Plants are a powerful proxy for global tidal marsh methane fluxes

Methane (CH 4 ) emissions can reduce the climate benefits of tidal marshes. Yet the drivers of tidal marsh CH 4 emissions remain poorly quantified, and salinity, the most well-established proxy for tidal marsh CH 4 fluxes, has low predictive performance. Here, we demonstrate that plant species out performs salinity as a single predictor of global tidal marsh CH 4 fluxes, providing a powerful and simple predictor on its own. A multiproxy approach combining plant species with other predictors (i.e., latitude, salinity, season) further improves predictions of global CH 4 fluxes. For our analysis, we compiled 87 studies with 2,094 mean measurements of CH 4 fluxes and used random forest and generalized additive modeling to investigate the relationship among CH 4 fluxes, salinity, and plants. We found that plant species was the most important predictor of CH 4 fluxes globally. Our model of plant species alone explained 62% of the variability in CH 4 fluxes and when including latitude, season, and soil salinity, explained 71%. We also developed a model with plant functional type for when plant species flux data are not available. We found that plant functional type alone explained 54% of the variability in CH 4 fluxes, underscoring the influential role of plants. Previously, polyhaline marshes were thought to have low CH 4 emissions. Here we show that CH 4 fluxes from these marshes offset between 1% and 39% of carbon sequestration, depending on the plant species present and selected global warming potential value. This plant species-based approach significantly improves global CH 4 flux estimates in tidal marshes and facilitates global carbon accounting.