Influence of temperature on the δ13C values and distribution of methanotroph‐related hopanoids in Sphagnum‐dominated peat bogs

Abstract Methane emissions from peat bogs are mitigated by methanotrophs, which live in symbiosis with peat moss (e.g. Sphagnum). Here, we investigate the influence of temperature and resultant changes in methane fluxes on Sphagnum and methanotroph‐related biomarkers, evaluating their potential as proxies in ancient bogs. A pulse‐chase experiment using 13C‐labelled methane in the field clearly showed label uptake in diploptene, a biomarker for methanotrophs, demonstrating in situ methanotrophic activity in Sphagnum under natural conditions. Peat cores containing live Sphagnum were incubated at 5, 10, 15, 20 and 25°C for two months, causing differences in net methane fluxes. The natural δ13C values of diploptene extracted from Sphagnum showed a strong correlation with temperature and methane production. The δ13C values ranged from −34‰ at 5°C to −41‰ at 25°C. These results are best explained by enhanced expression of the methanotrophic enzymatic isotope effect at higher methane concentrations. Hence, δ13C values of diploptene, or its diagenetic products, potentially provide a useful tool to assess methanotrophic activity in past environments. Increased methane fluxes towards Sphagnum did not affect δ13C values of bulk Sphagnum and its specific marker, the C23 n‐alkane. The concentration of methanotroph‐specific bacteriohopanepolyols (BHPs), aminobacteriohopanetetrol (aminotetrol, characteristic for type II and to a lesser extent type I methanotrophs) and aminobacteriohopanepentol (aminopentol, a marker for type I methanotrophs) showed a non‐linear response to increased methane fluxes, with relatively high abundances at 25°C compared to those at 20°C or below. Aminotetrol was more abundant than aminopentol, in contrast to similar abundances of aminotetrol and aminopentol in fresh Sphagnum. This probably indicates that type II methanotrophs became prevalent under the experimental conditions relative to type I methanotrophs. Even though BHP concentrations may not directly reflect bacterial activity, they may provide insight into the presence of different types of methanotrophs.


| INTRODUC TI ON
Peat bogs play an important role in the global carbon cycle since they sequester one-third of the Earth's terrestrial carbon (Smith et al., 2004). Peat bogs are responsible for approximately 10% of the total methane flux to the atmosphere, having a major impact on atmospheric methane concentrations (Gorham, 1991). High latitudes are especially vulnerable to climate change, and it is, therefore, important to understand the effect of changing environmental conditions on carbon and net methane flux rates in peat bogs.
Aerobic methane oxidation by bacteria (methanotrophs) is a significant terrestrial methane sink and hence plays an important role in the global methane cycle (Dean et al., 2018;Hanson & Hanson, 1996). In peat bogs, methanotrophs live in symbiosis with peat moss (Sphagnum), living inside the hyaline cells of Sphagnum and on the stems. There they reduce methane emissions and provide CO 2 and fixed nitrogen to Sphagnum when growing under submerged conditions Larmola et al., 2010Larmola et al., , 2014Raghoebarsing et al., 2005). In order to predict the consequences of perturbations in the global carbon cycle, a better understanding of the influence of environmental factors on methanotrophs and the methane cycle in peat bogs is essential. Mesocosm experiments have revealed that both methane production and oxidation increase with increasing temperature, albeit that the extent of methane retention by the Sphagnum-methanotroph consortium strongly decreases with increasing temperature (van Winden, Reichart, McNamara, Benthien, & Sinninghe Damsté, 2012). This confirmed the hypothesis that peat bogs may act as a positive feedback to global warming. Assessing the long-term impact of changing environmental conditions on methanotrophs and the methane cycle in peat bogs, however, also relies on reconstructions of methane fluxes in ancient peats.
Ombrotrophic peat bogs receive their nutrients solely via precipitation, which makes them relatively nutrient poor. The low pH buffering capacity, cation exchange and humic acid release by Sphagnum mosses result in a strongly acidic environment (Soudzilovskaia et al., 2010, and references therein). Due to the low pH of the bog water, the reservoir of exchangeable inorganic carbon is low and the level of diffusion of atmospheric CO 2 into the bog water is also little, which makes submerged-growing Sphagnum mosses CO 2 -limited and to a large extent dependent on CO 2 derived from organic matter degradation (Smolders, Tomassen, Pijnappel, Lamers, & Roelofs, 2001). Methanotrophs supply methane-derived CO 2 , which may account up to 35% of the carbon uptake by Sphagnum Larmola et al., 2010).
The knowledge on environmental controls, including temperature, on methanotrophy is still limited. If palaeoclimate archives could be used to study the controls on methane fluxes in peat bogs, including temperature and humidity, this would open up additional sources of information. While advanced microbial techniques exist to study current microbial community structures (see Yates, Nakatsu, Miller, & Pillai, 2016), the limited longevity of genetic material, however, limits the use of advanced microbial tools to reconstruct changes through time. Changes in past methanotrophic activity in peat bogs can potentially be studied using hopanoids. These bacterial markers are common lipids of methanotroph, also those occurring in Sphagnum peats (van Winden, Talbot, Kip, et al., 2012). However, they are not exclusive to methanotrophs since many other bacteria produce them as well (Belin et al., 2018;Farrimond et al., 1998;Rohmer, Bouvier-Nave, & Ourisson, 1984;Sinninghe Damsté, Rijpstra, Dedysh, Foesel, & Villanueva, 2017;Sinninghe Damsté et al., 2004). Bacteriohopanepolyols (BHPs) have a polyfunctionalized side chain attached to the pentacyclic carbon skeleton, to which a variety of polar groups can be attached .
In addition, in the Cobham lignite, a Paleogene wetland deposit, δ 13 C values down to −42‰ and −76‰ were observed for C 31 and C 29 17β,21β(H)-hopanes, respectively (Pancost et al., 2007). These depleted isotope signatures were interpreted to be indicative of increased methanotrophic activity. In our previous study, Sphagnum from settings with highest methane oxidation rates exhibited the most depleted hopene δ 13 C values, in contrast to δ 13 C values of the derived 17β,21β(H)-bisohomohopanol, and therefore, δ 13 C values of hopenes probably reflect methanotrophic activity .
In this study, we investigate the influence of temperature and the resultant changes in methane flux rates on Sphagnum and methanotroph-related hopanoids, evaluating their potential as proxies for methanotrophs and methane fluxes in ancient bogs. To this end, Sphagnum mosses, incubated at different temperatures in a previously described mesocosm experiment , were analysed for their compound-specific carbon isotopic composition and BHP composition. These parameters were subsequently compared to methane flux rates, to evaluate their potential as proxies for methane fluxes.

| Site description
The field site used in this study was in Moorhouse Nature Reserve in the North Pennines, UK, and the site has been described in detail elsewhere (McNamara et al., 2008 and references therein

| Field pulse-chase experiment
Large PVC tubes (height 1 m; width: 23.5 cm) were inserted into the peat, to prevent lateral migration of labelled methane to be injected. This was done six weeks prior to the in situ experiments to allow the environment to recover. Eriophorum was removed from the study site, to examine the Sphagnum-associated methanotrophs exclusively. We injected 100 ml demineralized water, saturated with 99 atom% 13 C-labelled methane, into the peat within the PVC tubes at 40 cm depth, roughly corresponding to a depth where methane is typically produced (Hornibrook, Bowes, Culbert, & Gallego-Sala, 2009). Injection was performed using a 100-ml syringe attached to a thin metal rod (1 mm inner diameter) via six 1 mm openings on the sides at the base of the rod. Sphagnum was harvested from parallel tubes at days 2, 6, 11 and 14 consecutively, and sectioned into top, middle and lower parts. Lower parts were analysed for the 13 C values of the diploptene. Sphagnum is a very basic plant, growing from the top and decaying at the bottom. After two months, when the photosynthetic capitulum was newly grown, the whole Sphagnum plant was harvested and sectioned into top parts (capitulum) and lower parts. Top parts of Sphagnum mosses were analysed for the δ 13 C of Sphagnum-derived C 23 n-alkane to capture only lipids grown under the experimental conditions. The lower parts, where methanotrophs reside (Raghoebarsing et al., 2005;van Winden, Talbot, Kip, et al., 2012), were analysed for BHPs and δ 13 C of a bacterial lipid (diploptene). The bulk carbon isotopic composition was measured of both the top and lower parts of the Sphagnum mosses.

| Extraction and derivatization of lipids
Total extracts of freeze-dried Sphagnum mosses (whole plant) derived from the field experiment were obtained with an Accelerated Solvent Extractor (Dionex), using a mixture of DCM:MeOH (9:1, v/v). Top parts and lower parts of freeze-dried Sphagnum mosses obtained from the temperature experiments were extracted using a modified Bligh and Dyer extraction procedure, to enable BHP analyses (Talbot, Rohmer, & Farrimond, 2007). An aliquot of the extract was separated into an apolar fraction and a polar fraction over an activated Al 2 O 3 column using hexane:DCM (9:1 v/v) and DCM:MeOH (1:1 v/v). To purify samples for hopene analyses (lower Sphagnum parts), apolar fractions were treated by urea adduction. Apolar fractions were dissolved in 200 μl urea-saturated methanol, 200 μl acetone and 200 μl hexane and shaken. After 30 min at −20°C, solvents were evaporated under a stream of nitrogen. The urea crystals, containing the adductable normal and iso alkanes, were washed with hexane three times. To obtain the non-adductable branched and cyclic hydrocarbons (including the hopenes), the urea crystals were washed with hexane three times.
The wash solvents were collected, dried under nitrogen and dissolved in hexane prior to GC analyses.
For BHP analyses, another aliquot of the total extract of the lower Sphagnum parts was acetylated using acetic anhydride and pyridine (1:1) at 50°C for 1 hr and left at room temperature overnight, after addition of the internal standard pregnanediol. The acetylated extract was dried at 50°C under a continuous nitrogen flow and dissolved in MeOH/propane-2-ol (60:40 v/v), prior to LC/MS analyses.

| Gas chromatography (GC), GC/mass spectrometry (GC/MS) and GC-isotope ratio monitoring mass spectrometry (IRMS), Elemental Analyzer (EA)-IRMS
Hopene-containing fractions were analysed on a gas chromatograph (HP 6890) equipped with a flame ionization detector (FID) set at constant pressure (100 kPa). A fused silica column (30 m × 0.32 mm i.d., film thickness 0.1 μm) coated with CP Sil-5CB was used with helium as a carrier gas. Extracts were injected on-column at 70°C. The temperature increased with 20°C/min to 130°C and 4°C/min to 320°C, followed by an isothermal hold for 20 min. Components were identified using gas chromatography-mass spectrometry (Thermo Trace GC Ultra), using the same column and temperature programme as for GC analyses.
Compound-specific δ 13 C values were determined by isotope ratio monitoring-gas chromatography mass spectrometry (GC-IRMS, Thermo Fisher Delta V), using the same column and temperature programme as for GC analyses. Carbon isotopic values are reported in the delta notation relative to the VPDB standard.
Bulk carbon isotopes were analysed on homogenized subsamples from the lower and upper parts of the Sphagna. To minimize the impact of isotopic differences between different plant parts about ten times, the required amount of carbon was weighed in tin sample cups. The CO 2 produced was subsequently diluted with He before being introduced online into the mass spectrometer (Finnigan Delta plus). Accuracy and precision based on replicate analyses of samples and an in-house standard (calibrated to VPDB using international standards) were <0.1‰.

F I G U R E 1
In situ pulse-chase experiment showing 13 C incorporation (in atomic %) into the bacterial lipid diploptene through time, after injection of 13 C-labelled methane into the peat. Compound-specific carbon isotopes were measured on the lower parts of the pool-derived Sphagnum, where potential methane oxidation rates were highest (van Winden, Reichart, et al., 2012)

| High-performance liquid chromatographyatmospheric pressure chemical ionization mass spectrometry (HPLC/APCI-MS)
Acetylated total extracts were analysed for BHPs by reversed-phase high-performance liquid chromatography (HPLC). The derivatization and analytical procedure have been described previously (Cooke, Talbot, & Farrimond, 2008). In short, reversed-phase HPLC analysis was carried out using a Surveyor HPLC system (ThermoFinnigan) fitted with a Gemini (Phenomenex) C 18 5 μm HPLC column (150 mm, 3.0 mm i.d.) and a security guard column of the same material.
Separation was achieved at a flow rate of 0.5 ml/min at 30°C with the following gradient profile: 90% A and 10% B (0 min); 59% A, 1% B and 40% C (at 25 min), then isocratic (to 40 min) and returning to the starting conditions over 5 min and finally stabilizing for 15 min before the next injection, where A = MeOH, B = water and C = propan-2-ol (all Fisher Scientific HPLC grade). LC/MS was performed using a ThermoFinnigan LCQ ion trap mass spectrometer equipped with an APCI source operated in positive ion mode. LC/MS analysis was carried out in data-dependent mode with three scan events: SCAN 1-full mass spectrum, range m/z 300-1,300; SCAN 2-datadependent MS 2 spectrum of most abundant ion from SCAN 1; and SCAN 3-data-dependent MS 3 spectrum of most abundant ion from SCAN 2. Quantification was performed using m/z traces targeting the characteristic base peak ions and calculated using averaged response factors for a limited suite of nitrogen and non-nitrogen containing BHP standards where N-containing BHPs give an averaged response 12 times that of the internal standard and those with no nitrogen 8 times that of the standard (Cooke et al., 2008). Accuracy of the quantification was ±20%.

| RE SULTS
To establish in situ methane uptake by methanotrophs and verify methanotroph-derived lipids, an isotopic pulse-chase experiment using 13 C-labelled methane was performed in the field. Within 14 days, a shift of 0.015% in 13 C, or 15‰ in δ 13 C, was observed in diploptene, the most abundant hopene in the lower parts of the poolderived Sphagnum (Figure 1). The magnitude and evolution of the shift in 13 C indicate uptake of the 13 C-labelled methane and, hence, evidence for diploptene as marker for methanotrophs.
To study potential methanotrophy proxies, intact peat cores containing living Sphagnum spp. obtained from the field were incubated at 5, 10, 15, 20 and 25°C for two months. We have previously demonstrated that with increasing temperature, methane production and consumption (presumably through methanotrophy) increases . Therefore, this set of experiments is useful to evaluate biomarker proxies for methanotrophy. The concentrations of diploptene were highest in the lower parts of the Sphagnum mosses obtained from the mesocosm temperature experiments but did not vary significantly (p > .01) with increasing incubation temperature (Figure 2a). However, the δ 13 C values of diploptene showed a strong negative relation with temperature (p < .01, R 2 = .84, Figure 2b), from −33.9‰ (at 5°C) to −40.7‰ (at 25°C). A strong correlation between methane production and diploptene δ 13 C values (p > .01, R 2 = .84; Figure 3) was also observed.
The summed BHP concentration in Sphagnum showed an increase with increasing temperature upon incubation (Figure 4a).
The concentrations of bacteriohopanetetrol (BHT) showed consistently increasing concentrations with increasing temperature from 10°C onwards (Figure 4b), while those of aminotriol did not increase until 25°C (Figure 4c), where a sudden fourfold increase was noted.  (2009) reported that bacteria in a vegetative state produce more BHPs than when active, implying that BHP abundance does not necessarily reflect bacterial activity. This potentially complicates the use of BHPs as a palaeoenvironmental proxy for methanotrophic activity.

Concentrations
Contrary to our previous study on fresh material from the same site where aminotetrol and aminopentol displayed equal abundances (van Winden, Talbot, Kip, et al., 2012), here aminotetrol concentrations were almost five times higher than those of aminopentol (Figure 4d,e). Aminotetrol is primarily produced by type II methanotrophs, but also, to a minor extent, by type I methanotrophs and some sulphate-reducing bacteria (Blumenberg et al., 2006;Neunlist & Rohmer, 1985b;Rohmer et al., 1984;Talbot et al., 2001;van Winden, Talbot, Kip, et al., 2012). Considering the low sulphate levels in peat bogs, sulphate-reducing bacteria are not expected to be abundant in these settings; hence, type II methanotrophs and, to a lower extent type I, methanotrophs are expected to be the predominant producers of aminotetrol here.
Therefore, most likely, type II methanotrophs were more prevalent than type I methanotrophs in our incubation experiments. This could be caused by experimental conditions being more favourable for the production of aminotetrol by type II methanotrophs, and/ or these methanotrophs were better able to adapt to the perturbation. Still, some caution should be taken with this interpretation, as Osborne et al. (2017) and the results of our study demonstrate that the bacterial BHP production in response to environmental conditions is non-linear. Furthermore, Osborne et al. (2017) found that aminotriol production stabilized after 10 days, while aminotetrol and aminopentol concentrations still increased with time, even after 28 days. Our incubation experiment lasted for 2 months, but it is possible that the duration of the experiment may have been insufficient for the methanotrophic community to fully adapt and express the effect of temperature on bacterial growth and BHP production.
Abundances of aminotetrol and aminopentol were small relative to the summed BHP abundance (Figure 4). This suggests that methanotrophs represent only a minor fraction of the BHP-producing bacterial community. In fact, the most abundant BHPs in Sphagnum mosses are unsaturated BHT pentose, BHT pentose, BHT cyclitol ether and adenosylhopanes, which are not produced by methanotrophs (van Winden, Talbot, Kip, et al., 2012). Not all type II methanotrophs produce aminotetrol, and this trait appears to be confined to type II methanotrophs belonging to the Methylocystaceae family.
Methylocella-like type II methanotrophs only produce BHT and aminotriol (van Winden, Talbot, Kip, et al., 2012;Talbot et al., unpublished data). Aminotriol and BHT are significantly more abundant Analyses of compound-specific isotopes of intact BHPs would enable a tremendous step forward in the understanding of their sources. Recently, a novel method was proposed to measure the 13 C composition of BHPs, using semi-preparative ultrahigh pressure liquid chromatography followed by high-temperature gas chromatography-isotope ratio mass spectrometry (Hemingway et al., 2018).
This method, or high-temperature gas chromatography, as recently developed by Lengger et al. (2018), would allow this and may provide more insight into the use of methanotroph-derived BHPs.
Aminotetrol and aminopentol were even present in the 55 Ma old Cobham lignite (Talbot, Bischoff, Inglis, Collinson, & Pancost, 2016), indicating that these BHPs can be used to trace back methanotro-

| Diploptene δ 13 C as a potential proxy for methanotrophy in ancient peat
The field pulse-chase experiment using 13 C-labelled methane clearly showed label incorporation into diploptene by the magnitude of the isotopic shift of 0.015% 13 C or 15‰ δ 13 C ( Figure 1). Therefore, these results confirm the in situ methanotrophic activity in Sphagnum under natural conditions Raghoebarsing et al., 2005) and make it likely this also occurs in the mesocosm experiments performed in this study. The amended 13 C-labelled methane represented only a fraction of the total available methane and is also not homogenously transported through the contained Sphagnum and peat. It is, therefore, not surprising that the isotopic shift in the δ 13 C of diploptene is not larger than the 15‰ observed. This label uptake in diploptene, which is in line with previous laboratory stud- ies (van Winden et al., 2010), confirms its, at least partial, methanotrophic origin (Rohmer et al., 1984). Moreover, these results show a direct and fast response of the δ 13 C of diploptene to changes in the δ 13 C of the methane. In the studied Sphagnum moss, diploptene (hop-22(29)-ene) was the most abundant C 30 hopene, while in other Sphagnum mosses or peats, its precursor lipid, diplopterol (Raghoebarsing et al., 2005)  The concentrations of diploptene in the Sphagnum moss from the mesocosm experiments did not show significant changes with increasing temperature (Figure 2a), suggesting no major increase in biomass of the associated methanotrophs. In contrast, diploptene δ 13 C values showed a strong correlation with temperature ( Figure 2b). Consequently, there is also a strong correlation of diploptene δ 13 C values with methane production (Figure 3), which has previously been established by measurement of the methane flux from the Sphagnum cores when the methanotroph-containing F I G U R E 4 The summed BHP abundances (a) and the abundances of (b) BHT, (c) aminotriol, (d) aminotetrol and (e) aminopentol, specific BHPs produced by methanotrophs (b-e), in the lower parts of the Sphagnum sp. of the mesocosm experiments versus incubation temperature (°C). All concentrations are expressed in μg/g dry weight. Note the differences in scale. The accuracy of the concentrations is ±20%. Stars indicate the average BHP abundance in the middle and lower parts of Sphagnum sp. mosses as determined in fresh field samples (data from van Winden, Talbot, Kip, et al., 2012). The summed BHP concentrations do not reflect the summed concentration of the BHPs shown in panels b-e, as the most abundant BHPs, BHT pentose, BHT cyclitol ether and adenosylhopanes, are BHPs not produced by methanotrophs (see text) and therefore not shown here Sphagnum layer was removed .
If diploptene δ 13 C values would be controlled by temperature directly, one would expect an opposite trend than observed ( Figure 2b); at higher temperatures, chemical reactions are faster and isotopic discrimination is expected to decrease with increasing temperature (Hayes, 2001). This was, for example, observed by Jahnke, Summons, Hope, and Des Marais (1999) in their study of the effect of growth temperature on isotopic fractionation in methanotroph cell cultures. Therefore, the decreasing δ 13 C values of diploptene could be interpreted as being caused by the enhanced methane flux related to the increase in temperature.
The decreasing δ 13 C values of diploptene could, thus, be caused by an increased contribution of 13 C-depleted methanotroph-derived diploptene to the total bacterial diploptene pool. The trend of more depleted δ 13 C values with increasing methanotrophic activity is, however, not supported by diploptene concentrations, which did not change significantly with temperature ( Figure 2a). Despite these considerations, an alternative explanation should be considered. It may well be that substrate availability influenced the extent of fractionation and, thereby, diploptene δ 13 C values. Jahnke et al. (1999) showed that δ 13 C values of methanotroph lipids strongly depend on methane concentrations, with much stronger isotopic fractionation at higher methane concen- trations. An increase from 3% to 37% methane in the headspace of their culture experiments resulted in up to 22‰ more negative δ 13 C values of total methanotroph lipids. Therefore, the decrease in diploptene δ 13 C values with increasing methane availability ( Figure 3) can also be explained through enhanced expression of the enzymatic isotope effect associated with methanotrophy, which was suppressed by substrate or mass transport limitations at lower temperatures or lower methane concentrations (Hayes, 2001).
Unfortunately, we did not analyse the isotopic composition of methane and carbon dioxide during the mesocosm experiments, which could have further assisted in the interpretation of the results.
In addition, it would be recommended to perform a set of mesocosm experiments with Sphagnum moss where the methane concentration is varied, while the temperature is kept constant, to decouple the temperature effect from the methane concentration effect, similar to the experiment performed by Sherry et al. (2016) andOsborne et al. (2017).
Diploptene and hopane δ 13 C values have been used previously to trace methanotrophy in past peatlands. In a Carex-dominated Holocene peat deposit from the Tibetan Plateau, diploptene δ 13 C values as negative as −50‰ have been reported (Zheng et al., 2014). In a recent study by Huang et al. (2018) in a peatland in Central China, the δ 13 C values of 17β,21β(H)-norhopane ranged between −28‰ and −40‰ and the more negative values were interpreted to be caused by enhanced methanotrophy. In both studies, the more negative hopanoid δ 13 C values corresponded to dryer conditions, which were explained to be the result of aeration of deeper peat deposits below the rooted zone, enhancing methanotrophy benefitting from an enhanced diffusive methane flux at the oxic-anoxic interface (Huang et al., 2018;Zheng et al., 2014). Extending the geological record, Inglis et al. (2015) observed no change in hopane δ 13 C values in a Paleogene peatland (Schӧningen, Germany), although there was a rise in the water  (Pancost et al., 2007). In the context of our data from experiments with extant peats, these depleted hopane δ 13 C values in the geological record suggest an intensified methane cycle during these key periods of Earth's past.
The environmental variability in peat bogs is much higher than

| Sphagnum bulk and compound-specific δ 13 C values as potential proxies for methanotrophy in ancient peat
No significant variation was observed in the δ 13 C values of bulk Sphagnum moss and the C 23 n-alkane, the most abundant n-alkane in S. cuspidatum and other Sphagnum sp. Nott et al., 2000), with incubation temperature ( Figure 2c) and, hence, with changes in methane production. This lack of variation in Sphagnum δ 13 C values suggests that Sphagnum does not take up more methane-derived CO 2 relative to CO 2 from other sources at increased levels of methane production and anticipated methanotrophy. This is somewhat surprising since it is thought that Sphagnum also uses the CO 2 produced by methanotrophs for photosynthesis Raghoebarsing et al., 2005). Increased levels of methanotrophy would thus expected to be reflected in a more negative 13 C signature of both specific lipids produced by Sphagnum or even in its bulk value.
Increasing temperatures could have stimulated organic matter degradation; hence, the increased methane flux at higher temperatures may be offset by a corresponding increase in the CO 2 flux from enhanced organic matter degradation. The isotopic composition of the combined recycled carbon source for Sphagnum would thereby remain the same. In addition, the isotopic fractionation between CO 2 and methane during methanogenesis may be balanced, with methane becoming more enriched in 13 C when more is produced, while CO 2 becomes less enriched in 13 C when less CO 2 is turned into methane. Such an inverse relationship between the carbon isotopic values of CO 2 and methane has been observed in peat bogs by Hornibrook et al. (2000). This way, changes in the isotopic composition of methane and CO 2 would cancel each other out. In any case, the lack of a clear temperature response of the stable carbon isotope values in both bulk Sphagnum and Sphagnum-derived biomarkers implies that reconstruction of methane oxidation and methane (re)cycling in peat bogs will have to rely on biomarkers specific for methanotrophs, such as δ 13 C values of hopenes and specific BHPs.

| CON CLUS IONS
A field pulse-chase experiment using 13 C-labelled methane clearly n-alkane did not show any significant variation with temperature.
The lack of any trend implies that reconstruction of methane oxidation and methane (re)cycling in peat bogs will have to rely on biomarkers specific for methanotrophs, such as δ 13 C values of hopenes and specific BHPs. Methanotroph-specific BHPs, aminotetrol and aminopentol, showed a non-linear response to temperature.
Aminotetrol was more abundant compared to aminopentol, which indicates that type II methanotrophs became prevalent during the experiment. Relative BHP concentrations may therefore provide insight into the presence of different types of methanotrophs in ancient peats. Finally, the δ 13 C values of diploptene, or its diagenetic products, potentially provide a useful tool to assess methanotrophic activity in past environments.

ACK N OWLED G M ENTS
The associate editor, David Naafs, and 3 anonymous reviewers are thanked for constructive comments on earlier versions of this

CO N FLI C T O F I NTE R E S T S
The authors declare that they have no conflicts of interests.