Dynamic methane emissions in a restored wetland: Decadal insights into uncertain climate outcomes and critical science needs

Wetland restoration is increasingly viewed as a strategy for long-term carbon sequestration. However, methane (CH₄) emissions from restored wetlands can significantly offset their climate benefits. In this study, we analyzed one of the longest quasi-continuous methane eddy covariance datasets, spanning over a decade from the Mayberry wetland in the Sacramento-San Joaquin Delta, CA. Methane emissions at Mayberry initially spiked post restoration but have since declined, and this interannual trend positively affects the natural climate change solution potential of the restoration project. Using random forest analysis we find that the decadal trend in decreasing emissions aligns with a decadal trend in vegetation infill. A recent uptick in methane emissions is aligned with a decrease in porewater conductivity, indicating that porewater chemistry may also play a dominant role in driving methane fluxes. The isotopic signal of methane accumulated in sediments were remarkably stable over the past decade, indicating minimal changes in carbon lability of the re-flooded peat. We find that on a diel scale, latent heat is by far the dominant predictor for methane emissions, highlighting the role of diurnal patterns in plant transpiration. On seasonal timescales changes in water table depth and surface water conductivity help explain methane emissions. Our results emphasize the unique value of eddy-covariance and ancillary measurements initiated at the start of restoration in elucidating long-term methane dynamics in restored wetlands. They also highlight the critical need for expanded environmental data, such as porewater chemistry and vegetation changes, to comprehensively capture the factors driving methane flux.