Permafrost and Periglac. Process. 14: 367–374 (2003)
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ppp.468
Simulation of Freezing-thawing Cycles in a Permafrost Microcosm for
Assessing Microbial Methane Production under Extreme Conditions
D. Wagner,* C. Wille, S. Kobabe and E.-M. Pfeiffer
Alfred Wegener Institute for Polar and Marine Research, Potsdam, Germany
The microbial process of methane (CH4) production during the back-freezing of permafrost soils in
autumn and the future fate of produced CH4 in the thawing phase of the following spring are not well
understood. Long-term CH4 flux studies in the Lena Delta (Siberia) indicate that back-stored CH4 adds
to the emission of newly-produced CH4 at the beginning of the vegetation period. Further field analysis
shows that microbial CH4 production already occurs at in situ temperatures of around 1 C in the
bottom layer of the soil. Therefore, a permafrost microcosm was developed to simulate the influence of
the annual freezing-thawing cycles on the CH4 fluxes in the active layer of permafrost soils. Two
cryostats ensure independent freezing and thawing the top and the bottom of the microcosm to
simulate different field conditions. The CH4 concentration (Rhizon soil moisture samplers), the soil
temperature (film platinum resistance temperature detectors [RTDs]) and the soil water content (time
domain reflectometry) are analysed in different depths of the microcosm during the simulation in
addition to the concentration of emitted CH4 in the headspace of the microcosm. The data obtained
contribute to the understanding of microbial processes and CH4 fluxes in permafrost environments in
the autumn and early winter. Copyright # 2003 John Wiley & Sons, Ltd.
methane production; winter methane fluxes; freezing-thawing cycles; permafrost microcosm
High-latitude wetlands are natural sources of methane
(CH4), which is one of the important climate-relevant
trace gases in the Earth’s atmosphere (Wuebbles and
Hayhoe, 2002). Methane emission rates from tundra
environments show large variations, ranging between
20 and 40 Tg CH4 yr1 (Christensen et al., 1996).
Winter methane fluxes have been estimated only in
North America and West Siberia (Whalen and
Reeburgh, 1988; Dise, 1992; Melloh and Crill, 1996;
Panikov and Dedysh, 2000). The reported winter emission rates amount to between 4–41% of the annual
methane fluxes. Friborg et al. (1997) observed a drastic
* Correspondence to: D. Wagner, Alfred Wegener Institute
for Polar and Marine Research, Telegrafenberg A 43, 14473
Potsdam, Germany. E-mail: [email protected]
Copyright # 2003 John Wiley & Sons, Ltd.
increase of CH4 release from a subarctic mire during
the thawing period, which reached approximately
25% of the mid-summer flux. The close relationship
between CH4 fluxes and the microbiological processes
of CH4 production and oxidation in permafrost soils
was reported by Wagner et al. (2003).
Methane fluxes from natural wetlands are basically
caused by two microbiological processes: (i) methane
production by methanogenic archaea in the anaerobic
soil horizons and (ii) methane oxidation by methanotrophic bacteria in the aerobic soil horizons (Hanson
and Hanson, 1996; Garcia et al., 2000). In permafrost
habitats microbial activity is influenced by extreme
gradients in temperature, moisture and chemistry (see
review in Wagner et al., 2001b). In spite of the extreme
conditions of permafrost soils, methane production
was revealed by in situ studies of the active layer
with temperature ranges between about 10 C to
Received 15 February 2003
Revised 26 June 2003
Accepted 4 July 2003
D. Wagner et al.
1 C (Samarkin et al., 1999; Wagner et al., 2003).
However, methane production and oxidation rates
during back-freezing of the active layer in autumn
and the future fate of produced methane in the thawing
phase of the following spring are not well understood.
In this paper, we describe a new technique for the
simulation of natural freezing-thawing cycles of the
active layer in a permafrost microcosm. The core
temperature, moisture and methane concentration
and release in different depths of the microcosm can
be measured continuously under controlled freezing
and thawing conditions. The simulation experiment
provides an insight into the activity of microorganisms and turnover rates under various cryogenic conditions (i.e. frozen, partly frozen, unfrozen).
Field Investigations
The field investigations during the expeditions LENA
1999 and LENA 2000 (Pfeiffer et al., 2000; Wagner
et al., 2001a) were carried out on Samoylov island
(72 22 N, 126 28 E) located in the Lena Delta,
Siberia. The study site represents an area of typical
polygonal patterned ground with ice-wedges. The
permafrost soils are classified according to the US
Soil Taxonomy (Soil Survey Staff, 1998) as Glacic
Aquiturbels and Typic Historthels with a maximum
thaw depth of between 30 and 55 cm. The average air
temperature in 1999 was 14.7 C with a minimum
in January (47.8 C) and a maximum in July
(þ18.3 C).
Methane flux measurements were conducted daily
at midday between the end of May and the beginning
of September 1999, and in August 2000. Five static
chambers (PVC transparent, 12.5 l) installed close to
each other in a polygon depression were used. The
condition of the chambers, the realization of the
measurements and the site characteristics are described in detail by Wagner et al. (2003). Methane
emissions were calculated from the chamber volume
and the linear increase in methane concentration.
Microbial methane production was studied in the
bottom layer of the polygon depression in July 1999.
Fresh soil material (20 g) from the bottom layer was
weighed into 100-ml glass jars and closed with a
screw cap containing a septum. The samples were
evacuated and flushed with ultra pure N2. The prepared soil samples were re-installed in the bottom
layer of the soil profile from which the samples had
been taken. Three replicates were used. Gas samples
were taken every 24 h out of the jars headspace with a
Copyright # 2003 John Wiley & Sons, Ltd.
gastight syringe. Methane production rates were calculated from the linear increase in CH4 concentration
analysed by gas chromatography.
Simulation Experiment
Back-freezing and thawing of the active layer was
simulated with an upgraded apparatus formerly applied for studying changes of the soil structure as a
function of freezing processes (Müller-Lupp, 2002).
The new configuration of the microcosm allows measurement of the in situ CH4 concentration (or the
concentration of other relevant gases like CO2) at
different core depths and the concentration of the
emitted CH4 in the headspace of the microcosm.
The data obtained can be used for the calculation of
CH4 turnover rates.
The permafrost microcosm was prepared from an
undisturbed soil core including the plant cover, taken
from the active layer of a permafrost soil (e.g. Typic
Historthel) in the depression of a low-centred polygon
in the Lena Delta, Siberia. The soil core was characterized by its natural structure, pore system and
water content. Different states of the active layer
(frozen, partly frozen, unfrozen) and varying freezing
times were simulated. Variations in the duration of
freeze-thaw cycles and their influence on the microbiological activity can then be studied. During the
experiment, the temperature, moisture and CH4 concentration were determined at 5-cm intervals of the
The permafrost microcosm consists of a Plexiglas
tube accommodating the undisturbed soil core
(Figure 1). The tube is closed gastight by an upper
Figure 1 Schematic view of the permafrost microcosm (height ¼
30 cm, inner diameter ¼ 10 cm).
Permafrost and Periglac. Process., 14: 367–374 (2003)
Microbial Methane Production 369
and lower flange plate. The Plexiglas tube has a length
of 30 cm, an inner diameter of 10 cm and a wall
thickness of 0.5 cm. One side of the tube wall was
cut open and the slot was filled with a polyurethane
elastomer. In this way, the tube wall can be exactly
adjusted to the undisturbed permafrost core using pipe
clamps without influencing the gastightness of the
microcosm. The flanges consist of two aluminium
plates with a cooling coil between. The aluminium
plates facing the soil core incorporate a rubber seal to
make a gastight connection to the Plexiglas tube.
For the insertion of temperature sensors, water
content probes (time-domain-reflectometry), and gas
samplers, the Plexiglas tube is prepared with compatible holes. The holes are placed at a distance of 120
around the tube, while the vertical distance between
the holes is 5 cm. In order to establish a gastight
connection to the Plexiglas tube, the temperature
sensors and gas samplers are glued into conical rubber
stoppers which are plugged into the holes. The TDR
probes have special thread flanges which are screwed
into the tube wall and sealed with a polyurethane
The installation of the permafrost microcosm is
carried out in a cooling chamber at 4 C. The soil
core is transferred to the Plexiglas tube in a frozen
condition. The core is arranged exactly aligned with
the bottom side of the tube and then clamped by
reducing the tube’s diameter with metal clamps. The
flange plates and the tube are then closed gastight
using four threaded bolts.
Five temperature sensors and gas samplers were
inserted into the frozen soil, after adequate holes were
drilled through the openings in the Plexiglas. Additionally, one temperature sensor and one gas sampler
were installed in the headspace above the soil core.
Finally, the microcosm was left in the cooling chamber to thaw the soil core. Afterwards, TDR probes
were plugged into the soft soil substrate.
Two Thermo Haake cryostats (C10-K15
and WKL26; Karlsruhe, Germany) were used to
freeze the permafrost microcosm by flowing a coolant
through the coils in the flanges. Two cooling
circuits ensured different cooling rates and temperatures for the flanges at the top and bottom of the
Measurements of volumetric water content were
carried out using a Campbell Scientific TDR Loughborough, UK) system, which consisted of a CR10 data
logger and a Tektronix 1502 C TDR cable tester. The
TDR probes used for the simulation were two-rod,
75 mm LP/ms laboratory probes from EASY TEST
Ltd. (Lublin, Poland). The distance between the rods
was 4.3 mm and the rod diameter was 0.8 mm.
Copyright # 2003 John Wiley & Sons, Ltd.
Gas samples were taken with Rhizon soil moisture
samplers (Rhizosphere Research Products, Wageningen,
The Netherlands). The sampler consisted of a porous
polymer tube of 50 mm length and an outer diameter of
2.5 mm connected to a PVC tube and a septum, which
was handmade from Luer-lock components. A stainless
steel wire inside the polymer and PVC tubes provided
support. The dead volume of the sampler was 0.5 ml.
Gas samples could be taken with a gastight syringe
through the septum for direct analysis by gas chromatography.
Temperature measurements were made with thin
film platinum RTDs (Honeywell HEL-705-U). These
sensors have cylindrical ceramic cases of 2.2 mm in
diameter and 5 mm in length.
In order to simulate the natural freezing-thawing
cycles of permafrost soils, the microcosm could be
independently frozen downwards and upwards from
the top and bottom side. Furthermore, it was possible
to adjust the freezing process on the top and on the
bottom side of the microcosm to the planned test
sequence (see below) using two separate cooling
circuits (Figure 1).
The simulation experiment begins in unfrozen conditions. Therefore, the permafrost microcosm was
incubated in the cooling chamber for about 48 h at
4 C to calibrate the system. Two different simulation
experiments can be performed: (i) the soil is completely frozen and thawed again or (ii) the soil is partly
frozen downwards and upwards from the top and
bottom side leaving a central zone which remains
unfrozen (Figure 2). Another option is to vary the
duration of the individual experimental phases. For
example, the period of final freezing can be varied, in
order to simulate the seasonal influence on microbial
activity during back-freezing of the active layer. In
general, a simulation cycle consisted of three to five
freezing-thawing phases to ensure statistical significance of the experiment.
During the experiment in the cooling chamber, the
microcosm was kept in a Styrofoam box. Measurements of temperature and water content were automatically logged at time steps of 15 min. Gas samples
were manually taken with a gastight syringe and
analysed with a gas chromatograph. A simulation
experiment with three freezing-thawing cycles lasted
between two and six weeks depending on the soil type
and its water content.
Further applications of the permafrost microcosm
are in progress to investigate the methane production
pathways by either using 14C-labelled substrates (bicarbonate, acetate or plant material) or by variation of
environmental conditions (e.g. the water content or
Permafrost and Periglac. Process., 14: 367–374 (2003)
D. Wagner et al.
Figure 2 Temperature profile of the soil core during a freezing process (soil depth 0 cm ¼ headspace above soil core).
Methane Analysis
The CH4 concentrations in the field experiments and
during the simulation experiments were determined
with a gas chromatograph (Chrompack GC 9003).
The instrument was equipped with a Poraplot Q (100/
120 mesh, 20 ft) capillary column, which operates
with pure helium as carrier gas at a flow rate of
20 ml min1. CH4 was analysed by a flame ionisation
detector. The injector/detector temperatures were set
at 160 C and the column oven at 80 C. All gas sample
analyses were done after calibration with standard
decreased again and attained a relatively constant rate
in August, when the maximum thawing depth of the
active layer was reached. However, the CH4 release in
August 1999 and 2000 revealed erratic fluctuations,
which differ clearly from the emission process at the
beginning of the season (Figures 3 and 4). Statistical
analysis indicated a significant correlation (r ¼ 0.94,
p < 0.0001) between CH4 release and thawing of the
Field Investigations
The CH4 fluxes of the polygon depression measured in
1999 showed, right from the start of the soil thawing
at the end of May, a continuously increasing flux
rate, which reached the highest values in mid-July
(Figure 3). In the course of the season, CH4 fluxes
Copyright # 2003 John Wiley & Sons, Ltd.
Figure 3 CH4 flux and thaw depth of the polygon depression
measured from the beginning of May to the beginning of September
1999 on Samoylov Island, Lena Delta (means SE, n ¼ 5).
Permafrost and Periglac. Process., 14: 367–374 (2003)
Microbial Methane Production 371
Figure 4 CH4 flux and thaw depth of the polygon depression measured in August 2000 on Samoylov Island, Lena Delta (means SE,
n ¼ 5).
active layer for the early vegetation period from the
end of May to the middle of July.
Further field studies of the microbial processes
showed significant activity of the methane-producing
microflora in the boundary of the frozen ground
(Figure 5). This layer was characterized by an in
situ temperature between 0.6 and 1.2 C. The measured soil temperature on our long-term study site in
Figure 5 Microbial CH4 production (three replicates) and in situ
temperature in the boundary to the frozen ground of the polygon
depression. The field experiment was carried out between 5 and 12
July 1999 on Samoylov Island, Lena Delta.
the Lena Delta (Boike et al., 2003) showed that a soil
zone persisted until the middle of October 1998 with
temperatures between 0.5 and 0 C (Figure 6). These
are temperatures in which unadapted microorganisms
are inactive (Morita, 1975). Due to our analyses of
Figure 6 Temperature profile in the active layer of the polygon depression from the beginning of September to the beginning of November
1998 (Lena Delta, Siberia).
Copyright # 2003 John Wiley & Sons, Ltd.
Permafrost and Periglac. Process., 14: 367–374 (2003)
D. Wagner et al.
Figure 7 Water content, temperature and methane concentration during a preliminary test sequence of a freezing-thawing cycle in a
forerunner model (modified after Müller-Lupp, 2002) of the described permafrost microcosm.
microbial activity it can be assumed that methane still
formed up to the complete back-freezing of the active
layer. In contrast to earlier assumptions, methane is
produced not only during the vegetation period but
also in the cold transitional season. This methane is
kept enclosed in the active layer during the backfreezing process and is released in the thawing phase
Copyright # 2003 John Wiley & Sons, Ltd.
of the following spring. Samarkin et al. (1999) already
assumed that such back-stored methane adds to the
emission of newly-produced methane. Both the currently-formed methane and the back-stored methane
from the year before determine the characteristic
increase of trace-gas release observed at the beginning
of the season 1999 (Figure 3).
Permafrost and Periglac. Process., 14: 367–374 (2003)
Microbial Methane Production 373
Permafrost soils are characterized by extreme variation in temperature (Yershov, 1998). Many microorganisms are able to survive in cold permafrost
sediments, but this adaptation can be either a tolerance
or a preference. In accordance with Panikov (1997),
psychrophilic bacteria are a significant part of the
microbial community in cold permafrost soils. Our
results indicate the existence of a methane-producing
microflora which is adapted to low in situ temperatures of permafrost-affected soils.
The seasonal variation of soil temperature also
influences the availability of pore water. The presence
of unfrozen water is an essential bio-physical requirement for the activity of microorganisms in permafrost.
Temperatures below zero stand for an increasing loss
of water. At the same time, freezing of water leads to
an increase in salt content in the remaining pore
solution. Nevertheless, permafrost soils are known to
contain significant amounts of unfrozen water (Boike
et al., 1998). The most important feature of this water
is the possible transfer of ions and nutrient (Ostroumov
and Siegert, 1996). Therefore, the microflora in permafrost soils and sediments is influenced by changing
gradients. These gradients, especially those of temperature and water content, are considered in the
simulation experiment.
To understand the carbon cycle in permafrost environments in the off-peak season further research is
needed. Field studies are difficult to accomplish during the back-freezing of the active layer and necessary
repetitions to validate results are often impossible in
the same season.
microcosm presented in this paper ensures independent freezing and thawing the top and the bottom of the
microcosm using two cooling circuits. This construction allows the simulation of different field conditions.
Furthermore, CH4 production rates can be determined
for different soil depth using gas samplers at each 5 cm
of the microcosm.
Simulation experiments can give insights into the
natural ecosystem and may yield important clues for
the understanding of microbial life under extreme
permafrost conditions. The use of undisturbed soil
cores keep the original structure, pore system and
stratification of the natural soil. The temperatures and
water contents measured during the simulation can be
adjusted with field data obtained from long-term
measurements in the Lena Delta, Siberia (Boike
et al., 2003). Thus, a direct correlation between the
field conditions and the involved processes can be
derived. Under this premise, the interaction between
microorganisms and soil matrix can be assessed in a
reasonable manner. The permafrost microcosm simulates the two freezing fronts (one from the top, one
from the bottom) while the centre of the active layer
remains unfrozen and methane production can still
continue. Such freezing-thawing experiments can
help to answer questions as to how the microbial
population will be influenced by the natural permafrost system and how microorganisms interact with
their environmental conditions.
Simulation Experiment
We acknowledge the good collaboration among the
members of the Russian-German field parties during
the expeditions LENA 1999 and LENA 2000. Special
thanks go to Waldemar Schneider for logistic support
and Günter ‘Molo’ Stoof for technical assistance.
In addition, we thank Dr B. Diekmann (Alfred
Wegener Institute for Polar and Marine Research)
for critical reading of the manuscript as well as
Professor Anatoli Brouchkov (Hokkaido University)
and an anonymous reviewer for their helpful comments on the manuscript.
A preliminary test sequence in a forerunner model
(modified after Müller-Lupp, 2002) of the described
permafrost microcosm showed the relationship between soil temperature, water content and CH4 concentration during the freezing-thawing process (Figure 7).
In the first 60 h the permafrost microcosm was incubated at 5 C to calibrate the system. Then the freezing
cycle was started and the temperature was lowered to
temperatures between 0 C and 4 C as a function of
the core depth. At this time the water content decreased from 80% to about 10% and the CH4 concentration in the pore space decreased as well. After 120 h
the thawing cycle was initiated again and immediately
after thawing of the soil the highest CH4 concentration
(63 ppm) was analysed. These first data support the
field observations that indicate methane production
occurs in the unfrozen soil zone in autumn, which
then leads to an inclusion of additional methane into
the active layer. The new configuration of the
Copyright # 2003 John Wiley & Sons, Ltd.
Boike J, Roth K, Overduin PP. 1998. Hydrological and
thermal dynamics of the active layer (Taymyr
Peninsula, Siberia). Water Resources Research 34:
Boike J, Hinzman L, Overduin PP, Romanovsky V,
Ippisch O, Roth K. 2003. A comparison of snow
Permafrost and Periglac. Process., 14: 367–374 (2003)
D. Wagner et al.
melt at three circumpolar sites: Spitzbergen, Siberia,
Alaska. In Proceedings of the 8th International Conference on Permafrost, Zurich, Switzerland, 21–25
July 2003. Balkema Publishers: Lisse; 79–84.
Christensen TR, Prentice IC, Kaplan J, Haxeltine A,
Sitch S. 1996. Methane flux from northern wetlands
and tundra: an ecosystem source modelling approach.
Tellus 48B: 651–660.
Dise NB. 1992. Winter fluxes of methane from Minnesota
peatlands. Biogeochemistry 17: 71–83.
Friborg T, Christensen TR, Søgaard H. 1997. Rapid
response of greenhouse gas emission to early spring
thaw in a subarctic mire as shown by micrometeorological techniques. Geophysical Research Letters 24:
Garcia J-L, Patel BKC, Ollivier B. 2000. Taxonomic,
phylogenetic, and ecological diversity of methanogenic Archaea. Anaerobe 6: 205–226.
Hanson RS, Hanson TE. 1996. Methanotrophic bacteria.
Microbiological Reviews 60: 439–471.
Melloh RA, Crill PM. 1996. Winter methane dynamics in
a temperate peatland. Global Biogeochemical Cycles
10: 247–254.
Morita RY. 1975. Psychrophilic bacteria. Bacteriological
Review 39: 144–167.
Müller-Lupp W. 2002. Freeze and thaw in Siberian
permafrost soils—methods and ecological relevance.
Reports on Polar and Marine Research, AWI: Bremehaven; 415: 145.
Ostroumov VE, Siegert C. 1996. Exobiological aspects
of mass transfer in microzones of permafrost deposits.
Advances in Space Research 18: 79–86.
Panikov NS. 1997. A kinetic approach to microbial
ecology in arctic and boreal ecosystems in relation
to global change. In Global Change and Arctic Terrestrial Ecosystems, Oechel WC (ed.). Springer: Berlin;
Panikov NS, Dedysh SN. 2000. Cold season CH4 and
CO2 emission from boreal peat bogs (West Siberia):
winter fluxes and thaw activation dynamics. Global
Biogeochemical Cycles 14: 1071–1080.
Copyright # 2003 John Wiley & Sons, Ltd.
Pfeiffer E-M, Wagner D, Becker H, Vlasenko A, Kutzbach
L, Boike J, Quass W, Kloss W, Schulz B, Kurchatova A,
Pozdnyakov VI, Akhmadeeva I. 2000. Modern processes in permafrost affected soils. In Russian-German
Cooperation System Laptev Sea 2000: The Expedition
LENA 1999, Rachold V, Grigoriev MN (eds). Reports
on Polar and Marine Research, AWI: Bremehaven; 354;
Samarkin VA, Gundelwein A, Pfeiffer E-M. 1999. Studies of methane production and emission in relation to
the microrelief of a polygonal tundra in Northern
Siberia. In Land-Ocean Systems in the Siberian Arctic,
Kassens H, Bauch HA, Dmitrenko IA, Eicken H,
Hubberten HW, Melles M, Thiede J, Timokhov LA
(eds). Springer: Berlin; 329–342.
Soil Survey Staff. 1998. Keys to soil taxonomy. Eighth
Edition. Soil Conservation Service, USDA: Washington,
Wagner D, Kurchatova A, Stoof G. 2001a. Modern
processes in permafrost affected soils. In RussianGerman Cooperation System Laptev Sea 2000: The
Expedition LENA 2000, Rachold V, Grigoriev MN
(eds). Reports on Polar and Marine Research, AWI:
Bremehaven; 388: 16–23.
Wagner D, Spieck E, Bock E, Pfeiffer E-M. 2001b.
Microbial life in terrestrial permafrost: methanogenesis and nitrification in Gelisols as potentials for
exobiological processes. In Astrobiology: The Quest
for the Conditions of Life, Horneck G, BaumstarkKhan C (eds). Springer: Berlin; 143–159.
Wagner D, Kobabe S, Pfeiffer E-M, Hubberten H-W.
2003. Microbial controls on methane fluxes from a
polygonal tundra of the Lena Delta, Northeast Siberia.
Permafrost and Periglacial Processes 14: 173–185.
Whalen SC, Reeburgh WS. 1988. A methane flux time
series for tundra environments. Global Biogeochemical Cycles 2: 399–409.
Wuebbles DJ, Hayhoe K. 2002. Atmospheric methane and
global change. Earth-Science Reviews 57: 177–210.
Yershov ED. 1998. General Geocryology. Cambridge
University Press: Cambridge.
Permafrost and Periglac. Process., 14: 367–374 (2003)
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF