Tidal marshes store large amounts of carbon; however, little is known about the patterns, magnitudes, and biophysical drivers that regulate CO 2 efflux from these ecosystems. Due to harsh environmental conditions (e.g., flooding), it is difficult to measure continuous soil CO 2 efflux in tidal marshes. These data are necessary to inform empirical and process-based models and to better quantify carbon budgets. We performed automated (30 min) and manual (bi-monthly) soil CO 2 efflux measurements, for ~20 months, at two sites in a temperate tidal marsh: tall Spartina (TS; dominated by S. cynosuroides) and short Spartina (SS; dominated by S. alterniflora). These measurements were coupled with water quality, canopy spectral reflectance, and meteorological measurements. There were no consistent diel patterns of soil CO 2 efflux, suggesting a decoupling of soil CO 2 efflux with diel variations in temperature and tides (i.e., water level) showing a hysteresis effect. Mean soil CO 2 efflux was significantly higher at SS (2.15 ± 1.60 μmol CO 2 m -2 s -1 ) than at TS (0.55 ± 0.80 μmol CO 2 m -2 s -1 ), highlighting distinct biogeochemical spatial variability. At the annual scale, air temperature explained >50% of the variability in soil CO 2 efflux at both sites; and water level and salinity were secondary drivers of soil CO 2 efflux at SS and TS, respectively. Annual soil CO 2 efflux varied from 287-876 to 153-211 g C m -2 y -1 at SS and TS, respectively, but manual measurements underestimated the annual flux by <67% at SS and <23% at TS. These results suggest that measuring and modeling diel soil CO 2 efflux variability in tidal marshes may be more challenging than previously expected and highlight large discrepancies between manual and automated soil CO 2 efflux measurements. New technical approaches are needed to implement long-term automated measurements of soil CO 2 efflux across wetlands to properly estimate the carbon balance of these ecosystems., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2021 Elsevier B.V. All rights reserved.)