Global warming is expected to affect methane (CH4) emissions from rice paddies, one of the largest human-induced sources of this potent greenhouse gas. However, the large variability in warming impacts on CH4 emissions makes it difficult to extrapolate the experimental results over large regions. Here, we show, through meta-analysis and multi-site warming experiments using the free air temperature increase facility, that warming stimulates CH4 emissions most strongly at background air temperatures during the flooded stage of ∼26 °C, with smaller responses of CH4 emissions to warming at lower and higher temperatures. This pattern can be explained by divergent warming responses of plant growth, methanogens, and methanotrophs. The effects of warming on rice biomass decreased with the background air temperature. Warming increased the abundance of methanogens more strongly at the medium air temperature site than the low and high air temperature sites. In contrast, the effects of warming on the abundance of methanotrophs were similar across the three temperature sites. We estimate that 1 °C warming will increase CH4 emissions from paddies in China by 12.6%─substantially higher than the estimates obtained from leading ecosystem models. Our findings challenge model assumptions and suggest that the estimates of future paddy CH4 emissions need to consider both plant and microbial responses to warming.
Keywords: meta-analysis; methane emissions; methanogens; methanotrophs; rice; warming.