A meta-analysis of methane emissions at the ecosystem level reveals a simple exponential dependence on temperature, despite the complex array of factors that control this process. See Letter p.488
Methane is the third-largest contributor to the greenhouse effect, after water vapour and carbon dioxide. The atmospheric concentration of methane rose for much of the twentieth century, was steady between 1999 and 2006, and is now again rising, at a rate of 0.4% annually1. The cause of this resumption is not fully understood, but it is probably related to a surge in methane emissions from wetlands — nearly half of global methane emissions comes from wetlands and rice paddies that are expected to be subject to temperature-related and other feedbacks from global climate change. Although a complex set of factors influences methane emission at the ecosystem level, Yvon-Durocher et al.2 report in a paper published on page 488 today that the average response of methane emissions to temperature across a range of ecosystems is well described by a simple mathematical relationship. Could it be as straightforward as that?
Most natural methane emissions originate with microorganisms called methanogens, the metabolic rates of which, during growth in culture with unlimited substrate, change with temperature according to the Arrhenius equation — a simple exponential dependence of rate constant on the reciprocal of absolute temperature. Such behaviour is unsurprising: although methanogenesis comprises a network of enzyme-catalysed reactions, the Arrhenius dependence reflects the kinetics of a single rate-limiting step. But Yvon-Durocher and colleagues' report of the same temperature dependence at the ecosystem level is notable, because an array of physical, chemical and ecological factors controls the production of methane and its release to the atmosphere3.
Methanogenesis in soils and sediments is ultimately fuelled by complex organic matter; what proportion of this organic matter is converted to methane, and at what rate, depends as much on ecosystem dynamics as on the enzyme kinetics of methanogens (Fig. 1). For example, organic carbon can be converted to carbon dioxide, rather than methane, when oxidants such as oxygen, nitrate, iron(III) and sulphate are available to fuel methanogens' microbial competitors. And when organic matter does get converted to methane, it must first be broken down by other microbes into the few simple substrates that methanogens can metabolize — an upstream supply that can limit the rate of methanogenesis.
Transport of methane from the site of production to the atmosphere is also subject to effects that might obscure a purely biochemical temperature dependence. Methane is consumed in both aerobic and anaerobic microbial processes, and transport through zones in which such consumption occurs can markedly reduce net emissions. But consumption can largely be bypassed when physical and metabolic factors combine to promote ebullition — the formation and release of methane bubbles, a sight not uncommon in productive freshwater lakes and swamps. The vascular system of plants can also facilitate methane transport, when roots penetrate the methane-producing portions of soils and provide a direct conduit to the atmosphere; but these same conduits also transport oxygen that inhibits methanogenesis and promotes methane consumption.
In their analysis of 127 studies of the ecosystem-level dependence of methane emission on temperature, Yvon-Durocher et al. acknowledge this complex array of factors, but conclude that the aggregate temperature response is nonetheless described by the Arrhenius equation, with an apparent activation energy (Ea) of 0.96 electronvolts, similar to the 1.10 eV observed in pure cultures of methanogens. Ea is a measure of temperature sensitivity; for example, 0.96 and 1.10 eV correspond, respectively, to a 3.5- and 4.2-fold increase in rate constant for an increase in temperature from 20 °C to 30 °C.
Statistically speaking, the large number of studies considered allows for a confident statement that the calculated mean Ea (0.96 eV) accurately reflects the mean temperature sensitivity of methane-emitting ecosystems — assuming that the sites that comprise the data set represent a random sample of all such environments. But the impact of factors other than temperature seems evident in the scatter and spread of the individual data sets considered by the authors. For example, about 40% of the studies considered had Arrhenius-plot correlation coefficients (r2) of less than 0.5, which indicates that less than half of the variance in those emission data is explained by the Arrhenius relationship, and about 10% of the studies measured methane emissions that were higher at lower temperatures (opposite to the effect predicted by the Arrhenius equation).
The reported ecosystem-level Ea is higher than what has been called the “universal temperature dependence” of aerobic metabolism4— an Ea of 0.67 ± 0.15 eV that encompasses the metabolism of a wide range of plants, protozoa, invertebrates and vertebrates — and is also higher than the average Ea (0.72 eV) observed for a diverse group of 50 aerobic and anaerobic microorganisms5. The higher average Ea reported here for methanogenic ecosystems could reflect either that the biochemistry of methanogens (which have an average Ea of 1.10 eV) directly limits methane emissions in some ecosystems, or that the organisms that supply methanogens with substrates have similarly high temperature dependence. Nevertheless, the clear implication of these findings is that methane production will increase more steeply with temperature than would be captured by climate-change models that assume methane emission is governed by more typical (lower) values of Ea. For example, over the range of global warming projected6 for this century (1.0–3.7 °C), an Ea of 0.96 eV suggests a 14–63% increase in methane emission compared with 10–40% for an Ea of 0.67 eV.
Yvon-Durocher and colleagues' findings constrain, and perhaps simplify, one piece of a much larger climate-change puzzle. Feedbacks from methane emissions in response to global climate change will ultimately derive from a combination of the direct temperature effects considered here and indirect effects such as thawing of permafrost (and the resultant availability of new organic matter), changes in vegetation, and large-scale inundation or drying of soils and wetlands. Moreover, methane's proportionate contribution to global warming (about 20% over the past century) may actually diminish as carbon dioxide takes an increasingly prominent role in the future7. But in this complex problem — unlike methane's flux to the atmosphere — every bit helps.
Kirschke, S. et al. Nature Geosci. 6, 813–823 (2013).
Bastviken, D. et al. Nature 507, 488–491 (2014).
Riley, W. J. et al. Biogeosciences 8, 1925–1953 (2011).
Gillooly, J. F. et al. Science 293, 2248–2251 (2001).
Tijhuis, L. et al. Biotech. Bioeng. 42, 509–519 (1993).
Stocker, T. F. et al. in Climate Change 2013: The Physical Science Basis. Working Group I Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (eds Stocker, T. F. et al.) Technical Summary, 33–115 (Cambridge Univ. Press, 2013).
Prather, M. et al. (eds) in Climate Change 2013: The Physical Science Basis. Working Group I Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (eds Stocker, T. F. et al.) Annex II, 1395–1445 (Cambridge Univ. Press, 2013).
About this article
Materials Research Bulletin (2021)
Journal of Industrial and Engineering Chemistry (2021)
Changes in archaeal ether lipid composition in response to agriculture alternation in ancient and modern paddy soils
Organic Geochemistry (2019)
Improved methane elimination by methane-oxidizing bacteria immobilized on modified oil shale semicoke
Science of The Total Environment (2019)
Journal of Chemical Technology & Biotechnology (2018)