Chapter 1 Introduction

Natural wetlands, with more than half of their geographical area covered with peat-rich ecosystems, are likely to be the single largest source of atmospheric methane, a potent greenhouse gas. Northern peatlands probably contribute about one-third of the world's total wetland emissions. However, there is a considerable uncertainty in current methane emissions estimates from peatalnds. Accurate observation of methane exchange between the peat and the atmosphere is a fundamental basis for any efforts to improve our knowledge of the methane cycles in peatlands. At present, most of the flux studies are based on infrequent, temporally discontinuous ground-based measurements, assuming that the flux was stationary during the substantial non-sampling period. However, this assumption has not been explicitly verified, presumably because diffusion of dissolved methane or releases of methane through aquatic plants, which may have small temporal variability, are believed to be the main transport mechanisms to the atmosphere.

　In contrast to conventional wisdom that wetland soils are saturated below the water table and methane exists in a dissolved state, presence of biogenic gas bubbles originating from anoxic methane fermentation has recently been suggested. If the occurrence of methane-containing bubbles in peat is found to be pronounced, it is plausible that a considerable portion of methane emitted from peatland might be via release of bubbles, i.e. ebullition. As ebullition seems highly variable in space and time, widely-used flux measurement scheme may not be able to capture the ebullition events. Hence, there seems an urgent need to improve our understandings regarding the volume, composition, and distributions of the bubbles as well as the factors that may cause possible ebullition.

　The objectives of this study are to investigate occurrence of methane bubble in peat and its possible release to the atmosphere. This research was conducted at Bibai wetland (an ombrotrophic bog), located at Hokkaido, northern Japan (141°48'E, 43°19'N).

Chapter 2 In situ accumulation of methane bubbles in a natural wetland soil

Compared with numerous papers on measurements of methane emission from wetland surfaces, there are few reports on methane configuration and distribution within the soil profiles. By using a newly designed gas sampler, we succeeded in collecting free-phase gas from beneath the water table. The volumetric percentage of methane in the gas phase increased with depth and was generally more than 50% beneath the zone within which the water table fluctuates (Fig. 1). The volume of the gas phase beneath the water table was estimated to be from 0-19%. Using the volume ratio of the gas and liquid phases and methane concentration in the gas phase, as well as assuming that methane was in equilibrium between the two phases, we calculated that 〜60% of the methane down to 1 m accumulates in the form of bubbles. These results suggest the importance of ebullition in methane emission. Most importantly, our results show the need to consider gaseous-phase methane for understanding the production, transport and emission mechanisms of methane in peatlands, which has largely been overlooked to date.

Chapter 3 Ebullition of methane from peat with falling atmosheric pressure

Among various potential parameters, which may trigger methane ebullition, barometric pressure is possibly one of the most important factors because a drop in atmospheric pressure may lead to gas generation from solution and the enlargement of the volume of the gas phase. We have investigated the quantitative relationship between the amount of methane emitted via ebullition and changes in the atmospheric pressure through a laboratory experiment. During flux measurement periods, ebullitions were recorded almost exclusively in air-pressure-declining phases (Fig. 2). The increased volume of the gas bubbles due to reduction in atmospheric pressure and the amount of released gas bubbles revealed a strong linear relation, suggesting that in situ methane emissions via ebullition can be estimated using this correlation. Our results clearly showed that atmospheric pressure can be one of the most important factors to control methane emissions from peatlands and that ebullition can be the main transport mechanism during the pressure-falling phase.

Chapter 4 Falling atmospheric pressure as a trigger for methane ebullition from peatland

　90-hour of field study was carried out in high-summer season to determine whether or not methane ebullition into the atmosphere from peat soil occurs in field condition, and if does, to identify factors that control it. We measured the air pressure, water table, and peat temperature as potential factors to control the ebullition. We found that the methane flux can change by two orders of magnitude within a matter of tens of minutes due to the release of free-phase methane and the contribution of the ebullition to the total methane flux during the measurements was significant (50-64%) (Fig. 3). Episodic ebullitions were always associated to the reductions in air pressure, indicating that increased buoyancy forces due to enlargement of the trapped gas caused the upward migration of the bubbles (Fig. 4). These results clearly revealed that field campaigns must be designed to cover this rapid temporal variability caused by ebullition, which may be especially important in intemperate weather. Process-based methane emission models should also be modified to include air pressure as a key factor for the control of ebullient methane release from peatland.

Chapter 5 Episodic release of methane bubbles from peatlnd during spring thaw

In most northern peatlands, it is clear that drastic changes in physical environments such as melting of snow and near-surface frozen peat take place during the spring thaw in relatively short periods of time, invoking a sudden change in methane emission rates. We have conducted 165 hours of intensive flux measurements and found a large methane flush at the very moment the surface ice cover thawed (Fig. 5). Bubbles were found to be trapped in the ice layer (Fig. 6) and very high concentration of methane (〜20%) was detected in them. The abundance of the bubble-form methane was likely to be sufficient to explain the observed episodic release during the thaw. Omission of the episodic release of stored methane during the spring thaw results in underestimation of annual methane emissions as well as misunderstanding of seasonal methane dynamics. The results also imply the gas-phase methane may play an important role also in cold season methane dynamics in northern peatlands.

Conclusions

The main conclusions of the study are:

1. In situ volumetric gas profiles and methane concentrations in the gas phase beneath the water table level were quantified for the first time, showing the occurrence of methane bubbles in waterlogged peat soil.

2. Approximately 60-70% of methane stored in water-logged peat from the surface to a depth of 1 m was found to exist as gas-phase gas (30-40% were in dissolved state).

3. The very frequent sampling regime adopted in the field study provided a clear evidence that ebullition represents an important mechanism of methane emission from peatland and that it occurs as episodic events.

4. Theoretical calculations followed by numerical computations confirmed our hypothesis that fluctuations in the atmospheric pressure play a dominant role in determining the timing and magnitude of the ebullition events.

5. Field campaigns must be designed to cover the rapid temporal variability caused by ebullition, which may be especially important under intemperate weather.

6. Because existing methane flux data may not capture ebullition events, widely-accepted process-based models might have been tested against erroneous data. Also, our findings may reveal the need to revise the model itself, i.e., air pressure should be included as a key factor for the control of methane release via ebullition.

7. Large methane emissions associated with melting of the surface ice layer occurred as a result of release of entrapped bubbles found in the ice layer, suggesting that the gas-phase methane is play an important role not only during the growing season but also in the cold season methane dynamics in northern peatlands.

Fig.1 Vertical profile of methane

concentration at Sphagnum-dominated site. Methane is expressed as a percentage of gas sampled from the peat beneath the water table.■,23 August 2002; ●,18 October 2002; ▲,26 October 2003.

Fig.2 Cumulative methane flux and change in atmospheric pressure starting at days 366(a), 378(b), 399(c), and 408(d) of the experiment. The timing of the flux measurement by a closed chamber is indicated by open squares.

Fig.3 Time series of methane flux and atmospheric pressure at two plots A (a) and B (b) (31.2 m apart). Both plots were located in Sphagnum-dominated site. The timing of the flux measurement is indicated by open squares. The vertical arrows indicate episodic fluxes, which are significantly greater (P<0.05, by one-tailed t-test) than the other emission rates. The difference in the scale of methane flux between a and b is noteworthy.

Fig.4 Effect of changing atmospheric pressure, peat temperature, water table level, and all the variables on the volumetric gas content from the water table level to a depth of 80 cm at plot A (a) and plot B (b). The timing of episodic emissions is indicated by vertical arrows.

Fig.5 Methane flux from either snow cover, standing water or peat surface from 13 April to 20 April, 2006 at plots　A and B. Symbols indicate the timing of the chamber measurements.

Fig.6 A picture of bubbles stored in the ice between the snow and the peat layers. The diameters of the bubbles ranged from several mm to 1 cm.

　本論文は、CO2に次ぐ温室効果ガスとして近年特に注目されている大気中のメタンガスについて、その最大の発生源である湿原(自然湿地)からの放出量を正確に測定し、もってメタンガス発生メカニズムを明らかにすることを目的としている。

　第1章では、研究の背景、位置づけ、重要性を、従来の研究を総括しつつ明らかにした。すなわち、メタンの大気中濃度は1750年の700ppbから2004年の1783ppbへと、最近250年で約2.5倍に増加したこと、しかも2005年には、大気中メタンの放射強制力の評価が従来の値より85.4%も多かったとの報告が出されるなど、地球温暖化との関わりがますます重視されていることを指摘した。しかし、それにも関わらず、湿原からのメタン放出量の推定精度はこれまでのところ著しく低く、水中の溶存メタンの拡散による放出、植物体を通る放出、地下水面下の気泡の上昇と噴出、というメタン放出の全体像については、ほとんど明らかにされていないことを指摘した。

　第2章では、北海道美唄湿原を対象とした現地観測による、湿原の地下水中に気泡としてのメタンが存在しているかどうかの確認について述べた。その結果、ミズゴケを主体とする美唄湿原中では、地下水面下に平均10%程度の気泡が存在すること、その気泡中ではおよそ60%がメタンガスで占められていることを突き止めた。さらに、ガス状メタン量と水中溶存メタン量の存在比率をヘンリーの法則を適用して推測し、ガス態が58〜72%、溶存態が28〜42%であると結論した。すなわち、湿原の地下水中の気泡メタン量と水中溶存メタン量との関係を世界で初めて定量的に確かめたものであり、この結果は国際誌、国際学会(学会賞受賞)で公表された。

　第3章では、前述の美唄湿原から採取した、直径20cm、高さ60cmの不撹乱ミズゴケ泥炭試料を、室温20℃、光条件12時間の明暗周期で制御した実験室内に静置し、約100日経過した後に、この不撹乱試料から発生するメタンガス量測定について述べた。試料は透明アクリル円筒内に隙間無く不撹乱で充填されており、地下水を与えて嫌気的な条件を維持した。この後約1年間、大気圧とメタンガス発生量との関係を継続的に測定した結果、大きな低気圧が通過することに、試料から大量のメタンガスが噴出することを発見した。この発生量は、試料中の気泡状メタンがシャルル法則に従って体積膨張した量と良く関連付けすることができ、このことによって、大気圧の低下に伴う湿原地下水面下からのメタンバブルの噴出現象を捉えたと結論した。この結果も世界的に最先端の知見であり、国際誌、国際学会(招待講演)にて公表された。

　第4章では、美唄湿原現地においても大気圧の低下がメタン噴出を起こすかどうかの測定結果を述べた。現地において2点の観測地点を選び、1.5時間おきに連続90時間のメタンフラックス測定を行った。この間、幸運にも、大気圧が1016hPaから1000hPaまで急激に低下する局面に遭遇し、その間のメタンフラックスを経時的に測定した結果、気泡状のメタンが、急激な気圧低下に応じて表面から噴出していることを捉えた。そこで、ガス状メタンと溶解状メタンの状態方程式をそれぞれ与え、総量保存則を適用した連立方程式をフリーソフトであるScilabによって数値計算したところ、地下水位変動や大気の温度変化の影響より、大気圧低下とその結果としてガスの体積膨張が、メタンガス噴出の引き金になっていたことを理論的に証明できた。この結果も国際学会において公表され、国際誌において印刷中となっている。

　第5章では、美唄湿原における融雪期のメタンガス放出について、現地測定を行った結果を述べた。すなわち、嫌気的な条件下におかれた湿原の地下水面下では、温度の低い冬季にもメタン生成が続いているとの報告があることを根拠とし、冬季発生のメタンが凍結氷中に封入され融解期に突然噴出する可能性があると考え、このことを160時間連続現地測定によって捉えた。その結果2006年4月17日に生じた表面氷の融解と同時に、氷中にトラップされていたメタンバブルが突発的に噴出することが分かった。氷中のメタンガス量は、132mg m(-2)と見積もられ、融解によるメタンガス放出量は2地点の平均値127.5mg m(-2)であり、よく一致した。この内容は、国際学会での発表を準備中で、国際誌へは投稿中である。

　第6章では、結論として、(1)湿原の地下水面下にはメタンを主成分とする気泡が存在すること、(2)湿原土壌中のメタンの6〜7割は気泡中にガス態として存在すること、(3)気泡の噴出は主要なメタン放出メカニズムの一つで大気圧の低下時に突発的なイベントとして生じること、(4)気圧が低下する悪天候時の観測が重要であること、(5)冬季には氷中にメタンバブルがトラップされ、融雪期の大きな放出の原因となること、などを述べた。

　以上要するに、本研究は、湿原の地下水中にガス状メタンが多量に存在し手いることを発見し、大気圧の低下と冬季の融雪とが、これらのガス状メタンを突発的に大気に噴出させていることを、室内実験及び現地観測に基づいて科学的に証明したものであり、学術上、応用上貢献するところが著しく大きい。よって、審査委員一同は、本論文が博士(農学)の学位論文として価値あるものと認めた。