Department of Thematic Studies Campus Norrköping
Bachelor of Science Thesis, Environmental Science Programme, 2020
Eldina Cehic
An annual evaluation of CH
4
Rapporttyp Report category Licentiatavhandling Examensarbete AB-uppsats C-uppsats D-uppsats Övrig rapport Språk Language Svenska/Swedish Engelska/English Titel
En årlig utvärdering av CH4 oxidation i en sötvattensjö Title
An annual evaluation of CH4 oxidation in a freshwater lake Författare
Author Eldina Cehic Sammanfattning
Utsläppen av metan (CH4) från sötvattensjöar begränsas genom CH4 oxidation. Det är två stabila kolisotoper som dominerar i CH4; 12C och 13C. Den ostabila kolisotopen 14C finns även i CH4, men den är mer ovanlig i naturen. Metantrofer (metanoxiderande bakterier) oxiderar den lättare kolisotopen snabbare. Förändringar i isotopsammansättningen kan användas för att beräkna hur mycket CH4 som oxideras i ett system. Denna studie undersöker en årlig CH4 oxidation i en sötvattensjö. Vattenprover och bubblor av CH4-gas samlades en gång i månaden, från mars till november, i Gundlebosjön. CH4 gasen separerades från vattenproverna med en ”headspace extraction” teknik. Koncentration och
isotopsammansättningen av CH4 analyserades i en ”cavity ring down spectrometer”. Isotopdata användes i två matematiska modeller, baserade på öppet-stabilt tillstånd och stängt system. Den stabila isotopmetoden för att uppskatta CH4 oxidation var endast användbar under perioder då tydliga skillnader i koncentrationen och isotopsammansättningen kunde observeras i vattenpelaren. CH4 oxidation kunde endast uppskattas i vattenpelaren i augusti, och i vattenpelarens ytskikt i juni och juli.
Abstract
Freshwater lakes constrain its methane (CH4) emissions through CH4 oxidation. CH4 includes three carbon (C) isotopes; the stable isotopes 12C,13C and the unstable and more uncommon isotope 14C. Methanotrophs (i.e. methane oxidizing bacteria) oxidize the lighter isotope more rapidly. Changes in relative isotopic composition can therefore be used to calculate how much CH4 is oxidized in a system. This study investigates an annual CH4 oxidation in a freshwater lake. Water samples and bubbles of CH4 gas were collected once a month, from March to November, in lake Gundlebosjön. The CH4 gas was separated from the water samples with a headspace extraction technique. The concentration and isotopic composition of CH4 was analyzed in a cavity ring down spectrometer. The isotopic data was used in two mathematical models, based on open-steady state and closed systems. It was found that the stable isotope method to estimate CH4 oxidation was only useful during periods when clear concentration and isotope differences could be observed in the water column. CH4 oxidation could only be estimated in the water column in August, and in the surface layer in June and July.
ISBN _____________________________________________________ ISRN LIU-TEMA/MV-C—20/09--SE _________________________________________________________________ ISSN _________________________________________________________________ Serietitel och serienummer
Title of series, numbering
Handledare Tutor David Bastviken Datum Date 2020-06-01
URL för elektronisk version
http://www.ep.liu.se/index.sv.html
Nyckelord
Metanoxidation, isotopfraktionering, stratifiering, öppet stabilt system och stängt system Keywords
Institution, Avdelning
Department, Division Tema Miljöförändring, Miljövetarprogrammet
Department of Thematic Studies – Environmental change Environmental Science Programme
Acknowledgements
First and foremost, I want to thank my tutor David Bastviken for his support, guidance and constructive feedback. I also want to thank Henrique Sawakuchi for his help and guidance with the sampling simulation, laboratory- and data analysis.
2020-05-16 Eldina Cehic
Abstract
Freshwater lakes constrain its methane (CH4) emissions through CH4 oxidation. CH4 includes
three carbon (C) isotopes; the stable isotopes 12C,13Cand the unstable and more uncommon isotope 14C.Methanotrophs (i.e. methane oxidizing bacteria) oxidize the lighter isotope more rapidly. Changes in relative isotopic composition can therefore be used to calculate how much CH4 is oxidized in a system. This study investigates an annual CH4 oxidation in a freshwater
lake. Water samples and bubbles of CH4 gas were collected once a month, from March to
November, in lake Gundlebosjön. The CH4 gas was separated from the water samples with a
headspace extraction technique. The concentration and isotopic composition of CH4 was
analyzed in a cavity ring down spectrometer. The isotopic data was used in two mathematical models, based on open-steady state and closed systems. It was found that the stable isotope method to estimate CH4 oxidation was only useful during periods when clear concentration
and isotope differences could be observed in the water column. CH4 oxidation could only be
estimated in the water column in August, and in the surface layer in June and July.
Keywords: Methane oxidation, isotopic fractionation, stratification, open steady-state and closed system.
Sammanfattning
Utsläppen av metan (CH4) från sötvattensjöar begränsas genom CH4 oxidation. Det är två
stabila kolisotoper som dominerar i CH4; 12C och 13C. Den ostabila kolisotopen 14C finns
även i CH4, men den är mer ovanlig i naturen. Metantrofer (metanoxiderande bakterier)
oxiderar den lättare kolisotopen snabbare. Förändringar i isotopsammansättningen kan
användas för att beräkna hur mycket CH4 som oxideras i ett system. Denna studie undersöker
en årlig CH4 oxidation i en sötvattensjö. Vattenprover och bubblor av CH4-gas samlades en
gång i månaden, från mars till november, i Gundlebosjön. CH4 gasen separerades från
vattenproverna med en ”headspace extraction” teknik. Koncentration och
isotopsammansättningen av CH4 analyserades i en ”cavity ring down spectrometer”.
Isotopdata användes i två matematiska modeller, baserade på öppet-stabilt tillstånd och stängt system. Den stabila isotopmetoden för att uppskatta CH4 oxidation var endast användbar
under perioder då tydliga skillnader i koncentrationen och isotopsammansättningen kunde observeras i vattenpelaren. CH4 oxidation kunde endast uppskattas i vattenpelaren i augusti,
och i vattenpelarens ytskikt i juni och juli.
TABLE OF CONTENTS 1. INTRODUCTION ... 1 2. AIM ... 3 2.1 RESEARCH QUESTION ... 3 3. BACKGROUND ... 4 3.1THERMAL STRATIFICATION ... 4
3.1.1 Mixing & stratification ... 4
... 5
3.2CH4 PRODUCTION ... 5
3.3CH4 OXIDATION ... 6
3.3.1 Spatial distribution of CH4 oxidation ... 6
3.4STABLE ISOTOPES OF CARBON ... 7
3.4.1MODELS BASED ON ISOTOPIC FRACTIONATION ... 8
4. METHOD ... 10
4.1RESEARCH PROJECT “METLAKE” ... 10
4.2STUDY AREA ... 10
4.3SITE SELECTION &SAMPLING PERIOD ... 11
4.4SAMPLING WATER IN THE DEPTH PROFILE ... 11
4.4.1 Headspace method ... 12
4.4.2 Headspace calculations ... 13
4.5MEASURING O2&TEMPERATURE ... 14
4.6SAMPLING CH4 BUBBLES ... 14 4.7SAMPLE QUANTITY ... 15 4.8LABORATORY ANALYSIS... 15 4.8.1 Preparation for CRDS ... 16 4.8.2 Analysis in CRDS ... 16 4.9MODEL APPLICATION ... 17 5. RESULTS ... 19
5.1 Δ13C-CH4,CH4,O2&TEMPERATURE PROFILES ... 19
5.2CH4 BUBBLES ... 22
5.3FRACTION OF OXIDIZED CH4 ... 23
6. DISCUSSION... 25
6.1RELIABILITY &VALIDITY ... 25
6.2EVALUATING THE MODEL APPLICATION ... 25
6.3 Δ13C-CH4,CH4,O2&TEMPERATURE PROFILES ... 26
6.4WATER COLUMN CH4 OXIDATION ... 28
7. CONCLUSION ... 30
8. REFERENCES ... 31
1.
Introduction
Greenhouse gases (GHG) have the ability of absorbing thermal infrared radiation (heat energy) that is emitted from the earth’s surface. GHG emit some of the heat energy back to earth’s surface, which warms the planet. Methane (CH4) is of one of the most abundant GHG
in the atmosphere (Montzka et al. 2011). It has an estimated residence time of 8-12 years in the atmosphere (Dlugokencky et al. 2009), whilst carbon dioxide (CO2) can remain for
approximately 120 years (Essenhigh 2009). In a time perspective of 100 years, one mass unit of CH4 has a global warming potential that is 28 times greater than CO2 (Myhre et al. (2013).
Sources of anthropogenic CH4 emissions can range from agricultural activities to fossil fuel
production. The largest natural CH4 emissions stemfrom wetlands and the second largest
come from freshwater lakes (Saunois et al. 2016).
The global CH4 budget is between 500-600 Tg yr-1 (Saunois et al. 2019), and it accounts for
the aquatic and terrestrial CH4 sources. It’s also balanced by the atmospheric sink, where CH4
is oxidized by hydroxyl radicals (OH) (Kirschke et al. 2013). The global CH4 emissions from
freshwater lakes is approximately 78-139 Tg yr-1 (DelSontro et al. 2017) and is constrained
through aerobic (with O2)- or anaerobic (without O2) microbial CH4 oxidation (Bastviken et
al. 2002). In a study by Bastviken et al. (2008), it was determined that 51-80% of the CH4 was
oxidized before reaching the surface water. This process is an important CH4 sink, since
reduced CH4 oxidation could result in CH4 emissions increasing from freshwater lakes (Shelly
et al. 2015).
In a study by Shelly et al. (2015), it was determined that the rate of CH4 production increased
with temperature, and that CH4 oxidation was eight times faster than production when the
water temperature reached 10 °C. Shelly et al. (2015) concluded that CH4 oxidation has the
potential to match or even exceed CH4 production at increasing temperatures. Fuchs et al.
(2016) reached similar conclusions, when investigating the balance between CH4 production
and oxidation. Furthermore, CH4 oxidation is regulated by the availability of CH4 and O2
(Lofton et al. 2013). A water column can undergo physical changes under a year, e.g. lake
mixing and stratification where the lake separates into three layers with different densities (see section 3.1). Such changes can affect the availability and distribution of CH4 and O2 in
the water column. Hence, if CH4 and O2 aren’t simultaneously present in sufficient
concentrations, then less CH4 could be oxidized, unless anaerobic CH4 oxidation takes over
(Bastviken 2009). Moreover, there is limited knowledge on how CH4 oxidation varies over a
year, since most of the existing studies provide measurements representing a shorter time period. An annual analysis would show when and where CH4 oxidation is most extensive in
the water column, and thereby indicate what months the CH4 emissions are more or less
constrained. Determining where CH4 is mostly oxidized in the water column and if this
2.
Aim
The aim of this thesis is to investigate an annual CH4 oxidation in a freshwater lake.
2.1 Research question
▪ Where does CH4 oxidation occur in a freshwater lake during different parts of the
3.
Background
3.1 Thermal stratificationThis section describes how lake stratification (formation of water layers) is governed by changes in temperature and water density. Understanding how lakes become stratified can be a basis for exploring how the circumstances for CH4 production and oxidation changes
(Pöschke et al. 2015). Moreover, temperature and water density have an inverse relationship. As water temperature increases, the water density will decrease. When the water temperature decreases to 4 °C, the water will reach its maximum density at 1 g/cm3 (Pöschke et al. 2015).
Differences in temperature and density can cause stratification, where layers are formed in the water column. Stratified layers are referred to as epilimnion (surface mixed layer),
metalimnion (middle layer where the temperature change rapidly with depth) and hypolimnion (bottom layer) (Woolway et al. 2014)
3.1.1 Mixing & stratification
A lake can be mixed, with the help of wind-induced currents, when the temperature and density is similar in the water column. During mixing, the surface water will move to the bottom and the bottom water will travel to the surface (Fig. 1) (Pöschke et al. 2015). The water temperature starts to increase when the lake is exposed to more thermal radiation from the sun. The amount of absorbed thermal radiation will decrease with depth (Fichot et al. 2019). As mentioned above, warm water is less dense than cold. Hence, a layer of warm water will develop on top of the colder water, creating the surface epilimnion and bottom
hypolimnion. The epilimnion and hypolimnion will be separated by a middle layer, known as the metalimnion (Fig. 1). The metalimnion has a distinct density and temperature gradient, where the temperature changes a lot along the depth scale (Woolway et al. 2014).
Consequently, the transport of compounds through the metalimnion will be slower, and the exchange between the hypolimnion and epilimnion will be limited (Thalasso et al. 2020).
When the epilimnion starts to cool in the fall, its water density will increase. The density difference between the epilimnion and hypolimnion will be reduced. The cooled epilimnetic water will sink towards the hypolimnion, and the layers will be mixed. The epilimnion will increase in depth over time and the temperature will eventually be similar throughout the water column (Holland & Kay 2003). When the surface water cools down to 0 °C, ice will form and cover the lake’s surface (Fig. 1). Stratification can still happen since water reaches
its highest density at 4 °C, which means that water with temperatures above 0 °C will remain below the frozen surface layer (Pöschke et al. 2015).
3.2 CH4 production
Methanogenesis is a metabolic process performed by a type of archaebacteria known as methanogens. This process occurs in anaerobic environments, such as sediments where electron acceptors, e.g. O2, SO42- & NO3- , are present in very low concentrations (Bastviken
2009). Two pathways of methanogenesis are; acetoclastic and hydrogenotrophic. In
acetoclastic methanogenesis, acetic acid (CH3COOH) is used as a substrate. CH3COOH can
be produced by e.g. fermentation and homoacetogenesis. With the help of coenzymes, acetoclastic methanogens can break down the CH3COOH to CH4 and CO2. (Wilson et al.
2019). In hydrogenotrophic methanogenesis, CO2 is reduced with H2 as an electron donor, and
the products are CH4 and H2O (Bastviken 2009). Both CO2 and H2 are used in
homoacetogenesis and hydrogenotrophic methanogenesis, which can lead to both processes competing for the same substrates (Ye et al. 2013). The reaction formulas (Eq 1-4) were retrieved from an article by Wilson et al. (2019)
Eq 1: Fermentation Eq 2: Homoacetogenesis 2C6H12O6 + 4H2O → 4CH3COOH + 4CO2 + 8H2 2CO2 + 4H2 → CH3COOH + 2H2O
Eq 3: Acetoclastic methanogenesis Eq 4: Hydrogenotrophic methanogenesis 5CH3COOH → 5CO2 + 5CH4 CO2 +4H2 →CH4 +2H2O
Figure 1: A simplified illustration of water column mixing and stratification, see section 3.1 for more
3.3 CH4 oxidation
CH4 oxidation is performed by a type of bacteria known as methanotrophs. This process
occurs in aerobic environments where methanotrophs use dissolved O2 as an electron acceptor
(Whalen 2005). CH4 can also be oxidized in anaerobic environments where electron
acceptors, e.g. SO42-, NO3- & Fe3+, are used instead (Thottathil et al. 2018). In comparison
with aerobic CH4 oxidation, a smaller fraction of CH4 is known to be oxidized anaerobically
in freshwater lakes (Schubert et al. 2011). CH4 oxidation has extensive reaction pathways,
however Whalen (2005) presented an abbreviated version by only including the intermediate products methanol (CH3OH), formaldehyde (HCHO) and formate (HCOOH) (Eq. 5). The first
step in oxidation where CH4 is converted to CH3OH, is catalyzed by the enzyme methane
monooxygenase (MMO). Methanotrophs can utilize the energy from oxidation and the intermediate formaldehyde (HCHO) as a carbon source for their growth.
Eq 5: CH4 oxidation
CH4 → CH3OH → HCHO → HCOOH → CO2
3.3.1 Spatial distribution of CH4 oxidation
CH4 oxidation is known to take place in the interface between the oxic and anoxic region,
which can be somewhere in the water column, surface or deeper sediments (Fig. 2) (He et al. 2012). The oxygenation of the water column can vary over time which can impact the location of the oxic-anoxic interface (Bastviken 2009). In a study by Bastviken et al. (2008), CH4 emissions and oxidation were investigated in 3 lakes. The findings showed that CH4 was
mostly oxidized in the area below the metalimnion, which was determined to be the oxic-anoxic interface. The interface or “CH4 oxidation zone” had simultaneous availability of CH4
and O2. The results showed that approximately 80% of the stock CH4 in the CH4 oxidation
zone was oxidized daily. The metalimnion had lower CH4 concentrations and consequently
lower oxidation rates. Bastviken et al. (2008) determined that the oxidation rate was non-significant in the deeper hypolimnion, due to low O2 levels. Anaerobic CH4 oxidation was not
observed in the studied lakes.
Thottathil et al. (2018) analyzed CH4 oxidation in 14 lakes and reached similar conclusions as
Bastviken et al. (2008). In the majority of the studied lakes, up 90% of CH4 was oxidizedin
the oxic-anoxic interface, located below the metalimnion. Thottathil et al. (2018) also found that CH4 oxidation was extensive in the epilimnion. The surface sediments in the epilimnion
can be a source of CH4 (Fig. 2). The epilimnion is also known to be rich in O2 and
methanotrophs tend to be present in the surface sediments (Bastviken 2009). However, ebullition can reduce the oxidation efficiency in the epilimnion. Ebullition is a process where pockets of CH4 gas are released from the sediments as bubbles (Fig. 2). This process can be
initiated by e.g. CH4 saturation in sediments or low hydrostatic pressure. Ebullition can be
intensive in the surface sediments, due to the low hydrostatic pressure found in the surface layer (De Mello et al. 2018). This process has an episodic nature which makes it difficult to analyze. The rapid release of CH4 can leave little to no time for methanotrophs to perform
oxidation (Bastviken et al. 2008).
Figure 2: A simplified illustration on some of the transport pathways of CH4 and the possible locations for
the oxic-anoxic interface, where CH4 oxidation is found to be extensive, in a stratified water column. See
section 3.3 for more information.
3.4 Stable isotopes of carbon
Approximately 99% of the carbon on earth consists of the stable isotope carbon-12 (12C) and circa 1.11% consists of the stable isotope carbon-13 (13C). Carbon-14 (14C) is a rarer and unstable isotope of carbon (Conrad & Claus 2009). The aforementioned carbon isotopes can be found in CH4, with 14CH4 being the most uncommon (Leonte et al. 2018). In biochemical
reactions such as CH4 oxidation, methanotrophs tend to discriminate against the heavier
isotope 13CH4. This is caused by the kinetic isotope effect, where molecules with a lighter
mass (12CH4) reacts at a faster rate than the heavier molecules with a larger mass (13CH4)
(Blaser & Conrad 2016). The remaining CH4 (after oxidation) will have a higher abundance
of 13C, compared to before oxidation. However, 12C will still be the dominant isotope in the remaining CH4, since it is the most abundant isotope of carbon (Whiticar 1999). This is
mathematical models that are based on the isotopic fractionation of 12C/13C and can be used to calculate how much CH4 is oxidized.
3.4.1 Models based on isotopic fractionation
Isotopic data is presented in a delta notation (δ), in order to show the isotopic change (Srivastava & Verkouteren 2018), and they can be referred to as signatures (Bastviken et al. 2008). Isotopic data is expressed in permil (‰), since isotopic ratios have multiple significant figures. The 13C/12C ratio is relative to the Pee Dee belemnite standard (PDB) (Eq. 6). The PDB standard is a type of limestone. If a sample has the same isotope ratio as the PDB, it will have a δ value of 0. A sample that has more 13C per 12C will have a more positive δ value, and a sample with less 13C per 12C will have more negative δ value (Srivastava &
Verkouteren 2018). CH4 produced from hydrogenotrophic methanogenesis will be depleted in
13C and will have a δ13C-CH4 signature between -110 to -60‰, whilst the acetoclastic
methanogenesis will produce 13C enriched CH4 with a signature from -60 to -50‰ (Fischer et
al. 2017). Equation 6 was retrieved from an article by Whiticar (1999).
Eq 6: δx (‰) = [Rsample/Rstandard - 1] * 103
▪ δx (‰) = δ13C in sample relative to the PDB standard
▪ Rsample = 13C/12C ratio in sample
▪ Rstandard = 13C/12C ratio in standard.
Happel et al. (1994) presented a model that is based on open-steady state conditions (Eq. 7). A lake can be at open-steady state when its e.g. mixed and no ice is covering the lake, and thereby open for gas exchange between the surface water and atmosphere (DelSontro et al. 2017).
Eq 7: fopen = (δs - δb)/((a - 1) × 1000)
fopen = Fraction of oxidized CH4
δs = δ13C-CH4 in the surface water (CH4 that has been exposed to oxidation)
δb = δ13C-CH4 in the bottom (CH4 that hasn’t been exposed to oxidation)
a = Isotopic fractionation factor 1000 = Conversion to permil (‰)
The isotopic fractionation factor represents the change that will happen in the abundance ratio of two isotopes during a chemical or physical process. There is no “standard” fractionation factor that is used, instead it’s determined experimentally (Visscher et al. 2004). Bastviken et al. (2002) determined the fractionation factors at 5 °C and 20 °C, by incubating water samples and monitoring the CH4 concentration and δ13C-CH4 signatures over time. The average values
for the fractionation factor were between 1.0184 and 1.0208, and Bastviken et al. (2002) used a 1.02 factor in the CH4 oxidation (MOX) calculations. Templeton et al. (2006) incubated
water samples and cultures of methanotrophs at 22-24 °C and determined a factor range of 1.002-1.03.
Liptay et al. (1998) presented a Rayleigh model for closed systems (Eq 8). The hypolimnion can be considered a closed system during the stratification period, due to the limited
compound transport (see section 3.1.1). An entire lake can be a closed system when e.g. the surface is fully covered by ice (DelSontro et al. 2017). Moreover, this model is derived from a Rayleigh distillation function (Eq 9), that assumes that a compound is continuously removed from a system by fractional distillation (Liptay et al. 1998). See appendix for how equation 8 is derived from equation 9.
Eq 8 ln(1 - fclosed) = [ln(δb + 1000) - ln(δs + 1000)]/[a - 1]
fclosed = Fraction of oxidized CH4
δs = δ13C-CH4 in the surface water (CH4 that has been exposed to oxidation)
δb = δ13C-CH4 in the bottom (CH4 that hasn’t been exposed to oxidation)
a = Isotopic fractionation factor 1000 = Conversion to permil (‰) Eq 9: Rr = Rr.i f (a-1)
▪ Rr = isotopic ratio of the compound after a period of distillation
▪ Rr.i = the initial isotopic ratio
▪ a = Isotopic fractionation factor ▪ f = Fraction of remaining substrate
4.
Method
4.1 Research project “METLAKE”
“METLAKE” is an ongoing research project at Linköping’s university. The purpose of the project is to increase the knowledge regarding CH4 emissions from lakes. The METLAKE
team have collected samples from multiple freshwater lakes in the provinces Västra Götaland, Östergötland, Småland and Norrbotten. The sampled lakes were selected to represent different lake types in different climate zones. For this thesis, samples were collected as a part of METLAKE. However, a sampling simulation was run to transfer practical experience and detailed understanding of the procedures and to allow me to describe the methods
independently. The laboratory preparations and analyses were performed by me independently.
4.2 Study area
The lake Gundlebosjön was sampled due to its eutrophic status and accessibility.
Gundlebosjön was selected for this thesis, because the samples from this lake stood next in line for analysis. Gundlebosjön is located within the municipality of Vänersborg, in the province of Västra Götaland. The lake is located 51.6 m above sea level. It’s 9 m deep and has surface area of 0,47 km2. Gundlebosjön is a part of the main catchment area called
Bäveån. The lake is drained in the nearby watercourse called Gundleboån (Fig 3). The lake is surrounded by forest (53%), agricultural land (22%) and open terrain (15%)
(Samhällsbyggnadsnämnden 2016).
4.3 Site selection & Sampling period
The deepest point of the lake was localized using a bathymetric map and selected for
sampling along a depth profile. The purpose with the depth profile was to determine how CH4
oxidation varies from the bottom to the surface of the water column. Gundlebosjön is 9 m deep, and to avoid sediment contamination the deepest sample was collected at 8 m. Samples were collected from a boat once a month, from March to November.
4.4 Sampling water in the depth profile
The ruttner sampler was by itself 1 m long, and the continuing lengths were marked on the attached rope (Fig. 4). The was rope was used to measure the depth where a sample was collected and to create the depth profile (0.3 m, 1.5 m, 3 m, 4.5 m, 6 m, 7 m, 8 m). The sampler was lowered to the desired depth and closed by dropping the messenger weight along the line, triggering the closing mechanism. After bringing the sampler up to the boat, the sample water was transferred to a 1.21-L polycarbonate bottle (Fig 5). During this process, the bottle was overfilled three times with the sample water in order to remove potential bubbles. The bottle was then closed with a rubber stopper that had two tubes attached to it, one longer that reached the bottom of the bottle and one shorter (Fig. 5). This sampling procedure was performed for all water samples collected in depth profile.
Figure 5: 1.21-L polycarbonate bottle
4.4.1 Headspace method
The purpose with the headspace technique is to separate volatile compounds from samples that are either in liquid or solid phase. One of the benefits with this technique is that it allows for fast separation and extraction of compounds (Kremser et al. 2016), which is why it was used to separate the CH4 from the collected water samples. The approach taken in this study
was applied in a very similar manner by Sawakuchi et al. (2016), where CH4 oxidation and
emissions were investigated in Amazonian rivers.
Procedure
Zero-air is synthetic air that has gone through hydrocarbon removal and is mainly composed of O2 and N2. Two empty 60 ml syringes were flushed with zero-air in order to remove
potential contaminants. One 60 ml syringe was then filled with zero-air and attached to the shorter tube (Fig. 5). The second 60 ml syringe was left empty and attached to the longer tube (Fig. 5). The zero-air was injected into the bottle whilst water was simultaneously withdrawn with the empty 60 ml syringe. This process created a headspace in the bottle. Zero-air was chosen to minimize the risk of contaminating the sample, in contrast to atmospheric air that includes some methane and other hydrocarbon compounds. The next step was for the sample and the headspace to reach equilibrium. This can be time consuming if the equilibrium should be reached by itself. Therefore, this process was accelerated by shaking the bottle vigorously for 2 min. Then 60 ml gas (sample) was extracted from the headspace whilst the previous extracted water was injected back. This volume was extracted in order to have enough sample for analysis. The temperature was measured in the bottle after extraction.
Before the gas sample was transferred to a 20 ml vial closed with a 3 mm rubber stopper, the vial was manually evacuated in the field. The first step in the vial-evacuation procedure was to attach a needle and a three-way valve to a new empty syringe (60 ml). The second step was to pierce the rubber stopper and use the syringe piston to repeatedly draw air out the vial, and release through the three-way valve. This was repeated until resistance was felt in the syringe. The gas sample was transferred to the evacuated vial by attaching a needle to the sample syringe and using it to pierce the stopper and inject the sample. An important note is that vials can be evacuated in the laboratory. However, there’s a risk of air leaking back into the vials over time, which is why they were manually evacuated in the field immediately before sample addition. Moreover, the described headspace and evacuation procedure was repeated for all water samples and vials. The sample vials were stored in polyethylene bags until analysis.
4.4.2 Headspace calculations
To determine the CH4 concentration in the water before the headspace extraction, a series of
calculations were performed. In a gas mixture, the partial pressure refers to the pressure exerted by one gas in the mixture. According to Dalton’s Law of partial pressures, the total pressure exerted by a gas mixture is the sum of the partial pressure of each gas (Hilgeman et al. 2007). Therefore, the first step was to convert the total air pressure (Ptot) from hectopascal
(hPa) to atmospheres (atm), in order to obtain pressure at standard conditions. This was done by multiplying the total air in hectopascal (1013hPa) with a conversion factor (100 *
0,000009869).
Eq 10: Ptot (atm) = P (hPa) * (100 * 0,000009869)
The conversion factor is based on: ▪ 1 hPa = 100 Pascal
▪ 1 Pascal = 0,000009869 atm
The second step was to calculate the partial pressure of CH4. This was done by dividing the
concentration (in ppm) with 106, in order to cancel out the ppm, and multiplying it with Ptot
(atm).
Eq 11: PCH4 = Cppm/106 * Ptot (atm).
The moles (amount of substance) of CH4 gas in the headspace was calculated by using the
(Vg). This was divided with the common gas constant (0.082056 L atm K-1 mol-1) multiplied
with the headspace temperature in Kelvin.
Eq 12: ngas = (PCH4 * Vg)/(R * T)
The moles (amount of substance) of CH4 gas in the water was calculated by multiplying the
partial pressure of CH4 (PCH4), Henry’s law constant (KH) and the water volume (Vaq).
Henry’s law constant for CH4 (1.4 * 10-3) was selected from Warneck & Williams (2012)
compilation of solubility coefficients for multiple gases.
Eq 13: naq = PCH4 * KH * Vaq
The total amount of CH4 in moles was calculated by adding the amount of CH4 in the
headspace with the amount of CH4 in the water. The original concentration in the water was
calculated by dividing the total amount of CH4 with the water volume.
Eq 14: ntot = ngas + naq Eq 15: Caq = ntot/Vaq
4.5 Measuring O2 & Temperature
An optical sensor probe (Hatch Intellical LDO101-probe), equipped with a temperature meter, was used to measure dissolved oxygen and temperature in the water column. For O2 the
sensor has a detection range of 0.05-20 mg/l. The optical sensor measures how dissolved O2
interacts with luminescent dyes when exposed to blue light. When the blue light (450-495 nm) is emitted, an active luminescent compound will be excited and luminesce red light (620-750 nm). When O2 diffuses across the sensor membrane, it will interact with the dye and
quench the intensity of the red luminescence. The reduced intensity is measured by a photodetector and used to determine the O2 concentration (Amao 2003).
4.6 Sampling CH4 bubbles
Bubbles of CH4 gas were sampled in the surface sediments. No samples were collected in the
bottom sediments, due to it being difficult to sample CH4 bubbles at that depth. An empty 60
ml syringe, with a connected three-way valve, was attached to an inverted funnel (Fig 6). The funnel was submersed under water whilst a long icepick was used to disturb the surface sediments (stimulating release of CH4 bubbles). The sample (60 ml) was collected by closing
the syringe with its piston under water. The syringe was then removed from the funnel. The excess water was pushed out in order to isolate the CH4 bubbles in the syringe. In the final
step, the sample was transferred to a 20 ml vial that was evacuated in the field (see section 4.4.1 for the evacuation procedure). This sampling procedure was repeated for all samples of CH4 bubbles, and the vials were stored in polyethylene bags until analysis.
Figure 6: An inverted funnel that is connected to a syringe,
used to collect CH4 bubbles
4.7 Sample quantity
In the depth profile, one 1.21-L polycarbonate bottle (Fig 5.) was filled with sample water at every depth (0.3 m, 1.5 m, 3 m, 4.5 m, 6 m, 7 m, 8 m), which resulted in 7 bottles of sample water every month. The only exception is March were water was only sampled at 3 depths. Two samples of CH4 bubbles were collected each month. In total (water samples + CH4
bubbles), 77 samples were collected.
4.8 Laboratory analysis
All samples were analyzed at the Department of Thematic Studies – Environmental Change laboratory, located in Linköping’s University. CH4 concentration and stables isotopes of
carbon were analyzed by using a cavity ring down spectrometer (CRDS) coupled with a Small Sample Isotopic Module (SSIM Picarro G2201-i, Picarro Inc., CA, USA). The CRDS utilizes an optical cavity and single-frequency laser to measure the rate of absorption of gaseous compounds. The instrument has a high precision and sensitivity in measuring stables isotopes of carbon and CH4 concentration (Dickinson et al. 2017), which is why it was chosen for
4.8.1 Preparation for CRDS
Standards
A CH4 standard of 2500 ppm was used for both isotopes (13C/12C). However, the CRDS was
already calibrated for the stable isotopes, before the analysis began. The calibration for the CH4 concentration was performed as a first step in the analysis, where two CH4 standards (10
ppm) were prepared. The first step was to flush an empty 30 ml glass syringe with CH4 gas
(taken from a tank filled with CH4 gas). This was an important step since the syringe could
have been used by someone else working with other compounds. In the next step, 20 ml of CH4 gas was needed however 21 ml was extracted. The reason was that 1 ml of sample was
lost during the CRDS analysis (see section 4.8.2).
Samples
A needle was attached to a three-way valve that was connected to a 30 ml glass syringe. The needle was used to pierce the rubber stopper on the sample vial and transfer the sample to the syringe. Each 20 ml sample vial was charged with 60 ml gas, which created an overpressure inside the vial. This resulted in the gas being “pushed” into the glass syringe, and the needed volume was acquired by stopping the piston at 21 ml and closing the three-way valve. 21 ml sample was extracted due to 1 ml being lost during the CRDS analysis (see section 4.8.2). Precautions taken during sample transfers was to hold the needle and syringe securely when piercing the stopper, since the needle was brittle and there was a risk of it breaking. The needle weakened after 3-4 sample transfer and it had to be changed often. Between every sample transfer, the sample-syringe was flushed with zero-air, in order to clean the syringe. The valve on the syringe was always closed when the syringe wasn’t in use. This was done in order to reduce the risk of air entering the syringe and contaminating it.
4.8.2 Analysis in CRDS
The CRDS is connected to a software called the “SSIM coordinator launcher”. This is where the CRDS logs the results and from where the analysis is run. The analysis was run in time intervals; 30 sec, 2 min 45 sec & 11 min 5 sec. The times were displayed on the computer screen. In conjunction, the software provided information on the procedural steps after each time interval had passed. The first step was to attach the syringe to the CRDS sample port. Then the instrument needed 30 seconds to evacuate the sample compartment. After 30 seconds, the syringe’s valve was opened and due to the under pressure in the sample compartment, 1 ml of sample was pulled in. After 2 minutes and 45 seconds, the rest of the
sample was injected, and the valve was closed. The final analysis was 11 minutes and 5 seconds long. The analysis steps were applied in a highly similar manner by Dickinson et al. (2017) that analyzed the ability of CRDS of detecting stable isotopes of carbon.
4.9 Model application
In the data analysis, when estimating CH4 oxidation from stable isotopes of carbon, the
models for both open-steady state and closed systems (Eq 7 & 8) were applied. See section 3.5 for a description of the equation 7 and 8. The fraction of oxidized CH4 was calculated
between each depth in the profile, in order to analyze the oxidation at a smaller scale. The deepest point out of two depths was regarded as the δ13C-CH4 source, i.e. the CH4 that hasn’t
been exposed to oxidation between two depths. This approach was used for all of the depths. In stratified water columns it has been shown that very little deep CH4 can cross the
metalimnion and reach the surface water (Thalasso et al. 2020). Therefore, most of the surface water CH4 was assumed to come from the surface sediments during the assumed stratification
period. The δ13C-CH4 signature in surface sedimentbubbles was thus used as the source CH4
in the calculations for the epilimnion.
During the assumed stratification period, the total fraction of oxidized CH4 was calculated in
each layer, by using the layer’s bottom and surface δ13C-CH4 signature. The same approach
was applied for the water column, in order to obtain a value that accounts for the entire water column CH4 oxidation. This was done by only using surface and bottom δ13C-CH4 signatures
in the water column. However, the δ13C-CH4 signature in the sedimentbubbles were not
included because this estimate regarded the water column only. Two isotope fractionation factors were used; 1.020 & 1.027. They were experimentally determined in a study by Thottathil et al. (2018), in which the water column CH4 oxidation was analyzed in six lakes.
According to Thottathil et al. (2018), the 1.020 factor is suitable for the surface layer, whilst 1.027 is suitable for the bottom layers, but used both factors for all layers and presented the results as a range, and the present study follows this approach.
4.9.1. Depth profile in stratified water column
Based on the temperature and O2 profiles, the depth profile was divided into three layers,
when Gundlebosjön was assumed to be stratified (June-August). The purpose was to calculate the fraction of oxidized CH4 in each layer. Figure 7 illustrates the depth profile in a water
column, and shows what depths were selected for the epilimnion, metalimnion and hypolimnion for June-August.
Figure 7: Selected depths for the epilimnion, metalimnion and hypolimnion, from June to August, when
5. Results
5.1 δ13C-CH4, CH4, O2 & Temperature profiles
The monthly results were displayed in Figure 8, to clearly show the variation over the measured 9-month period. In Figure 8 the δ13C-CH4 signatures are presented in permille ‰
on the bottom x-axis. The CH4 and O2 concentration are presented in µM and mg/l.
Temperature is displayed in degrees Celsius. The depth profile is visualized on the y-axis. CH4, O2 and temperature are displayed on the top x-axis, and the variables have different
units, but the scale is applicable for all of them. Data labels have been placed in order to show the measured values.
Temperature & O2
Figure 8 shows that the water column has stable temperature and O2 profiles in March and
April. The temperature and O2 levels start to decrease along the depth profile in May (0.3 → 8
m). The decrease is more distinct in June, where the temperature and O2 levels reach 13.3 °C
and 3 mg/l at 8 m. A similar pattern in July, but the temperature is higher along the depth profile and the bottom layer is more anoxic. In August, there is a small variation in the temperature and O2 levels from 0.3 to 4.5 m. A distinct temperature decrease is seen from 4.5
to 7 m (18-14 °C) In the same depth range, O2 levels decrease from 6.9 mg/l to below the
sensor’s detection limit (0.05 mg/l). The water column is still anoxic at 8 m in September. The remaining depths have slightly higher O2 levels and lower temperatures, in comparison
with August. The water temperature is much lower along the depth profile in October
(approx. 8 °C) and November (approx. 5.5 °C). The water column becomes more oxygenated towards November, where the O2 levels reach 10 mg/l.
CH4 & δ13C-CH4
There is a small variation in the δ13C-CH4 signatures and CH4 concentration in March and
April (Fig. 8). The only exception is seen in April, where the CH4 concentration increase from
1 to 5 µM, between 8 and 6 m. In May, the bottom layer has a higher CH4 concentration and
δ13C-CH4 signature than the surface. From 4.5 to 3 m, the δ13C-CH4 signature decreases
from -49.1‰ to -60‰. The δ13C-CH4 signature returns to -54‰ in the surface layer. There is
a similar pattern in June, where the highest CH4 concentration (1.6 µM) and δ13C-CH4
signature (-38.8‰) are found at 8 m and decrease towards the surface. In July, the CH4
concentration and the δ13C-CH4 signature is 3.5 µM and -22.8‰ at 8 m. The CH4
-60‰ towards 1.5 m. In comparison with previous months, the bottom layer in August has the highest CH4 concentration (201-641 µM), and lowest δ13C-CH4 signature (approx. -72‰).
From 7 to 4.5 m, the CH4 concentration decreases to 4.9 µM, whilst the δ13C-CH4 signature
increases to approx. -54‰. In September, the water column has high δ13C-CH4 signatures
along the depth profile, and there isn’t large variation between the depths. The CH4
concentration is relatively low in the water column. In October, the δ13C-CH4 and CH4
profiles are similar throughout the water column. The same pattern is seen in November, but with lower δ13C-CH4 signatures.
Figure 8: The δ13C-CH4, O2, CH4 and temperature profiles for one sample occasion in March-November. Values highlighted in a “box” will further be analyzed in section 6. The data labels showing an O2 concentration of “0 mg/l” means that it was below the sensor’s detection limit of 0.05 mg/. NOTE: The profiles in August are multiplied with a factor of 10, due to large concentration values, the measured values (without multiplication) are displayed in the data labels
5.2 CH4 bubbles
Table 1 shows the CH4 concentration and δ13C-CH4 signatures in the surface (epilimnetic)
sediments in June-August. The δ13C-CH4 signatures were used as source values for the
epilimnetic MOX (methane oxidation) calculations in section 5.3. The CH4 concentration and
δ13C-CH4 signature in the surface sediments for the remaining months are presented in Table
3 in the appendix. In Table 1, the CH4 concentration is presented in parts per million (ppm)
and the δ13C-CH4 signatures in permil ‰. The results are rounded up to two significant
figures. Table 1 shows that the surface sediments had high CH4 concentrations from June to
August. The δ13C-CH4 signature was more negative in July, in comparison with June and
August.
Table 1: The CH4 concentration (ppm) and δ13C-CH4 signatures (‰) in the surface sediments in June-August
June 940000 -59 July 620000 -64 August 750000 -57
5.3 Fraction of oxidized CH4
The difference in δ13C-CH4 signatures in March-May, at certain depths in June-August and in
September-November was too small to estimate the fraction of oxidized CH4. In addition, the
change in the δ13C-CH4 signatures became more negative (i.e. less 13C per 12C) along the
depth profile at those sampling times (Fig. 8). The calculations for the aforementioned months are presented in Tables 4-6 in the appendix.
The depths where the fraction of oxidized CH4 could be estimated are presented in Table 2.
The fractions are written in percentage (%) in order to clearly convey how much is oxidized. The percentage values are rounded up to two significant figures, since they represent an estimation and not a precise value. The percentages are presented as a range since two isotopic fractionation factors were used. The results from equations 7 and 8 are presented alongside each other in order to detect potential similarities or differences. Note that there wasn’t a 6 m measurement in August (Fig. 8), and the 4.5 m depth was instead included in the hypolimnion in August.
Table 2 shows that more CH4 was oxidized in the epilimnion (up to 42% in open-steady state
and up to 37% in closed) in June in comparison with July (up to 22% in open-steady state and 21% in closed). In August, CH4 was mostly oxidized in the hypolimnion, specifically from 7
to 4.5 m. In the hypolimnion, the bottom to surface layer oxidation was 69-93% in open-steady state and 73-99% in closed system. In the epilimnion, 12-16% of the epilimnetic sediment CH4 was oxidized at 3 m, in both open-steady state and closed system. The bottom
to surface epilimnetic oxidation was 13-18% in open-steady state and 11-14% in closed system. The total water column oxidation in August was 68-92% for open-steady state, and 55-68% for closed system. For all MOX calculations, the open-steady state and closed system equations produced relatively similar results.
Table 2: The fraction of oxidized CH4 in open-steady state (eq 7) & closed system (eq 8). The calculated values in are expressed in %. The “-“stands for oxidized CH4. Total B→S represents the water column. Layer B→S represents the entire layer. ES is an abbreviation for ”epilimnetic sediments”. The δ13C-CH4 (‰) value in the epilimnetic sediments was used a source value in the epilimnion, see Table 1 for source values.
June
Epilimnion ES →3 m 1,5 →0.3 m Layer B→S
fopen ≈ (-) 24–33% (-) 15–21% (-) 31–42%
fclosed ≈ (-) 23–30% (-) 15–20% (-) 29–37%
July
Epilimnion ES→ 3 m 1,5 →0.3 m Layer B→S
fopen ≈ (-) 31–41% (-) 14–19% (-) 16–22% fclosed ≈ (-) 28–36% (-) 16–23% (-) 16–21% August Epilimnion ES →3 m Layer B→S fopen ≈ (-) 12–16% (-) 13–18% fclosed ≈ (-) 12–16% (-) 11–14% Hypolimnion 7 →4.5 m Layer B→S fopen ≈ (-) 67–90% (-) 69–93% fclosed ≈ (-) 71–97% (-) 73–98% Total B →S fopen ≈ (-) 68–92% fclosed ≈ (-) 55–68%
6. Discussion
6.1 Reliability & ValidityAn analytical technique can be deemed reliable if it has a good precision, where it consistently produces the same results for one sample. Reliability can be assessed by e.g. repeated measurements on every sample or by viewing the consistency of the results over time. Validity is the extent to which a measurement accurately measures what it is supposed to measure. Validity can be assessed by e.g. comparing the results to previous research (Skoog, Holler & Crouch 2016). In the present study, no repeated measurements were performed for one sample. This can question whether the measurements were reliable.The results were compared to previous studies (see section 6.3-6.4) where the same variables were investigated under similar lake conditions, measured with the same analytical techniques and mathematical models. The comparison showed similar findings and suggested that the
measurements were valid. Moreover, the results in Tables 4-6 (see appendix) were calculated from more negative isotopic signatures. The models by Happel et al. (1994) and Liptay et al. (1998) are based on the fractionation against 13C, where estimations are made from more positive isotopic signatures. However, this doesn’t mean that the estimations in Tables 4-6 are invalid. Instead they suggest that CH4 oxidation was low or didn’t occur in the water column
during those sampling times. However, in order to answer this study’s research question, only the results from Table 2 was analyzed in section 6.4.
6.2 Evaluating the model application
The open-steady state and closed system model was used to estimate the CH4 oxidation in
Gundlebosjön. The depth selection for the stratified layers was based on the O2 and
temperature profiles, which were measured once a month. Stratified layers can vary in depth during the summer months (De Crop & Verschuren 2019). Therefore, the depth selections in the present study can be seen as rough estimates for the summer months. De Crop &
Verschuren (2019) established a monthly “picture” on the depth profiles in the stratified layers, by logging temperature and O2 data at 1 h intervals every day, for a whole year.
However, this can be difficult to achieve, especially if resources (e.g. money, time & people) are limited. In the present study, an improvement would be to log temperature and O2
consistently during the stratification period, which can be during the summer and winter (Fichot et al. 2019). During the winter period, measurements could be done when the lake is not covered with ice.
The MOX calculations only worked for certain depths in August and for the surface layer in June and July, where clear concentration (CH4) and isotope (δ13C) differences was observed
in the water column (Fig. 8). An improvement for future studies is to sample CH4 from
bottom sediments and use its δ13C-CH4 signature as a source value for every depth in the
depth profile. With this improvement, CH4 oxidation could be estimated at a smaller scale
(i.e. between depths) and in the entire water column, more accurately. Mayr et al. (2020) stated that CH4 oxidation is mostly investigated during summer stratification, which was seen
in some of the aforementioned studies, e.g. Lofton et al. (2013); Thottathil et al. (2018); Thalasso et al. (2020). This present study shows that it might not be enough to analyze the water column CH4 oxidation only during summer stratification, since the CH4 dynamics seem
to change over the year.
The fractionation factors from Thottathil et al. (2018) were chosen since they were of a recent study and within the ranges determined in previous studies (see section 3.4.1). Isotopic fractionation factors are dependent on the methanotrophs preference of the lighter isotope of carbon, which can vary with microbial communities (Liptay et al. 1998). The isotopic change that happened during the CH4 oxidation in Gundlebosjön can be different from the change
that Thottathil et al. (2018) detected and quantified. Additionally, biological processes are difficult to precisely quantify (Paterson et al. 2009). Therefore, the calculated values are estimates and not precise values. In future studies, this error can be reduced by experimentally determining the fractionation factor for the analyzed system, by e.g. performing incubation experiments. This wasn’t done for this thesis due to lack of time.
6.3 δ13C-CH4, CH4, O2 & Temperature profiles
The water column was mixed in March and April (Fig. 8). In April, there was a distinct increase in the CH4 concentration and a decrease in δ13C-CH4 signature, from 8 to 6 m. A
disturbance in the bottom sediments, caused by e.g. the release of CH4 bubbles, could explain
the observations between 8 and 6 m. However, it is uncommon that CH4 is transported
through bubbles from the bottom sediments to the surface layer (Thalasso et al. 2020). There could have been an oversaturation of CH4 in bottom sediments and a lower hydrostatic
pressure in April, that eventually caused the release of CH4 bubbles. The beginning of
stratification was visible in May and June, with a distinct temperature gradient between 3 and 4.5 m. Water cannot easily move when there’s a density difference (Pöschke et al. 2015). Hence, the CH4 transport from the bottom sediments will be slower, which can increase its
residence time in the bottom layer (Bastviken 2009). As a result, the bottom layer can have 13C enriched δ13C-CH4 signatures, since the CH4 was longer exposed to oxidation, which is
seen in May and June.
The water column seems to be stratified in July, with a temperature gradient between 3 and 1.5 m. The 8 m measurement had anoxic levels and a high CH4 concentration, but the
δ13C-CH4 signature indicated 13C enrichment. Anaerobic CH4 oxidation could have occurred in the
bottom layer, due to the anoxic levels. For future studies, the concentration of electron acceptors (e.g. SO42-, NO3- and Fe3+) needed for anaerobic oxidation could be measured, in
order to determine if anaerobic CH4 oxidation does occur in a freshwater lake. The CH4
concentration and the δ13C-CH4 signature increased in the surface layer (1.5 → 0.3 m) in July.
The increase in concentration could be due to CH4 production in the surface sediments. The
higher CH4 and O2 availability could have enabled the methanotrophs to oxidize some of the
CH4 before it reached the surface water, which could explain the 13C enrichment from 1.5 to
0.3 m.
In August, the bottom layer (8→7 m) had high CH4 concentrations and δ13C-CH4 signatures
depleted in 13C. The anoxic levels could have provided a suitable environment for
methanogenesis. Similar findings were found in a study by Bastviken et al. (2008), where the anoxic bottom layer in the studied lakes had CH4 concentrations ranging from 50 to 600 µM,
and δ13C-CH4 signatures between -60 to -80‰. CH4 oxidation is dependent on the
availability of CH4, but an electron acceptor such as O2 is needed (Lofton et al. 2013). If other
electron acceptors are available, e.g. SO42-, NO3- and Fe3+, then anaerobic oxidation could
take over (Whalen 2005). In this present study, the depleted δ13C-CH4 signature indicated
that oxidation didn’t occur from 8 to 7 m. Nevertheless, the accumulation of CH4 could have
been due to a density difference between the bottom and surface layer. This would have limited the transport of the bottom CH4 to the surface layer. The water column seemed to be
mixed from 4.5 to 0.3 m. This suggests that the water column was not fully stratified, which questions the depth selections in Figure 7.
In comparison with August, CH4 didn’t accumulate in the bottom layer in September. Figure
8 showed that the temperature profile was stable along the depth profile in September, suggesting that the water column was mixed. Water column mixing can cause storage flux, where accumulated CH4 in the hypolimnion is rapidly transported to the surface water and
emitted to the atmosphere (Bastviken 2009). This could have occurred between the measurement in August, where CH4 accumulation was found, and the measurement in
September. The surface layer in September had a lower concentration of CH4, which could be
a result of CH4 emissions from the surface layer. The transport of CH4 could have been slower
in the water column, thereby exposing CH4 longer to oxidation, which could explain the 13C
enrichment along the depth profile. The same interpretation, regarding slow CH4 transport,
can be made for the δ13C-CH4 profiles in the water column in October and November. In
addition, the temperature profiles suggested that the water column was still mixed in October and November.
6.4 Water column CH4 oxidation
In June-August, the δ13C-CH4 signature in the surface sediments (Table 1) was used as a
source value for the epilimnion. There was a clear difference in CH4 concentration and
δ13C-CH4 signature, which could be the reason why MOX calculations worked. However, the
MOX results for June and July only showed how much of the surface sediment CH4 was
oxidized. They didn’t give an indication of the water column CH4 oxidation. Nevertheless, the
epilimnion can be at open-steady state, assuming there’s an equal gas-exchange between the surface layer and the atmosphere (DelSontro et al. 2017). However, Thottathil et al. (2018) argued that the closed system model is also applicable for the epilimnion, due to gas evasion from the epilimnion resulting in CH4 continuously leaving the layer. This coincides with the
Rayleigh distillation function (Eq 9), from which the closed system model (Eq 8) is derived from, that assumes that a compound is continuously removed from a system (Liptay et al. 1998). Therefore, at open-steady state and in closed system, more of the surface sediment CH4 was oxidized in June, in comparison with July and August. CH4 is known to quickly be
emitted from the surface layer and escape CH4 oxidation (Bastviken 2009). There could have
also been disturbances in the surface sediments, caused by e.g. the sampling procedure or ebullition, resulting in less oxidized sediment CH4 in July and August.
As mentioned in section 3.5.1, the hypolimnion can be regarded as a closed system during stratification. Assuming the hypolimnion was a closed system in August, then up to 97% of CH4 was oxidized between 7 to 4.5 m. According to Figure 8, this could be the interface
between the oxic and anoxic region in the water column. As shown in Figure 8, there was an accumulation of CH4 in the anoxic hypolimnion. O2 could have been the limiting factor
the CH4. Similar findings were found in the study by Bastviken et al. (2008), where approx.
80% of CH4 was oxidized daily in the oxic-anoxic interface.
In August, the MOX values in the hypolimnion were higher in comparison with epilimnion. A similar pattern was seen in the study by Thottathil et al. (2018), but most of the study’s MOX values in the hypolimnion were overestimates indicating that more than 100% of CH4 was
oxidized. Both Thottathil et al. (2018) and the present study had very similar source values in the hypolimnion. However, the δ13C-CH4 signature for the CH4 exposed to oxidation,(i.e. the
δs value), could have been different in both studies. In terms of total water column CH4
oxidation, the open-steady state model yielded the highest estimate of oxidized CH4
(68-92%), in comparison with the closed system model (55-68%). Considering the difference seen in the concentration (CH4) and δ13C-CH4 signature between 8 and 0.3 m in August (Fig. 8),
the open-steady state estimation could be more representative of the water column CH4
oxidation. Nevertheless, the estimates show the importance of CH4 oxidation, especially when
CH4 accumulates in the bottom layer (Fig. 8). CH4 oxidation could have resulted in less
7. Conclusion
The conclusion is formed as a response to the research question in section 2.
Water column CH4 oxidation happened in August, with the extensive CH4 oxidation
occurring in the region between oxic and anoxic layer. CH4 oxidation also occurred in the
surface layer in June and July. The stable isotope method to estimate CH4 oxidation, between
depths and in the entire water column, was only useful during periods when clear
concentration (CH4) and isotope (δ13C) differences could be observed in the water column.
Sampling CH4 from the bottom sediments and using δ13C-CH4 signature as a source value in
8. References
Amao, Y. (2003). Probes and Polymers for Optical Sensing of Oxygen. Microchimica Acta, Vol. 143, pp.1-12.
Bastviken, D. et al. (2002). Measurement of Methane Oxidation in Lakes: A Comparison of Methods. Environmental Science & Technology, Vol. 36, no. 15, pp.3354-3361.
Bastviken, D. et al. (2008). Fates of methane from different lake habitats: Connecting whole-lake budgets and CH4 emissions. Journal of Geophysical Research, Vol. 113, pp.1–13.
Bastviken, D. 2009. Methane. Elsevier. Encyclopedia of inland Waters, Vol. 2, pp.783–805.
Berden, G. et al. (2010). Cavity ring-down spectroscopy: Experimental schemes and applications. International Reviews in Physical Chemistry, Vol. 19, no. 4, pp.565-607.
Blaser, M. & Conrad, R. (2016). Stable carbon isotope fractionation as tracer of carbon cycling in anoxic soil ecosystems. Current Opinion in Biotechnology, Vol. 41, pp.122-129.
Conrad, R. & Claus, P. (2009). Characterization of stable isotope fractionation during methane production in the sediment of a eutrophic lake, Lake Dagow, Germany. Limnology Oceanography, Vol. 54, no. 2, pp.457-471.
De Crop, W. et al. (2019). Determining patterns of stratification and mixing in tropical crater lakes through intermittent water-column profiling: A case study in western Uganda. Journal of African Earth Sciences, Vol. 153, pp.17-20.
De Mello, N. et al. (2018). Spatial variability of methane (CH4) ebullition in a tropical hypereutrophic reservoir: silted areas as a bubble hot spot. Lake and Reservoir Management, Vol. 34, no. 2, pp.105-114.
DelSontro, T. et al. (2017). No Longer a Paradox: The Interaction Between Physical Transport and Biological Processes Explains the Spatial Distribution of Surface Water Methane Within and Across Lakes. Ecosystems, pp.1-15.
Dickinson, D. et al. (2017). System for δ13C-CO2 and xCO2 analysis of discrete gas samples
by cavity ring-down spectroscopy. Atmospheric Measurement Techniques, Vol. 10, pp.4507-4519.
Dlugokencky, J, E. et al. (2009). Observational constraints on recent increases in the atmospheric CH4 burden. Geophysical Research Letters, Vol. 36, no. 18, pp.1-5.
Essenhigh, H, R. (2009). Potential Dependence of Global Warming on the Residence Time (RT) in the Atmosphere of Anthropogenically Sourced Carbon Dioxide. Energy & Fuels, Vol. 23, no. 5, pp.2773-2784.
Fichot, G, C. et al. (2019). Assessing change in the overturning behavior of the Laurentian Great Lakes using remotely sensed lake surface water temperatures. Remote Sensing of Environment, Vol. 235, pp.1-16.
Fischer, E, R. et al. (2017). Measurement of the 13C isotopic signature of methane emissions
from northern European wetlands. Global Biogeochemical Cycles, Vol. 31, pp. 605-623.
Fuchs, A. et al. (2016). Effects of increasing temperatures on methane concentrations and methanogenesis during experimental incubation of sediments from oligotrophic and mesotrophic lakes. Journal of Geophysical Research: Biogeosciences, Vol. 121, no. 5, pp.1394-1406.
Happel, D, J. et al. (1994). The influence of methane oxidation on the stable isotopic
composition of methane emitted from Florida swamp forests. Geochimica et Cosmochimica Acta, Vol. 58, pp.4377-4388.
He, R. et al. (2012). Identification of functionally active aerobic methanotrophs in sediments from an arctic lake using stable isotopes probing. Environmental Microbiology, Vol. 14, no. 6, pp. 1403-1419.
Hilgeman, R, F. et al. (2007). Using Dalton’s Law of Partial Pressures To Determine the Vapor Pressure of a Volatile Liquid. Journal of Chemical Education, Vol. 84, no. 3, pp.469-470.
Holland, R, P. & Kay, A. (2003). A review of the physics and ecological implications of the thermal bar circulation. Limnologica, Vol. 33, no. 3, pp.153-162.
Kirschke, S. et al. (2013). Three decades of global methane sources and sinks. Nature Geoscience, Vol. 6, no. 10, pp.1-66.
Kremser, A. et al. (2016). Systematic comparison of static and dynamic headspace sampling techniques for gas chromatograph. Bioanalytical Chemistry, Vol. 408, no. 24, pp.6567-6579.
Leonte, M. et al. (2018). Using Carbon Isotope Fractionation to Constrain the Extent of Methane Dissolution Into the Water Column Surrounding a Natural Hydrocarbon Gas Seep in the Northern Gulf of Mexico. Geochemistry, Geophysics, Geosystems, Vol. 29, no. 11, pp.4459-4475.
Liptay, K. et al. (1998). Use of stable isotopes to determine methane oxidation in landfill cover soils. Journal of Geophysical Research, Vo. 103, no. D7, pp.8243-8250.
Lofton, D, D. et al. (2013). Effect of temperature on methane dynamics and evaluation of methane oxidation kinetics in shallow Arctic Alaskan lakes. Hydrobiologia, Vol. 721, no. 1, pp.209-222.
Mayr, J, M. et al. (2020). Growth and rapid succession of methanotrophs effectively limit methane release during lake overturn. Communications Biology, Vol. 3, no. 168, pp.1-9.
Montzka, A, S. et al. (2011). Non-CO2 greenhouse gases and climate change. National
Oceanic and Atmospheric Administration, Vol. 476, no. 7358, pp.43-50.
Myhre, G. et al. (2013). Anthropogenic and Natural Radiative Forcing, in Stocker, T., Qin, D., Plattner, G.-K., Tignor, M., Allen, S., Boschung, J., Nauels, A., Xia, Y., Bex, V., Midgley, P. (eds.), Climate Change 2013: The Physical Science Basis. Contribution of Working Group
I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change,
Pang, J. et al. (2016). Intercomparison of two cavity ring-down spectroscopy analyzers for atmospheric 13CO2/12CO2 measurement. Atmospheric Measurement Techniques, Vol. 9, no. 8,
pp.3879-3891.
Paterson, E. et al. (2009). Through the eye of the needle: a review of isotope approaches to quantify microbial processes mediating soil carbon balance. New Phytologist, Vol. 184, no. 1, pp.19-33.
Pöschke, F. et al. (2015). Upwelling of deep water during thermal stratification onset- A major mechanism of vertical transport in small temperature lakes in spring? Water Resources Research, Vol. 51, pp.9612-9627.
Samhällsbyggnadsnämnden (2016). Blåplan 1: Översiktlig del. Vänersborgs kommun. https://www.vanersborg.se/download/18.6eacfb51158c2f92394c478/1480949759318/Bl%C3 %A5plan%20del%201%20-%20%C3%B6versiktlig%20del.pdf
Saunois, M. et al. (2016). The global methane budget 2000-2012. Earth System Science Data Discussions, Vol. 8, no. 2, pp.607-751.
Saunois, M. et al. (2019). The global methane budget 2000-2017. Earth System Science Data Discussions, pp.1-136.
Sawakuchi, O, H. et al. (2016). Oxidative mitigation of aquatic methane emissions in large Amazonian rivers. Global Change Biology, Vol. 22, no. 3, pp. 1075-1085.
Schubert, C, J. et al. (2011). Evidence for anaerobic oxidation of methane in sediments of a freshwater system (Lago di Cadagno). FEMS Microbiology Ecology, Vol. 76, no. 1, pp.26-38.
Shelly, F. et al. (2015). Microbial methane cycling in the bed of a chalk river: oxidation has the potential to match methanogenesis enhanced by warming. Freshwater Biology, Vol. 60, no. 1, pp.150-160.
Skoog, Douglas A., Holler, James F. & Crouch, Stanley R. (2016). Principles of Instrumental
Srivastava, A. & Verkouteren, M, R. (2018). Metrology for stable isotope reference materials:
13C/12C and 18O/16O isotope ratio value assignment of pure carbon dioxide gas samples on the
Vienna PeeDee Belemnite-CO2 scale using dual-inlet mass spectrometry. Analytical and
Bioanalytical Chemistry, Vol. 410, pp.4153-4163.
Templeton, S, A. et al. (2006). Variable carbon isotope fractionation expressed by aerobic CH4-oxdizing bacteria. Geochimica et Cosmochimica Acta, Vol. 70, pp.1739-1752.
Thalasso, F. et al. (2020). Sub-oxycline methane oxidation can fully uptake CH4 produced in
sediments: case study of lake in Siberia. Scientific Reports, Vol. 10, no. 1, pp.1-7.
Thottathil, D, S. et al. (2018). The Extent and Regulation of Summer Methane Oxidation in Northern Lakes. Journal of Geophysical Research: Biogeosciences, Vol. 123, no. 10, pp.3216-3230.
Visscher, D, A. et al. (2004). Isotope fractionation effects by diffusion and methane oxidation in landfill cover soils. Journal of Geophysical Research, Vol. 109, pp.1-8.
Warneck, Peter. & Williams, Jonathan. (2012). The Atmospheric Chemist’s Companion.
Numerical Data for Use in the Atmospheric Sciences.
Whalen, C, S. (2005). Biogeochemistry of Methane Exchange between Natural Wetlands and the Atmosphere. Environmental Engineering Science, Vol. 22, no. 1, pp. 73–94.
Whiticar, J, M. (1999). Carbon and hydrogen isotope systematics of bacterial information and oxidation of methane. Chemical Geology, Vol. 161, pp.291-314.
Wilson, M, R. (2019). Microbial Community Analyses Inform Geochemical Reaction Network Models for Predicting Pathways of Greenhouse gas Production. Frontiers in Earth Science, Vol. 7, no 59.
Woolway, L, R. et al. (2014). A novel method for estimating the onset of thermal
stratification in lakes from surface water measurements. Water Resources Research, Vol. 50, pp.5131-5140.
Ye, R. et al (2013). Homoacetogenesis: A potentially underappreciated carbon pathway in peatlands. Soil Biology and Biochemistry, Vol. 68, pp. 385–391.
