Abstract
Coastal marshes and their carbon sequestration capacity are threatened by development and accelerated sea level rise, yet we lack an understanding of how carbon dynamics vary among dominant zones due to confounding feedbacks between environmental gradients and vegetation. We implemented a three‐year mesocosm experiment (48 379‐L tanks) manipulating plant species ( Spartina alterniflora , Phragmites australis , unvegetated tidal flat) under controlled hydroperiod and brackish salinity conditions to limit confounding physiochemical gradients. We quantified all biomass pools, net ecosystem exchange (NEE), and soil respiration (CO 2 ) during the 2021 and 2022 growing seasons, and soil microbial communities in 2021. We measured field‐based soil respiration (CO 2 ) rates in P. australis and S. alterniflora zones at one site in Stonington, Connecticut, USA, to compare with experimental fluxes. We observed greater biomass in vegetated than in tidal flat treatments, but no differences between vegetated P. australis and S. alterniflora treatments. Spartina alterniflora had the highest rates of NEE yet also had the highest soil respiration rates; both patterns were likely due to differential physiology between S. alterniflora and P. australis . Field‐based fluxes were comparable in magnitude to our experimental fluxes and similar across vegetation zones. Soil microbial community diversity was higher in vegetated treatments compared to unvegetated, whereas community composition moderately differed between vegetated treatments. Overall, growing two dominant coastal marsh grasses under similar physiochemical conditions neutralized many species differences typically observed in situ, suggesting strong environmental control on carbon cycling in coastal marshes.