents an
unrecognized methane sink in the deep sea. Nat. Commun. 5, 5094 (2014).
36. Bastviken, D., Ejlertsson, J., Sundh, I. & Tranvik, L. Methane as a
source of carbon and energy for lake pelagic food webs. Ecology 84,
969–981 (2003).
37. Lough, A. C. et al . Seismic detection of an active subglacial magmatic complex
in Marie Byrd Land, Antarctica. Nat. Geosci. 6, 1031–1035 (2013).
38. Beem, L. H., Jezek, K. C. & Van Der Veen, C. J. Basal melt rates beneath
Whillans Ice Stream, West Antarctica. J. Glaciol. 56, 647–654 (2010).
39. Yvon-Durocher, G. et al . Methane fluxes show consistent temperature
dependence across microbial to ecosystem scales. Nature 507,
488–491 (2014).
Acknowledgements
This study was funded by National Science Foundation – Division of Polar Programs
grants (0838933, 1346250, 1439774 to J.C.P.; 0838941 to B.C.C.) awarded as part of the
Whillans Ice Stream Subglacial Access Research Drilling (WISSARD) project. We thank
the WISSARD Science Team (see http://wissard.org for the full list of team members) for
their assistance in expedition planning and with collecting and processing samples.
Partial support was provided by graduate fellowships from the NSF-IGERT Program
(0654336), Montana Space Grant Consortium and NSF-Center for Dark Energy
Biosphere Investigations (A.B.M.); a dissertation grant from the American Association of
University Women (T.J.V.-M.); a NSF-Graduate Research Fellowship (A.M.A.); and a Sêr
Cymru National Research Network for Low Carbon, Energy and the Environment Grant
from the Welsh Government and Higher Education Funding Council for Wales (A.C.M.).
We thank R. Scherer and R. Powell for sediment cores. B.B. Jørgensen, M. A. Lever and
S. Nielsen provided support and assistance with DNA extraction and pmoA/mcrA
amplification. Logistics were conducted by the 139th Expeditionary Airlift Squadron of
the New York Air National Guard, Kenn Borek Air, and Antarctic Support Contractor,
managed by Lockheed-Martin. Hot-water drill support was provided by University of
Nebraska-Lincoln and directed by F. Rack and D. Duling (chief driller). D. Blythe,
J. Burnett, C. Carpenter, D. Gibson, J. Lemery, A. Melby and G. Roberts provided drill
support at SLW. This is C-DEBI contribution #371.
Author contributions
A.B.M., J.E.D., T.J.V.-M., J.C.P. and M.L.S. wrote the manuscript. A.B.M., J.E.D., M.L.S.
and T.J.V.-M. conducted and analysed methane concentration and isotopic data. A.M.A.,
A.B.M. and B.C.C. processed, analysed and interpreted the molecular data. A.C.M.
conducted thermodynamic calculations. All authors contributed to the study design,
collection of samples and approved the final draft of the manuscript.
Additional information
Supplementary information is available in the online version of the paper. Reprints and
permissions information is available online at www.nature.com/reprints. Publisher’s note:
Springer Nature remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations. Correspondence and requests for materials should be
addressed to A.B.M. or J.C.P.
Competing financial interests
The authors declare no competing financial interests.
Methods
Sample collection.We used a microbiologically clean hot-water drill to directly
sample the water column and the upper 40 cm of sediment of Subglacial Lake
Whillans (SLW; 84.240◦ S, 153.694◦W) to assess the CH4 dynamics40,41. SLW water
column and sediment were sampled through a 800-m-deep,∼0.6-m-diameter
borehole on 30 January 2013. The clean access hot-water drill system has been
shown to reduce cell concentrations within the drilling water to<100 cellml−1,
which is acceptable based on the predicted cell concentration in the lake water and
the National Research Council 2007 report on subglacial lake access40,42. The
2.2-m-deep SLW water column was sampled with a 10 l Niskin bottle, suspended
microbial cells were concentrated using an in situ water filtration system, and
surficial sediments were collected with a gravity multicorer (60 cm long× 6 cm
diameter). For complete drilling and sampling details see ref. 40,41.
Geochemical analysis. Sediment from a gravity core (MC-2A) was sampled every
2 cm by extrusion and subsampling of each newly exposed layer. Sediment
subsamples for methane (CH4) were collected using a sterile cut-off 5ml syringe
and immediately placed into 20ml sterile serum vials and stoppered with a sterile
butyl rubber stopper, then crimped with an aluminium cap. Three empty vials were
sealed in the field to capture atmospheric air as procedural blanks. Ten ml of 2.5%
NaOH was added by sterile syringe to each sample vial and the three blanks,
stopping biological activity and creating a pressurized headspace within each vial43.
A CH4 sample from the SLW water column was collected from cast 1 from a Niskin
bottle by placing the tube to the bottom of the serum vial and filling from top to
bottom. The water sample was fixed with Lugol’s solution to prevent biological
activity. All vials were stored inverted at 4 ◦C for transport back to Montana State
University (MSU) for CH4 quantification. Headspace CH4 was quantified on a
Hewlett-Packard 5890 Series II gas chromatograph (GC) equipped with a flame
ionization detector (FID) with a detection limit of 3 nM for water column samples
and 190 nM for the sediment samples. Headspace gas was introduced to the GC
using a 10-port injection valve configured for back flushing of a precolumn
(25 cm× 0.32 cm OD, packed with Porapak-T 80/100 mesh) to prevent water
vapour from reaching the analytical columns. The vial overpressure was used to
flush and fill a 1 cm3 sample loop using a syringe needle inlet; measured laboratory
air temperature and pressure were used to calculate the total moles of gas contained
within the loop, assuming gas ideality. Gases were separated on two analytical
columns in series (both 183 cm× 0.32 cm OD, packed with
Chromosorb 102 80/100 mesh and Porapak-Q 80/100 mesh, respectively). The
columns were maintained at 55 ◦C and the FID at 240 ◦C. The carrier gas was an
ultra-high purity N2, which was further purified through Molecular Sieve 5A,
activated charcoal and an O2 scrubber. The carrier flow was 30mlmin−1; under
these conditions, CH4 eluted to the FID at 1.97min. Instrument calibration was
performed using certified 500 and 51 ppmv CH4 in air standards (Air Liquide;±1% accuracy), and volumetric dilutions thereof into carrier N2. Dissolved CH4
concentrations were calculated using Henry’s law based on measured headspace
mole fractions and Bunsen solubility coefficients estimated from temperature and
sample salinity (including added NaOH) as parameterized by ref. 44. Porewater
volumes were determined from mass loss after drying the sediment at 95 ◦C until
the mass stopped decreasing (∼24 h), and dry sediment volume was similarly
determined assuming a density of 2.60 g cm−3 for the sedimentary particles45. The
total volume of the vials was determined weighing the vials with sediment and
NaOH fixative, then completely filling the headspace with deionized water and
weighing again. The headspace volume was determined by difference. The extent of
pressurization of the headspace was determined from total headspace volume and
the volume of NaOH solution added. The total CH4 within each vial, after
correction for the small amount of CH4 present in the headspace air when
originally sealed (characterized by the blank vials), was then used to determine the
initial CH4 concentration of the porewater.
Gravity core MC-3A was collected from SLW, capped and immediately frozen
(−20 ◦C). The frozen core was returned to MSU and thawed at 4 ◦C overnight in a
class 1,000 clean, cold room in the MSU SubZero Science and Engineering Facility.
The core was extruded and cut every 2 cm, and sediment for CH4 stable isotope
analysis was subsampled and fixed using the same method as for CH4
concentration analysis fromMC-2A described above. One ml of room-temperature
headspace gas from the fixed sediment vials was transferred to a gas-tight
laminated foil bag using a gas-tight, glass syringe and diluted 1:100 with CH4-free
(zero grade) air. The bag was connected to the inlet of a Picarro G2201-i Cavity
Ring-Down Spectrometer (CRDS) specific for high-precision concentration and
δ13C analyses of CH4. Sample was introduced to the instrument at a flow rate of
100mlmin−1; δ13C–CH4 values were determined using factory calibrations and
were averaged over≥30 s of 1Hz measurements. Between samples, atmospheric air
was measured for at least 5min to ensure lack of instrument drift. The δD–CH4
values were measured at the University of California Davis Stable Isotope Facility
(UCD-SIF) using a PreCon concentration system (ThermoScientific) in line with a
Delta V plus isotope ratio mass spectrometer (ThermoScientific)46. Two δ13C–CH4
samples (MC-3A samples from 18 to 20 cm and 34–36 cm depths) were also run at
UCD-SIF to compare their independent results with our values obtained on the
Picarro CRDS. There was a<4% difference in the δ13C–CH4 values reported from
the two methods. The carbon and hydrogen stable isotope ratios are reported in
δ-notation (δ13C, δD) with respect to Vienna Pee Dee Belemnite (VPDB) and
Vienna Standard Mean Ocean Water (VSMOW) standards, respectively. The
running average (with depth) of the CH4 concentration and isotope values was
calculated in SigmaPlot v. 11 using a locally estimated scatterplot smoothing
(Loess) function with smoothing parameters set to first-degree polynomial and a
sampling frequency of 0.45, which determines the number of local data points used
in the weighted regression carried out by the Loess smoothing function.
Sediments used for dissolved NH+4 concentration measurements were collected
fromMC-3A (ref. 47). The sediment was transferred to acid-washed (10% HCl),
ultra-pure water-rinsed (6×), combusted (4 h at 450 ◦C) glass vials with
polytetrafluoroethylene lined caps, frozen at−20 ◦C and thawed prior to analysis.
Sediments were transferred from the glass vials to acid-washed and ultra-pure
water-rinsed 50ml conical centrifuge tubes and centrifuged at 3,500g for 20min.
The supernatant was transferred to acid-washed and ultra-pure water-rinsed 15ml
conical centrifuge tubes and spun for an additional 20min at 4,500g to pellet
fine particulates. The clean supernatant from the 15ml centrifuge tube was
transferred to an acid-washed and ultra-pure water-rinsed glass vial. The
supernatant was diluted (1:10) to a final volume of 5ml with ultra-pure water for
colorimetric analysis48.
Particulate organic carbon and nitrogen values were determined with an
elemental analyser as described in ref. 1. Acetate, formate and oxalate
concentrations were determined using ion chromatography, following methods
in ref. 14.
Molecular analyses. DNA was extracted using a modular method to allow for
optimization of the DNA extraction procedure, specific to the SLW sediments49.
DNA extraction yield from SLW sediments was greatest when sediments were
pre-treated with 450 µmol g−1 deoxynucleotide triphosphate to prevent
adsorption of lysed DNA to the abundant clay particles in SLW49. The particulate
methane monooxygenase (pmoA) gene clone libraries were constructed by
Polymerase Chain Reaction (PCR) amplification using A189F (5′-GGNGACTGG
GACTTCTGG-3′) and m680R (5′-CCGGMGCAACGTCYTTACC-3′)24. The PCR
was set up using 0.13 µl of ExTaq at 5 units µl−1 (Takara), 2.5 µl of 10× ExTaq
buffer (Takara), 2 µl dNTP mixture at 2.5mM per nucleotide (Takara), 2.5 µl of
A189F and Mb661R primers (10 pmol µl−1), 2 µl molecular biology-grade bovine
serum albumen (BSA; 1.6mgml−1 final concentration) (New England BioLabs),
4 µl of template DNA (0.01–0.09 ng DNA µl−1), and 11.37 µl of PCR-grade water
for a final reaction volume of 25 µl. The PCR thermocycling conditions were 1 cycle
of 98 ◦C for 2min; 40 cycles of 98 ◦C for 15 s, 55 ◦C for 1min, and 72 ◦C for 1min;
followed by a final 72 ◦C for 7min. PCR was conducted with DNA extraction
blanks and no template blanks (PCR-grade water) as negative controls. Negative
controls were not carried forward for cloning, as no PCR bands were detected. PCR
products were run on a 1.5% agarose gel and the 491 basepair pmoA fragment was
excised from the gel with sterile razor blade and DNA was purified using a Wizard
SV gel clean-up system (Promega). Cleaned pmoA fragments were immediately
ligated and cloned with a TA Cloning kit (Invitrogen). Positive clones were
transferred to LB+ampicillin broth and grown overnight at 37 ◦C, then sequenced
(288 total sequenced clones) (Functional Bioscience). The pmoA DNA sequences
were processed by removing the forward and reverse primer sequences and
removing poor-quality sequences (<20 phred score)50. Quality controlled pmoA
sequences (176 total) were clustered into operational taxonomic units (OTUs) at
the 97% similarity level and one representative sequence from each OTU51, along
with representative pmoA sequences from type Ia and II methanotrophs24, were
aligned using ClustalW using the default alignment parameters within the program
MEGA6 (ref. 52). A phylogenetic tree was built using the neighbour-joining
method with 1,000 bootstrap replications52. The pmoA sequences have been
deposited in GenBank under accession numbers KX589304-KX589461 and
KX784213-KX84230.
We attempted to amplifymcrA gene sequences from SLW sediment DNA
extracts using a primer set designed to amplify the diversity ofmcrA-containing
methanogens18 with a nested PCR amplification scheme. The primer pair used to
detect themcrA gene sequence were mcrIRD18. The primer pair is capable of
detecting a wide diversity of known and several novelmcrA gene clusters18.
The first reaction was set up using 0.13 µl of Takara ExTaq at 5 units µl−1, 2.5 µl
10× ExTaq buffer, 2 µl dNTP mixture at 2.5mM per nucleotide (Takara), 2.5 µl of
forward and reverse primer (10 pmol µl−1), 2 µl of BSA (1.6mgml−1 final
concentration), 9.38 µl PCR-grade water and 4 µl DNA extract (0.01–0.09 ng DNA
µl−1) for a final reaction volume of 25 µl. This first reaction was run with an initial
denaturation step at 98 ◦C for 2min followed by 40 cycles of 98 ◦C for 15 s, 53 ◦C
for 1min and 72 ◦C for 1min, and a final elongation at 72 ◦C for 7min. The second
reaction was set up using 0.25 µl Takara ExTaq, 5 µl 10x ExTaq buffer, 4 µl dNTP
mixture at 2.5mM per nucleotide (Takara), 5 µl of forward and reverse primers
(10 pmol µl−1), 4 µl of BSA (1.6mgml−1 final concentration), 21.75 µl of
PCR-grade water and 4 µl of product from the first reaction as template DNA. The
second reaction was run with the same thermocycler program as the first reaction.
PCR was conducted with DNA extraction blanks and no template blanks
(PCR-grade water) as negative controls. Details of the 16S rRNA molecule sample
collection and preservation, extraction, reverse transcription, sequencing and
processing are described in ref. 21. Extraction blanks were conducted, processed
and analysed in parallel with the SLW sediment samples as described in ref. 21.
Chemical affinity calculations. An assessment of CH4 as a potential chemical
energy source for the surficial (0–2 cm) sediment layer was undertaken. The
chemical affinity of coupled oxidation–reduction reactions involving CH4 and
other potential metabolic reactions was determined. The chemical affinity (Ar) is
the maximum amount of energy that can be obtained for a reaction based on in situ
conditions. Ar is defined as the change in the overall Gibbs energy under
non-equilibrium conditions (1Gor ) with a change in the progress of the reaction,
which quantifies the reactions proximity to equilibrium30,31 and is given by:
Ar=RT ln(Kr/Qr) (3)
where Kr is the calculated equilibrium constant for the reaction, which is derived
from1Gor of the reaction according to1G
o
r =Gof products—Gof reactants, where Gof
is the standard Gibbs energy of formation for the products and reactants53. Kr is
given by:
Kr=e−1Gr o/RT (4)
where R is the gas constant 0.008314 kJmol−1, and T is SLW temperature in Kelvin
[−0.5 ◦C=272.65K] (ref. 53). Thermodynamic values were derived from ref. 31
using values for 2 ◦C, the closest available values for the temperature of SLW
(−0.5 ◦C); the impact of the temperature difference on1Gor and resulting Kr values
will be small30,31.
Qr is the activity product for the reaction, determined as;
Qr=
∏
i
(ai)Vi,r (5)
where ai represents the activity of the ith compound in the reaction raised to its
stoichiometric coefficient in the rth reaction, vi,r , which is positive for products and
negative for reactants. Activities are calculated from molal concentrations (mi)
using activity coefficients (γi) and the relationship ai=miγi (ref. 30). These
activities were calculated using the geochemical model PHREEQC54 using the
empirical SLW geochemistry7,14. The O2 concentration in the 0–2 cm layer was not
measured, but for the chemical affinity calculations we consider two scenarios of
O2 concentration set at 50% (36.5 µM) and 10% (7.3 µM) of average SLW lake
water to account for the decrease in sedimentary O2 concentration due to
consumption and diffusion55. Given that oxygen is inferred to penetrate to∼16 cm
based on redox-sensitive trace metal concentrations14, it is reasonable to model
chemical affinity using these two concentrations of O2 in the surficial sediment.
Temperature, pH, redox (pE) and concentrations of acetate, formate
(Supplementary Fig. 1), dissolved inorganic carbon (DIC), O2 (aq), CH4 (aq),
SO2−4 , NO
−
3 , NH
+
4 , total dissolved Fe, Ca
2+, Mg2+, Na+, K+, P, Li+, Br−, Cl− and
F− were defined7,14,32,47. Redox-sensitive elements that were measured as total
dissolved elemental concentration (that is, C, Fe) were assumed to be speciated to
the redox states and species activities as determined by PHREEQC. Conversely,
ions measured in specific redox states (that is, SO2−4 , NO
−
3 , NH
+
4 ) were maintained
in their respective redox states by the model, and the species activities including
these ions were calculated.
The chemical affinities are expressed in per electron yields (Ae−r ) and also
shown in terms of energy density, the energy per kg H2O (Akgr ), which scales the
energy availability to the limiting reactant, calculated as:
Akgr =
∣∣∣∣Arvi
∣∣∣∣ [i] (6)
where [i] refers to the concentration of the limiting electron donor or acceptor56.
This scaling (equation (6)) of chemical affinity has been shown to better correlate
with actual microbial communities and metabolisms than the chemical affinity
normalized to moles of electrons transferred56,57.
Methane oxidation rate modelling. CH4 oxidation rates were modelled by
calculating the flux of CH4 into the 0–2 cm sediment layer. The CH4 concentration
gradient was determined using CH4 values from 15 cm to 3 cm. The flux was
calculated using Fick’s first law and the error of the flux determined from the error
associated with the diffusional gradient. Water content was measured and
calculated by weighing a known volume of wet weight sediment, then measuring
the sediment again after drying at 95 ◦C for three days43,45. Porosity was calculated
from the water content and density of the sediment43,45. The diffusion coefficient
for CH4 at 0 ◦C was corrected for porosity (Supplementary Fig. 3) and tortuosity of
SLW sediments calculated according to equation (3.11) from ref. 58 with C= 2.02
(refs 58,59). We modelled the rate of biological CH4 consumption according to
equation (1) (see main text).
The control volume of our model can be defined by the relationship:
V=A×HL+ (HSS×ϕ) (7)
where HL and HSS are the height of the lake and surficial sediments, respectively,
and ϕ is the sediment porosity. Assuming steady-state conditions (that is,
dC/dt=0) and substitution of equation (7) into equation (1), R can be calculated
as shown in equation (2). R represents the sum of both microbial CH4 oxidation to
CO2 and incorporation of CH4 into biomass. We estimated the amount of CH4
removal due to oxidation and incorporation of biomass by assuming that the
biomass partitioning factor of CH4 going to biomass is 0.5 (x ; equation (8)). The
value of 0.5 has been shown to be a good approximation for the fraction of biomass
incorporated by type I methanotrophs during CH4 oxidation and is a median value
across many habitats33,60,61. We calculated the impact of varying x from 0.06 to 0.77
(ref. 36) on biomass C production and methanotrophy oxygen demand
(Supplementary Table 1). From the CH4 removal rate and the fraction of CH4
incorporated into biomass, we can then calculate the O2 consumption by CH4
oxidation, which follows the stoichiometric relationship:
CH4+ (2−x)O2→ (1−x)CO2+xCH2O+ (2−x)H2O (8)
where x is the fraction of CH4 partitioned into biomass formation33,60,62. The inputs
of O2 to the lake are from atmospheric gases released by melting of the overlying
meteoric ice and advection of water into the lake during the filling phase of the
hydrologic cycle9,32,63. Based on the concentration of gas in the overlying ice and the
basal ice melt rate, which has been estimated at 1.8 cm yr−1 (ref. 9), the overlying
ice sheet supplies 1.0×106 mol O2 yr−1 (67% of O2 supply to SLW)32. Advection
into the lake provides 5×105 mol O2 yr−1 (33% of O2 supply to SLW)32, assuming
the incoming water has the same concentration measured in the SLW water
column32,63. When the fraction of carbon from CH4 going to biomass is varied
(Supplementary Table 1), the oxygen demand on the system changes as well. We
used the SLW oxygen budget from ref. 32 to determine the impact the biomass
partitioning factor (x) could have on the oxygen demand for biological processes in
SLW (Supplementary Table 1).
Data availability. Data generated for this study are available through the Microbial
Antarctic Resource System database (http://mars.biodiversity.aq/resources/97).
Molecular data were accessed from NCBI Sequence Read Archive
(https://www.ncbi.nlm.nih.gov/sra) project PRJNA244335.
References
40. Priscu, J. C. et al . A microbiologically clean strategy for access to the Whillans
Ice Stream subglacial environment. Antarct. Sci. 11, 1–11 (2013).
41. Tulaczyk, S. et al . WISSARD at Subglacial Lake Whillans, West Antarctica:
scientific operations and initial observations. Ann. Glaciol. 55,
51–58 (2014).
42. National Resource Council Exploration of Antarctic Subglacial Aquatic
Environments (The National Academies Press, 2007).
43. Riedinger, N. et al . Methane at the sediment–water transition in Black Sea
sediments. Chem. Geol. 274, 29–37 (2010).
44. Wiesenburg, D. A. & Guinasso, N. L. Jr Equilibrium solubilities of methane,
carbon monoxide, and hydrogen in water and sea water. J. Chem. Eng. Data 24,
356–360 (1979).
45. Avnimelech, Y., Ritvo, G., Meijer, L. E. & Kochba, M. Water content, organic
carbon and dry bulk density in flooded sediments. Aquacult. Eng. 25,
25–33 (2001).
46. Yarnes, C. 13C and 2Hmeasurement of methane from ecological and geological
sources by gas chromatography/combustion/pyrolysis isotope-ratio mass
spectrometry. Rapid Commun. Mass Spectrom. 27, 1036–1044 (2013).
47. Vick-Majors, T. J. Biogeochemical Processes in Antarctic Aquatic Environments:
Linkages and Limitations (Montana State University, 2016).
48. Solorzano, L. Determination of ammonia in natural waters by the
phenolhypochlorite method. Limnol. Oceanogr. 14, 799–801 (1969).
49. Lever, M. A. et al . A modular method for the extraction of DNA and RNA, and
the separation of DNA pools from diverse environmental sample types. Front.
Microbiol. 6, 476 (2015).
50. Paegel, B. M., Emrich, C. A., Wedemayer, G. J., Scherer, J. R. & Mathies, R. A.
High throughput DNA sequencing with a microfabricated 96-lane capillary
array electrophoresis bioprocessor. Proc. Natl Acad. Sci. USA 99,
574–579 (2002).
51. Schloss, P. D. et al . Introducing mothur: open-source, platform-independent,
community-supported software for describing and comparing microbial
communities. Appl. Environ. Microbiol. 75, 7537–7541 (2009).
52. Tamura, K., Stecher, G., Peterson, D., Filipski, A. & Kumar, S. MEGA6:
molecular evolutionary genetics analysis version 6.0.Mol. Biol. Evol. 30,
2725–2729 (2013).
53. Stumm, W. & Morgan, J. J. Aquatic Chemistry: Chemical Equilibria and Rates in
Natural Waters (Wiley-Interscience, 1996).
54. Parkhurst, D. L. & Appelo, C. A. J. US Geological Survey Techniques and
Methods 497 (US Geological Survey, 2013).
55. Boetius, A. & Wenzhöfer, F. Seafloor oxygen consumption fuelled by methane
from cold seeps. Nat. Geosci. 6, 725–734 (2013).
56. LaRowe, D. E. & Amend, J. P. inMicrobial Life of the Deep Biosphere
(eds Kallmeyer, J. & Wagner, D.) 279–302 (Walter de Gruyter, 2014).
57. Osburn, M. R. et al . Chemolithotrophy in the continental deep subsurface:
Sanford Underground Research Facility (SURF), USA. Front. Microbiol. 5,
610 (2014).
58. Shen, L. & Chen, Z. Critical review of the impact of tortuosity on diffusion.
Chem. Eng. Sci. 62, 3748–3755 (2007).
59. Broecker, W. S. & Peng, T.-H. Gas exchange rates between air and seal. Tellus
26, 21–35 (1974).
60. Shelley, F., Abdullahi, F., Grey, J. & Trimmer, M. Microbial methane
cycling in the bed of a chalk river: oxidation has the potential to match
methanogenesis enhanced by warming. Freshw. Biol. 60,
150–160 (2015).
61. Whalen, S. C., Reeburgh, W. S. & Sandbeck, K. A. Rapid methane
oxidation in a landfill cover soil. Appl. Environ. Microbiol. 56,
3405–3411 (1990).
62. Urmann, K., Lazzaro, A., Gandolfi, I., Schroth, M. H. & Zeyer, J. Response of
methanotrophic activity and community structure to temperature changes in a
diffusive CH/O counter gradient in an unsaturated porous medium. FEMS
Microbiol. Ecol. 69, 202–212 (2009).
63. Siegfried, M. R., Fricker, H. A., Carter, S. P. & Tulaczyk, S. Episodic ice velocity
fluctuations triggered by a subglacial flood in West Antarctica. Geophys. Res.
Lett. 43, 2640–2648 (2016).