Global change

Rice, microbes and methane

Methane is present at about only 1.8 parts per million in the atmosphere, but is a key player there — it is a greenhouse gas, it is central to atmospheric oxidation chemistry, and it is ultimately a source of stratospheric water vapour, which influences ozone depletion. Moreover, the concentration of methane is increasing rapidly. Hence the interest in the paper by Bodelier et al., on page 421of this issue1, which deals with methane emissions from rice paddies.

Most of the methane in the Earth's atmosphere comes from biological processes, and rice paddies are one of the main sources. A large fraction of the methane produced in rice soils is consumed, however, being oxidized to carbon dioxide by methane oxidizing bacteria (methanotrophs) in the soil, and so never makes it to the atmosphere. In upland soils, ammonium, which is formed naturally but is also a major constituent of nitrogen fertilizers, can inhibit methane oxidation and methanotroph growth. It has been a common assumption that this should occur in other ecosystems as well. So it comes as a surprise that Bodelier et al. find that, in rice-paddy soils, ammonium actually stimulates methane oxidation and methanotroph growth. This phenomenon may dominate the overall response of methane cycling to fertilization in rice-paddy ecosystems.

According to current estimates, rice agriculture will expand by up to 70% over the next 25 years to support the growing human population2. This will involve both increasing the area under cultivation and maximizing productivity by crop breeding and fertilizer management. Until now, it was thought that using nitrogen fertilizers on rice would increase trace-gas emissions. When nitrate-based fertilizers are used, much of the nitrate is denitrified, causing increased emissions of nitrous oxide, another potent greenhouse gas and ozone depleter. Ammonium fertilization also has the potential to increase methane emissions3 — not only does it increase plant growth and carbon flow to methane-producing bacteria (Fig. 1a, overleaf), but it can also inhibit methane oxidation (Fig. 1c).

Figure 1: Effects of ammonium fertilizer on methane dynamics in a rice ecosystem.

a, At the plant/ecosystem level, nitrogen increases plant growth and carbon supply to the methane producers, stimulating methane efflux from the soil. b, As Bodelier et al.1 show, at the level of the microbial community, nitrogen stimulates the growth and activity of methane-oxidizing (methanotroph) bacteria, leading to reduced net efflux. c, At the biochemical level, ammonium inhibits methane consumption because of competition for methane monoxygenase, and so increases efflux. The balance between these processes may vary. But Bodelier et al.1 find that effects at the level of the microbial community dominated the rice-ecosystem response to ammonium fertilization.

How does ammonium inhibit methane consumption? Several explanations have been proposed, the most solidly substantiated of which is competitive inhibition at the enzyme level4,5. This occurs because, at the molecular scale, methane and ammonium are similar in size and structure. As a result, the enzyme that oxidizes methane (methane monoxygenase) can bind to ammonium and react with it (Fig. 1c). Because the possibility of competitive inhibition is fundamental to the biochemistry of methane oxidation, it was generally thought that inhibition should occur in rice paddies as well as in upland systems. In fact, it may, and the work of Bodelier et al . does not rule it out.

However, the extent of competitive inhibition is proportional to the ratio between inhibitor and substrate. Methane concentrations are extremely high around a rice root, and plant nitrogen uptake may limit the exposure of bacteria to excess ammonium concentrations. These conditions should limit competitive inhibition. Equally importantly, because carbon (as methane) is readily available in rice soils, methanotrophs may become nitrogen-limited. So, at the level of the microbial community (Fig. 1b), addition of ammonium could stimulate methanotroph growth and activity to a much greater degree than it would inhibit them at the biochemical level (Fig. 1c ). The net effect could be to reduce overall methane efflux from the ecosystem.

Bodelier et al.1 show that, in a rice ecosystem, ammonium fertilization does indeed stimulate methanotroph growth and activity. Their research had three components: measurements of the rates of the reactions involved; studies using isotope tracers; and molecular analysis of the bacterial communities in the soil. The rate measurements showed a clear stimulation of methane consumption with fertilization. The tracer studies showed that, when 14C-methane was added, the 14C appeared in membrane lipids that are unique to methanotrophs. Finally, DNA-based molecular analysis identified the species of methanotrophs present, and showed that the rice plant stimulated the growth of type I methanotrophs, the group that responded most favourably to nitrogen fertilization.

In a rice paddy, the interactions of nitrogen fertilizer and the methane cycle are complex, with different effects occurring at different levels of organization (Fig. 1). At the ecosystem and biochemical levels, fertilization would lead to increased methane emissions ( Fig. 1a,c). As Bodelier et al. show, however, at the level of the microbial community, fertilization can stimulate methane consumption and reduce its efflux from the paddy (Fig. 1b). Which set of effects dominates the overall methane efflux may vary between systems, but responses at the level of the microbial community have generally not been considered. By incorporating these responses into their analysis, Bodelier et al. find that increasing rice production may not have such a severely detrimental effect on the atmosphere as has been assumed.

Further, their work highlights the importance of understanding ecological processes at the microbial-community level. This is a level of organization that has been little studied. Probably less than 1% of bacterial species have ever been isolated6, and many microbial processes have been studied in only a few ecosystem types. Only recently have researchers begun asking meaningful questions about how variations within physiological groups of microorganisms (methano-trophs, for instance) affect ecosystem-level processes such as trace-gas fluxes7. In large part, this has been because the molecular tools necessary for analysing microbial communities in situ have become available only in the past decade. Bodelier et al. have identified a gap in our understanding of microbial-community dynamics. As research continues, we are likely to find that some predictions about ecosystem behaviour fail because we are unaware of other such gaps.


  1. 1

    Bodelier, P. L. E., Roslev, P., Henckel, T. & Frenzel, P. Nature 403, 421–424 (1999).

  2. 2

    Neue, H. U. Soil Use Mgmt 13, 258–267 (1997).

  3. 3

    Banik, A., Sen, M. & Sen, S. P. Biol. Fertil. Soil 21, 319–322 (1996).

  4. 4

    Dunfield, P. & Knowles, R. Appl. Environ. Microbiol. 61, 3129–3135 (1995).

  5. 5

    Gulledge, J. & Schimel, J. P. Appl. Environ. Microbiol. 64, 4291–4298 (1998).

  6. 6

    Amann, R. I., Ludwig, W. & Schliefer, K.-H Microbiol. Rev. 59, 143– 169 (1995).

  7. 7

    Schimel, J. P. & Gulledge, J. Global Change Biol. 4, 745–758 (1998).

Download references

Author information



Corresponding author

Correspondence to Joshua Schimel.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Schimel, J. Rice, microbes and methane. Nature 403, 375–377 (2000).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.