Carbon Isotopic Evidence for Gas Hydrate Release and Its Significance on Seasonal Wetland Methane Emission in the Muli Permafrost of the Qinghai-Tibet Plateau

In order to determine the significant role of gas hydrate in seasonal wetland methane emission at the drilling-affected permafrost, the carbon isotopic monthly field monitoring of methane (CH4), as well as carbon dioxide (CO2), emitted from near-surface soil and a gas hydrate drilling well (DK-8) was conducted in the Muli permafrost of the Qinghai-Tibet Plateau. The methane source effused from the well DK-8 was calculated as −25.9 ± 1.4‰ and −26.5 ± 0.5‰, respectively, by the Keeling and Miller Tans plots, with the carbon isotope fractionation (εC) between CO2 and CH4 from −25.3‰ to −32.1‰. The carbon isotopic signatures are indicative of thermogenic origin associated with gas hydrate dissociation. The near-surface soil-emitted methane has δ13CCH4 values between −52.0 ± 1.2‰ and −43.2 ± 1.8‰ with the heaviest in December and the lightest in July. Further, the εC values of near-surface soil-emitted gases were between 28.6‰ and 47.9‰, significantly correlated with the δ13CCH4 values. The linear correlation between εC and δ13CCH4 values indicated binary end-member of microbial and thermogenic sources control the seasonal variation of wetland methane emission. The thermogenically derived methane was identified as the dominant methane source in autumn and winter, compared with the increasing contribution of microbially derived methane in spring and summer. The finding provides reliable evidence for gas hydrate release on the seasonal wetland methane emission in the Muli permafrost affected by drilling activities. The combined application of εC and δ13CCH4 to distinguish thermogenic from biogenic methane is well established and powerful in complex environments, which can provide an improved constraint on source apportionment for wetland emitted methane in the permafrost of the Qinghai-Tibet Plateau.


Introduction
Methane (CH 4 ) is an important greenhouse gas that provides the second-largest contribution to historical global warming, with stronger climate warming potential than carbon dioxide (CO 2 ) [1]. Global atmospheric methane concentrations have increased rapidly since 2007 at a renewed growth rate of 6.9 ± 2.7 ppb yr −1 [2]. Wetland methane emissions, the largest natural source in the global CH 4 budget, play an emerging role in driving 21st-century climate change [3]. The carbon pools in permafrost regions are large reservoirs susceptible to climate change and can drive positive feedback to climate warming [4][5][6][7][8]. Methane emission in the thawing permafrost is key to the greenhouse gas budget on climate-relevant timescales [9]. Therefore, the wetland methane emission process strongly associated with wetland carbon cycling in permafrost regions is of great significance to understanding climate feedback and mitigating global warming.

Study Area
The study area is located in the Muli permafrost of the Qilian Mountains in the northeastern margin of the Qinghai-Tibet Plateau (Figure 1a,b), with an altitude of approximately 4000-4300 m and an annual average temperature of −5.1 • C. The wetland ecosystem is dominated by alpine meadow, which is influenced by permafrost freezing and thawing driven by seasonal temperature variation. The alpine-typed permafrost covers an area of 1 × 10 5 km 2 in the Qilian Mountains, with a thickness regionally ranging from 60 to 95 m and an active layer of 1.5-2.0 m [34]. determine their stable carbon isotopic signatures. This study is aimed to provide new evidence for gas hydrate release as a significant source of wetland methane emission in the Muli permafrost, to reveal the effect of gas hydrate release on wetland methane seasonal emission, and further, to imply the influence of anthropogenic activities, such as drilling, on wetland methane emission in the permafrost of the Qinghai-Tibet Plateau.

Study Area
The study area is located in the Muli permafrost of the Qilian Mountains in the northeastern margin of the Qinghai-Tibet Plateau (Figure 1a,b), with an altitude of approximately 4000-4300 m and an annual average temperature of −5.1 °C. The wetland ecosystem is dominated by alpine meadow, which is influenced by permafrost freezing and thawing driven by seasonal temperature variation. The alpine-typed permafrost covers an area of 1 × 10 5 km 2 in the Qilian Mountains, with a thickness regionally ranging from 60 to 95 m and an active layer of 1.5-2.0 m [34]. The study area is tectonically situated in the western Middle Qilian block formed during the Caledonian Movement (513-386 Ma), adjacent to the South Qilian structural zone [35], and it is also situated in the Muli depression of the south Qilian basin. The exposed strata are Quaternary, Middle Jurassic, and overturned Triassic, which thrust onto the Middle Jurassic. Large NW-SE trending thrust faults developed in the study area ( Figure 1c). These thrusts or fractures may serve as paths for gas to migrate upwards from the deep layers, and some of the fractures may serve as temporary gas hydrate-bearing reservoirs [36]. The study area is tectonically situated in the western Middle Qilian block formed during the Caledonian Movement (513-386 Ma), adjacent to the South Qilian structural zone [35], and it is also situated in the Muli depression of the south Qilian basin. The exposed strata are Quaternary, Middle Jurassic, and overturned Triassic, which thrust onto the Middle Jurassic. Large NW-SE trending thrust faults developed in the study area ( Figure 1c). These thrusts or fractures may serve as paths for gas to migrate upwards from the deep layers, and some of the fractures may serve as temporary gas hydrate-bearing reservoirs [36].
Gas hydrate exploration and scientific drilling work have been performed in permafrost regions in the Qilian Mountains since 2000 [37,38]. A large amount of gas was released when the gas hydrate-bearing cores were extracted under air-tight conditions. Raman spectroscopy identified CH 4 as the primary hydrate-forming gas, with secondary components of C 2 H 6 , C 3 H 8 , and CO 2 [14]. The geological and geochemical field exploration work indicated the existence of natural gas hydrate in the target area [36][37][38][39]. The gas hydrate is a structure II gas hydrate, mainly stored in siltstone and mudstone of the Middle Jurassic Jiangcang Formation, and secondly, sandstone is unstable and discontinuous in distribution. The occurrence of gas hydrates is closely related to faults, and the burial depth is generally 133-396 m [40].

Field Monitoring Method
The monthly dynamic monitoring of methane in the free gas from the hydrate gas drilling well DK-8 and near-surface soil was conducted simultaneously in situ from January to December in 2017 in the study area. The monitoring frequency is once or twice a month.
The DK-8 well with a diameter of 16.8 cm was drilled to a depth of approximately 400 m, with the presence of gas hydrate and its associated anomalies mainly within intervals of 150-300 mbs (meter below surface), shown in Figure 2a. The core lithology is mainly composed of sandstone, siltstone, mudstone, and oily shale. The base of the permafrost is around 110 m, revealed by drilling, coexisting with underground ice around 70 mbs. The DK-8 was twice tested for exploring gas hydrates in 2011 and 2016, respectively. During the monitoring period, the mouth of the DK-8 was sealed, and the headspace gas can be only derived from the gas emission of the 150-300 mbs layers.

Measurements and Data Processing
The sampling ports from the well DK-8 and the soil chamber were connected, respectively, to different gas inlets of a portable Picarro isotope analyzer (G2201-i; Cavity Ring-Down Spectroscopy, Picarro, USA). On all sampling occasions, the contents (ppm) and carbon isotope composition (δ 13 C, ‰ vs. PDB) of CH4 together with CO2 were simultaneously measured in situ using the Picarro G2201-i isotope analyzer. The Picarro G2201-i was checked regularly against working standards, and adjustments were made before each field monitoring. The working standards used were a CH4 standard gas of 2 ppm and 10 ppm with the δ 13 C gradients of 0‰, −20‰, −40‰, and −65‰, and used CO2 standard gas of 300 ppm and 600 ppm with the δ 13 C gradient of 0‰, −10‰, −20‰, and −30‰, respectively. These standard gases were calibrated by the National Institute of Metrology, China. The free gas from the near-surface soil was monitored using a detachable cylindrical soil respiration chamber that was 14.5 cm in diameter and 53.5 cm in height (Figure 2b). The cylindrical chamber was embedded in a pre-excavated hole with a depth of 30 cm near the well DK-8. After sealing the chamber in the hole, the whole system was circulated with the surrounding soil only through the several sieve holes in the chamber wall 10-30 cm blew the ground. Disturbance from the mixing of air and soil gas through the soil surface was assumed to be negligible because of the fine texture and compact nature of the soil and gas-tight measurement chamber.

Measurements and Data Processing
The sampling ports from the well DK-8 and the soil chamber were connected, respectively, to different gas inlets of a portable Picarro isotope analyzer (G2201-i; Cavity Ring-Down Spectroscopy, Picarro, USA). On all sampling occasions, the contents (ppm) and carbon isotope composition (δ 13 C, ‰ vs. PDB) of CH 4 together with CO 2 were simultaneously measured in situ using the Picarro G2201-i isotope analyzer. The Picarro G2201-i was checked regularly against working standards, and adjustments were made before each field monitoring. The working standards used were a CH 4 standard gas of 2 ppm and 10 ppm with the δ 13 C gradients of 0‰, −20‰, −40‰, and −65‰, and used CO 2 standard gas of 300 ppm and 600 ppm with the δ 13 C gradient of 0‰, −10‰, −20‰, and −30‰, respectively. These standard gases were calibrated by the National Institute of Metrology, China.
The readings were collected under a high-accuracy CH 4 -CO 2 compound mode at intervals of 5 min for a total duration of at least 4 h per monitoring. The performance drift of the instrument was checked during the continuous determination for 24 h. The content precision (5 min, 1σ) is 5 ppb (+0.05% of reading) for CH 4 and 200 ppb (+0.05% of reading) for CO 2 , with the guaranteed δ 13 C precision (5 min, 1σ) <1.0‰ for CH 4 and <0.2‰ for CO 2. Moreover, the instrument itself has a temperature control function with an accuracy of 0.5%. The measurement can only be started when the temperature of the instrument chamber reaches 45°C, thus avoiding the interference of ambient temperature change.
The two gas inlets were alternately switched and measured in succession for each sampling and measurement. To avoid physical fluctuations caused by switching gas inlets, only the measured data during the intermediate 3-min stationary period were taken as valid data to calculate. All valid readings of the total duration per monitoring are calculated as mean ± standard error and presented as the final result for each measurement. The calculated measurements of CH 4 and CO 2 content, δ 13 C CH4 and δ 13 C CO2 , as well as the calculated carbon isotope fractionation ε C between CO 2 and CH 4 (ε C = δ 13 C CO2 − δ 13 C CH4 ) for the free gas of the DK-8 hydrate gas drilling well and the active soil layer are shown in Table 1. Table 1. Methane content and carbon isotopic compositions of methane and carbon dioxide in the gas effused from the DK8 hydrate gas drilling well and in the free gas of the soil active layer.

Monthly Temperature Variation of the Soil upper Active Layer
The monthly temperature variation of the soil's upper active layer, as well as the low-level air, at the gas hydrate drilling area in the Muli permafrost is shown in Figure 3, based on the average daily monitoring data in 2017. The soil temperature varied with the atmospheric temperature in a range of −16-16 • C, lower than that of the atmospheric temperature (−25-18 • C). The soil temperature was observed above 0 • C first in early April and was always greater than 0 • C between May and October, which is the main period of permafrost thawing. The frozen period characterized by the soil temperature lower than 0 • C was from November to next April.

Monthly Temperature Variation of the Soil upper Active Layer
The monthly temperature variation of the soil's upper active layer, as well as the lowlevel air, at the gas hydrate drilling area in the Muli permafrost is shown in Figure 3, based on the average daily monitoring data in 2017. The soil temperature varied with the atmospheric temperature in a range of −16-16 °C, lower than that of the atmospheric temperature (−25-18 °C). The soil temperature was observed above 0 °C first in early April and was always greater than 0 °C between May and October, which is the main period of permafrost thawing. The frozen period characterized by the soil temperature lower than 0 °C was from November to next April.

Methane Content and Carbon Isotopic Compositions Derived from the Drilling Well DK-8
The monthly variation in content and carbon isotopic composition of methane derived from the hydrate drilling well DK-8 during the period from January to August in 2017 is shown in Figure 4. The measured methane contents showed a large variation from 1.904 ± 0.001 to 8.530 ± 1.983 ppm. The methane was constant at a low concentration of

Methane Content and Carbon Isotopic Compositions Derived from the Drilling Well DK-8
The monthly variation in content and carbon isotopic composition of methane derived from the hydrate drilling well DK-8 during the period from January to August in 2017 is shown in Figure 4. The measured methane contents showed a large variation from 1.904 ± 0.001 to 8.530 ± 1.983 ppm. The methane was constant at a low concentration of 1.967 ± 0.130 ppm from January to March and of 2.015 ± 0.108 ppm from June to August, whereas it significantly increased to 2.920 ppm in early April and soared to 8.530 ppm in early May and 4.201 ppm in late May. The measured δ 13 C CH4 values varied between −49.6‰ and −34.6‰, parallel with methane contents (Figure 4). Overall, the higher CH 4 content had less negative δ 13 C CH4 , while lower CH 4 content had more negative δ 13 C CH4 . Specifically, for the higher content of CH 4   . Content and stable carbon isotopic composition (δ 13 CCH4) of methane and carbon isotope separation (εC) between CH4 and CO2 effused from the gas hydrate drilling well DK-8.

Methane Content and Carbon Isotopic Compositions of Free Gas from the Upper of Soil Active Layer
The CH4 content of the soil chamber varied from 1.906 ± 0.003 to 2.390 ± 0.303 ppm in 2017, with an average of 1.973 ± 0.113 ppm. Their δ 13 CCH4 values, representing that of the upper soil active layer, ranged between −52.0 ± 1.2‰ and −43.2 ± 1.8‰ with an average of −48.4 ± 2.7‰. The δ 13 CCH4 values measured in March-August (−50.2 ± 1.1‰) were lower than those measured in October-February (−46.7 ± 2.1‰). Particularly, the CH4 with soaring content (2.390 ± 0.303 ppm) in early April had significantly higher δ 13 CCH4 values (−43.2 ± 1.8‰) than that in other months (Figure 5), suggesting a significant increase of the methane source that has more positive δ 13 CCH4. The carbon isotope fractionation (εC) between CO2 and CH4 ranged between 28.6‰ and 47.9‰, with the highest in July and the lowest in December. The significant increasing and decreasing trend of the εC values were observed in the periods from April to July and from July to December, respectively. The εC values were negatively correlated with the δ 13 CCH4 values with a Pearson's r of −0.9115. . Content and stable carbon isotopic composition (δ 13 C CH4 ) of methane and carbon isotope separation (ε C ) between CH 4 and CO 2 effused from the gas hydrate drilling well DK-8.

Methane Content and Carbon Isotopic Compositions of Free Gas from the Upper of Soil Active Layer
The CH 4 content of the soil chamber varied from 1.906 ± 0.003 to 2.390 ± 0.303 ppm in 2017, with an average of 1.973 ± 0.113 ppm. Their δ 13 C CH4 values, representing that of the upper soil active layer, ranged between −52.0 ± 1.2‰ and −43.2 ± 1.8‰ with an average of −48.4 ± 2.7‰. The δ 13 C CH4 values measured in March-August (−50.2 ± 1.1‰) were lower than those measured in October-February (−46.7 ± 2.1‰). Particularly, the CH 4 with soaring content (2.390 ± 0.303 ppm) in early April had significantly higher δ 13 C CH4 values (−43.2 ± 1.8‰) than that in other months (Figure 5), suggesting a significant increase of the methane source that has more positive δ 13 C CH4 . The carbon isotope fractionation (ε C ) between CO 2 and CH 4 ranged between 28.6‰ and 47.9‰, with the highest in July and the lowest in December. The significant increasing and decreasing trend of the ε C values were observed in the periods from April to July and from July to December, respectively. The ε C values were negatively correlated with the δ 13 C CH4 values with a Pearson's r of −0.9115.

Carbon Isotopic Signatures for Methane Derived from Gas Hydrate Dissociation
We used Keeling and Miller Tans plots [41][42][43] to explore the potential source isotopic signature of fresh CH4 and CO2 escaping from the drilling well DK-8. The source isotopic signature for CH4 (δs) was calculated as the intercept value (−25.9 ± 1.4‰) of the regression associated with the Keeling plot ( Figure 6a) and as the slop value (−26.5 ± 0.5‰) of the regression associated with the Miller Tans plot (Figure 6b), respectively. There was no clear distinction in identified δs values between the two methods. The less negative δ 13 CCH4 values are typical of thermogenically-derived methane sources enriched in 13 C (−50‰ to −20‰, [32]). The isotopic signature of the CH4 source for the well DK-8 is also consistent with δ 13 CCH4 values (−17.7‰ to −34.7‰) of headspace gases from gas-hydratebearing DK-8 drill cores at depths of 184-291m [36]. Gases from the gas hydrate samples in the Muli permafrost were mainly thermogenic, either of coal-derived or oil-derived origin, discerned by the combined plots of δ 13 C1-C1/(C2 + C3) [30,40,[44][45][46], δ 13 C1-δ 13 C2-δ 13 C3 [47,48], and δ 13 C1-δD1 [44]. However, the Pearson's correlation coefficients between δ 13 CCO2 and 1/[CO2] and between δ 13 CCO2*[CO2] and CO2 content were low (r = 0.1768 and r = −0.5690, p > 0.05, respectively). Thus, the traditional "Keeling plot" or "Miller Tans plots" two end-member mixing model cannot estimate the fresh CO2 source. The CO2 contents (422.0 ± 37.7 ppm) were considerably greater than the CH4 contents (1.98 ± 0.11 ppm), suggesting the importance of some other mechanisms of CO2 production. Therefore, besides the hydrate-forming gas, the CO2 source is not specified in this study, which may come from the deep depth or from soil organic decomposition. Figure 5. Content and stable carbon isotopic composition of methane (δ 13 C CH4 ) and carbon isotope fractionation (ε C ) between CO 2 and CH 4 in the free gas of the soil active layer.

Carbon Isotopic Signatures for Methane Derived from Gas Hydrate Dissociation
We used Keeling and Miller Tans plots [41][42][43] to explore the potential source isotopic signature of fresh CH 4 and CO 2 escaping from the drilling well DK-8. The source isotopic signature for CH 4 (δ s ) was calculated as the intercept value (−25.9 ± 1.4‰) of the regression associated with the Keeling plot (Figure 6a) and as the slop value (−26.5 ± 0.5‰) of the regression associated with the Miller Tans plot (Figure 6b), respectively. There was no clear distinction in identified δ s values between the two methods. The less negative δ 13 C CH4 values are typical of thermogenically-derived methane sources enriched in 13 C (−50‰ to −20‰, [32]). The isotopic signature of the CH 4 source for the well DK-8 is also consistent with δ 13 C CH4 values (−17.7‰ to −34.7‰) of headspace gases from gashydrate-bearing DK-8 drill cores at depths of 184-291 m [36]. Gases from the gas hydrate samples in the Muli permafrost were mainly thermogenic, either of coal-derived or oilderived origin, discerned by the combined plots of δ 13 C 1 -C 1 /(C 2 + C 3 ) [30,40,[44][45][46], δ 13 C 1δ 13 C 2 -δ 13 C 3 [47,48], and δ 13 C 1 -δD 1 [44]. However, the Pearson's correlation coefficients between δ 13 C CO2 and 1/[CO 2 ] and between δ 13 C CO2 *[CO 2 ] and CO 2 content were low (r = 0.1768 and r = −0.5690, p > 0.05, respectively). Thus, the traditional "Keeling plot" or "Miller Tans plots" two end-member mixing model cannot estimate the fresh CO 2 source. The CO 2 contents (422.0 ± 37.7 ppm) were considerably greater than the CH 4 contents (1.98 ± 0.11 ppm), suggesting the importance of some other mechanisms of CO 2 production. Therefore, besides the hydrate-forming gas, the CO 2 source is not specified in this study, which may come from the deep depth or from soil organic decomposition.
The increasing CH 4 content of well DK-8 in April-May was observed with less negative δ 13 C CH4 values than −40‰ (Figure 4), indicating a substantial proportion of releasing CH 4 from gas hydrate dissociation. The CH 4 from gas hydrate dissociation in the permafrost layer at a depth of 150−305 m can diffuse into well DK-8 and be trapped and accumulated in the ice of the well during upward diffusion when the soil temperature is lower than 0 • C. The occurring of a considerable increase of CH 4 content of the well DK-8 coincides with the initial melting of the ice of the well in early April when the soil temperature first raises above 0 • C. The peak of the CH 4 content also coincides with the complete melting process in early May when the soil temperature is always above 0 • C. The increasing CH4 content of well DK-8 in April-May was observed with less negative δ 13 CCH4 values than −40‰ (Figure 4), indicating a substantial proportion of releasing CH4 from gas hydrate dissociation. The CH4 from gas hydrate dissociation in the permafrost layer at a depth of 150−305 m can diffuse into well DK-8 and be trapped and accumulated in the ice of the well during upward diffusion when the soil temperature is lower than 0 °C. The occurring of a considerable increase of CH4 content of the well DK-8 coincides with the initial melting of the ice of the well in early April when the soil temperature first raises above 0 °C. The peak of the CH4 content also coincides with the complete melting process in early May when the soil temperature is always above 0 °C.
The carbon isotope fractionation (εC) between CO2 and CH4 can be indicative of the particular methanogenic pathway [32]. The εC values of gas hydrates in the Muli permafrost are calculated based on the reported δ 13 C data of gas components (CH4 and CO2) of the drilled gas hydrate samples summarized by Dai et al [48]. The εC values are characterized by 21.5-35.8‰, with an average of 29.8‰ (Figure 7), associated with the average δ 13 CCH4 and δ 13 CCO2 of −45.3 ± 5.8‰ and −13.2 ± 7.3‰. The calculated εC values of gases from well DK-8 that correspond to more CH4 content and higher δ 13 CCH4 were 25.3-32.1‰, which is consistent with the thermogenic gas hydrate dissociation. This consistency shows no to little carbon isotopic fractionation during upward transport through the drilling wells. Moreover, the εC values of gas hydrates are distinctly lower than those associated with microbial methanogenesis predominantly by carbonate reduction (εC from 49‰ to over 100‰, [32]) or by fermentation of methylated substrates (εC typically between 40‰ and 55‰, [32]) and distinctly higher than those associated with microbial methane oxidation (εC generally less than 10‰, [32]). Therefore, the εC of 20-36‰ can be a carbon isotopic signature of thermogenic gas hydrate. The lower εC values associated with the δ 13 CCH4 higher than −40‰ further suggest considerable contribution from gas hydrate in April-May. The higher εC values (44.4-46.3‰) associated with lower δ 13 CCH4 (~−49‰) indicate an increasing contribution of a microbial source in June-July due to the elevated temperatures. The carbon isotope fractionation (ε C ) between CO 2 and CH 4 can be indicative of the particular methanogenic pathway [32]. The ε C values of gas hydrates in the Muli permafrost are calculated based on the reported δ 13 C data of gas components (CH 4 and CO 2 ) of the drilled gas hydrate samples summarized by Dai et al. [48]. The ε C values are characterized by 21.5-35.8‰, with an average of 29.8‰ (Figure 7), associated with the average δ 13 C CH4 and δ 13 C CO2 of −45.3 ± 5.8‰ and −13.2 ± 7.3‰. The calculated ε C values of gases from well DK-8 that correspond to more CH 4 content and higher δ 13 C CH4 were 25.3-32.1‰, which is consistent with the thermogenic gas hydrate dissociation. This consistency shows no to little carbon isotopic fractionation during upward transport through the drilling wells. Moreover, the ε C values of gas hydrates are distinctly lower than those associated with microbial methanogenesis predominantly by carbonate reduction (ε C from 49‰ to over 100‰, [32]) or by fermentation of methylated substrates (ε C typically between 40‰ and 55‰, [32]) and distinctly higher than those associated with microbial methane oxidation (ε C generally less than 10‰, [32]). Therefore, the ε C of 20-36‰ can be a carbon isotopic signature of thermogenic gas hydrate. The lower ε C values associated with the δ 13 C CH4 higher than −40‰ further suggest considerable contribution from gas hydrate in April-May. The higher ε C values (44.4-46.3‰) associated with lower δ 13 C CH4 (~−49‰) indicate an increasing contribution of a microbial source in June-July due to the elevated temperatures.

Sources of Seasonal Wetland Methane Emission Affected by Gas Hydrate Drilling
The flux and area-weighted mean global source signature from wetlands is −62‰ [49], which is similar to the mean microbial signature of 61.5 ± 0.6‰ [50]. The source sig-

Sources of Seasonal Wetland Methane Emission Affected by Gas Hydrate Drilling
The flux and area-weighted mean global source signature from wetlands is −62‰ [49], which is similar to the mean microbial signature of 61.5 ± 0.6‰ [50]. The source signature map resolves a latitudinal difference of~10‰ between northern high-latitude (mean 67.8‰) and tropical (mean 56.7‰) wetlands [49]. However, the annual average of δ 13 C CH4 in the study area is −48.4 ± 2.7‰, significantly more enriched than the global source isotopic signature. The soil-emitted δ 13 C CH4 in summer (−51.3 ± 0.9‰) of the study area is similar to that in two bogs (−53.0 ± 5.5‰ and −55.3 ± 5.5‰) separated by >1000 km over the Qinghai-Tibetan Plateau [51], but distinctly different from that of the northern European wetlands (−71 ± 1‰) [52]. In particular, more enriched δ 13 C CH4 of the soilemitted methane was observed in December than in July-August, reflecting highly variable seasonal wetland methane sources. Compared with wetlands away from gas hydrate in the Qinghai-Tibetan Plateau, there was no significant difference of soil-emitted δ 13 C CH4 between summer (from −60.6‰ to −54.6‰) and winter (−65.5‰ to −55.8‰) [53].
The gases of thermogenic gas hydrates samples in the Muli permafrost of the Qilian Mountains had distinctive δ 13 C CH4 values heavier than −50‰ and ε C values lower at 20-36‰ from the microbially-derived methane with δ 13 C CH4 lighter than −50‰ and ε C values more than 40‰. Therefore, the isotope signatures of δ 13 C CH4 combined with ε C can be adequate to reliably distinguish microbially or thermogenically derived methane. The two dominated microbial and thermogenic sources controlling seasonal variation of the wetland methane emission in the study area are indicated by linear correlation of binary mixing in the diagram of ε C versus δ 13 C CH4 (Figure 8). The samples collected in spring and summer plotted in the microbial zone indicated the soil-emitted CH 4 being dominated by fermentation of methylated substrates (δ 13 C CH4= −50‰ to −60‰, ε C = 40-60‰, [32]). Further, the samples in autumn and winter are plotted beyond the microbial zone and along the fitting line into the thermogenic zone, indicating an increasing contribution of thermogenic gas hydrates. A response of soil methane emission to the CH 4 effusion peak out of well DK-8 occurred on April 6 and can be observed by the CH 4 surge (2390 ppm), less negative δ 13 C CH4 (−43.2‰), and lower ε C (33.9‰) during the same period. The soil-emitted CH 4 on April 6 plotted in the thermogenic zone (Figure 8) suggests the accumulated thermogenic gas hydrate dissociation in the ice of drilling wells is an important source of soil-emitted methane. The combined effects of both microbially and thermogenically derived methane control the seasonal variation of wetland methane emission. This finding further suggests gas hydrate dissociation affected by drilling activities is an important source of methane emis- The combined effects of both microbially and thermogenically derived methane control the seasonal variation of wetland methane emission. This finding further suggests gas hydrate dissociation affected by drilling activities is an important source of methane emission from wetlands in the Muli permafrost. During spring and summer, microbes are the dominant source of soil-emitted methane, although the thermogenic methane emission from subterranean gas hydrates can be accelerated through new channels resulting from permafrost thawing. The increase in temperature accompanied by permafrost thawing promoted an exponential boost in the scale of microbial communities and activity [54][55][56] during the transition from spring to summer, which results in increasing methane produced by methanogenic microbes from May to July. However, with an increase in temperature to a certain extent, the upper limit of the environmental carrying capacity of methanogenic microbial communities is reached [57], which leads to a lower contribution of microbial methane in August than in July.
During the transition from late summer to autumn, microbial activities become weaker as the temperature decreases and the consequential reduction in methane production by methanogens. As a result, the methane contribution from the microbial source decreased, and that from the thermogenic source relatively increased in October. During winter, significant decreases occur on the scale of microbial communities and microbial activity [58], especially at temperatures below −10 • C. This leads to a substantial reduction in the amount of methane produced by methanogens. Thus, the contribution of the thermogenically derived methane relatively increased. The carbon isotope compositions of δ 13 C CH4 and ε C of wetland emission consistently reflected the dynamic variation of relative contributions of microbial and thermogenic methane during autumn and winter. In addition, as the average daily temperature falls below zero, freezing occurs in the permafrost, which leads to the decreased effusion of methane from subterranean gas hydrates. The dominance of thermogenically derived methane in December is shown in Figure 8, suggesting that the subterranean gas hydrates dissociation may seep through faults or drainage systems or drilled-through permafrost layers upward into the wetland.

Conclusions
The twelve-month systematic field monitoring of methane sources at a typical gas hydrate drilling site in the Muli permafrost of the Qilian Mountains provided carbon isotopic evidence that thermogenic gas hydrate dissociation is an important source of wetland methane emission affected by drilling activities. The combined application of carbon isotope fractionation (ε C ) between CH 4 and CO 2 and carbon isotopic composition of methane (δ 13 C CH4 ) to distinguish thermogenic from microbial methane is well established and appears to be useful in complex environments. The carbon isotopic signatures for methane derived from gas hydrate dissociation determined by the gas hydrate drilling DK-8 is distinctly characterized by heavier δ 13 C CH4 of −26.5 ± 0.5‰ and lower ε C of −25.3‰ to −32.1‰. The seasonal variation of methane sources of near-surface soil fluxes at the drilling site confirmed the significance of methane emission from the thermogenic gas hydrates on local permafrost wetland methane emission, especially in autumn and winter. The effect of gas hydrate dissociation on methane emission of the permafrost wetland may become more significant with the further exploration of gas hydrate resources. The effect of gas hydrate on regional wetland methane emission in the Muli permafrost requires more research to fully understand, which will allow for exploration of the mechanisms of soil methane emissions in the permafrost wetlands of the Qinghai-Tibet Plateau as the global climate warms.
Author Contributions: X.L., interpreted the data and wrote the paper and acquired funding; S.P., conceived and designed the research; J.X. performed data analysis and figure drawing; Y.Z., performed fieldwork supervision and funding acquisition; S.Z., R.X., and C.L. performed field investigation and monitoring. All authors have read and agreed to the published version of the manuscript. Data Availability Statement: The data generated or analyzed during this study were included in this manuscript.