Abstract
Freshwater lakes emit large amounts of methane, some of which is produced in oxic surface waters. Two potential pathways for aerobic methane production exist: methanogenesis in oxygenated water, which has been observed in some lakes, or demethylation of small organic molecules. Although methane is produced via demethylation in oxic marine environments, this mechanism of methane release has not yet been demonstrated in freshwater systems. Genes related to the C-P lyase pathway, which cleaves C-P bonds in phosphonate compounds, were found in a metagenomic survey of the surface water of Lake Matano, which is chronically P-starved and methane-rich. We demonstrate that four bacterial isolates from Lake Matano obtain P from methylphosphonate and release methane, and that this activity is repressed by phosphate. We further demonstrate that expression of phnJ, which encodes the enzyme that releases methane, is higher in the presence of methylphosphonate and lower when both methylphosphonate and phosphate are added. This gene is also found in most of the metagenomic data sets from freshwater environments. These experiments link methylphosphonate degradation and methane production with gene expression and phosphate availability in freshwater organisms, and suggest that some of the excess methane in the Lake Matano surface water, and in other methane-rich lakes, may be produced by P-starved bacteria. Importance: Methane is an important greenhouse gas, and contributes substantially to global warming. Although freshwater environments are known to release methane into the atmosphere, estimates of the amount of methane emitted by freshwater lakes vary from 8 to 73 Tg per year. Methane emissions are difficult to predict in part because the source of the methane can vary: it is the end product of the energy-conserving pathway in methanogenic archaea, which predominantly live in anoxic sediments or waters, but have also been identified in some oxic freshwater environments. More recently, methane release from small organic molecules has been observed in oxic marine environments. Here we show that demethylation of methylphosphonate may also contribute to methane release from lakes, and that phosphate can repress this activity. Since lakes are typically phosphorus-limited, some methane release in these environments may be a byproduct of phosphorus metabolism, rather than carbon or energy metabolism. Methane emissions from lakes are currently predicted using primary production, eutrophication status, extent of anoxia, and the shape and size of the lake; to improve prediction of methane emissions, phosphorus availability and sources may also need to be included in these models.