Measuring and Modeling Ecosystem Carbon Budgets and Greenhouse Gas Exchange from Restored Marshes across a Salinity Gradient

Wetland restoration is a high priority as we try to mitigate climate and adapt to sea level rise. However, quantification of carbon budgets in these systems is complex and difficult to model at the annual time scale. Net ecosystem carbon budgets are even more difficult to quantify in tidal marshes which have lateral or hydrologic exchanges of carbon. We will present the state of our knowledge concerning drivers of net ecosystem carbon balance and greenhouse gas exchange across a salinity gradient in the San Francisco Bay-Delta, synthesizing measurements and analyses from field sites in the Sacramento-San Joaquin River Delta, Suisun Bay, and San Francisco Bay. Results to date suggest that flooding events can inhibit or enhance photosynthesis, depending on height of vegetation and impacts on VPD. While photosynthesis was lower in polyhaline sites compared to mesohaline and freshwater marshes, lower ecosystem respiration was also observed resulting in overall high annual carbon uptake (Polyhaline: -408 g C-CO₂ m^-2 yr^-1, Mesohaline: -182 to -397 g C- CO₂ m^-2 yr^-1; Freshwater: -223 to -454 g C- CO₂ m^-2 yr^-1). Preliminary results from sediment accretion measurements, ⁷Be dating, and hydrologic carbon flux measurements in the tidal marshes suggests that 40-45% of annual net atmospheric carbon uptake is lost to hydrologic transport. CH₄ emissions are high in freshwater wetlands and low to undetectable at the mesohaline and polyhaline sites (Polyhaline: 0.15 g C-CH₄ m^-2 yr^-1, Mesohaline: 1 to 1.3 g C- CH₄ m^-2 yr^-1; Freshwater: 38.7 to 53 g C- CH₄ m^-2 yr^-1). However, high temperature events led to high CH₄ emissions even in high salinity sites, suggesting that increasing temperatures and heat waves may threaten greenhouse gas sequestration potential of meso- to polyhaline marshes. We will also present a biogeochemical process-based model used to predict annual greenhouse gas budgets in these ecosystems using a model-data fusion approach. Modeling results suggest greenhouse gas budgets can be predicted within 5-11% for CO₂ and 15-20% for CH₄ at these field sites. Future research is aimed at improving predictions of atmospheric and hydrologic carbon exchange in restored marshes and upscaling model predictions to the Bay-Delta region using remote sensing.