Dynamics of methane flux in permafrost-affected wetlands: A meta-analysis of permafrost continuity effects and hydrological controls.

Permafrost wetlands are critical and vulnerable components of northern ecosystems, with their methane (CH) emissions representing a major uncertainty in Earth system models. Previous syntheses have disagreed on how permafrost continuity modulates CH fluxes, leaving a blind spot in climate projections. We hypothesize that degradation of permafrost continuity from continuous to discontinuous to sporadic creates a gradient of environmental conditions that drives an exponential shift in CH emissions. To test this, we synthesized 153 growing-season CH flux observations from 40 global permafrost wetland studies. Our analysis reveals a clear emission gradient: continuous, discontinuous, and sporadic permafrost wetlands exhibited median CH fluxes of 0.09, 0.24, and 4.73 mg CH m h, respectively. Water-table depth (WT) emerged as the dominant predictor of CH flux (XGBoost RMSE = 0.78 mg CH m h, R = 0.59), with a nonlinear threshold at ∼-10 cm. Fluxes were negligible below this depth but increased markedly above it, indicating a hydrological tipping point. Structural equation modelling showed that the composite permafrost factor (β = 0.26) and hydrological conditions (β = 0.26) jointly explained ≈0.37 of the total R = 0.48. Environmental analyses highlighted systematic shifts across categories: sporadic sites had the shallowest WT (-4.73 cm) and deepest active layers (115.3 cm), conditions that expand methanogenic soil volumes while compressing oxidation zones. Continuous sites showed higher soil bulk density and pH, whereas sporadic sites accumulated more organic matter. These biophysical changes, coupled with vegetation turnover, further modulated CH dynamics: tree-dominated sites suppressed emissions, while sedge-dominated zones enhanced them under wetter conditions. Together, our findings demonstrate a cascading, nonlinear relationship linking permafrost degradation, thaw depth, hydrology, microbial processes, and vegetation to CH flux. This mechanistic understanding underscores the need to integrate permafrost-hydrology-microbe-vegetation feedback into models to reduce uncertainty in Arctic methane-climate projections.