Short-term variations in air-water gas transfer velocity and CO₂ fluxes over a temperate seagrass meadow

Abstract. Seagrass meadows store and sequester carbon for long periods of time in their tissues and sediments. They can also emit or take up gases such as carbon dioxide (CO₂) across the air-water interface. Air-water gas fluxes over seagrass meadows are poorly constrained, primarily due to methodological limitations and a persistent lack of fine-scale in situ spatial and temporal data. In this study, we present a thorough assessment of air-water CO₂ flux dynamics over a shallow, temperate coastal bay (South Bay, Virginia, USA) that has a benthic environment dominated by a temperate eelgrass meadow (Zostera marina). We evaluated major sources of variability in the flux calculation by analyzing in situ and regional wind speeds, comparing in situ and modeled gas transfer velocities (k₆₀₀), and conducting a long-term flux study to disentangle diurnal CO₂ flux patterns over South Bay. All wind stations located in proximity to land features significantly underestimated wind speeds over South Bay, while a wind station located in a nearby estuary over open water (Chesapeake Bay) was most representative of in situ wind speed measurements. We also derived in situ k₆₀₀ by deploying an “upside-down” aquatic eddy covariance system right below the air-water interface and using the data to ground truth the best-fitting empirical parameterizations for our site. The best-fitting models for South Bay were originally derived over the ocean and had an intercept of zero. Using the best-fitting model and continuous water column  measurements, we derived hourly air-water CO₂ fluxes over South Bay during the peak seagrass density period (early July) over three years. CO₂ fluxes exhibited a distinct diurnal pattern that correlated well with photosynthetically active radiation, demonstrating that fluxes are driven by photosynthesis and respiration from photosynthetic organisms, seagrasses, epiphytes, benthic microalgae and water column microalgae. Overall, South Bay was a small sink of CO₂ (−0.60 mmol CO₂ m⁻² d⁻¹) during the peak seagrass period across all three years. Due to the continuous nature of our study, we were able to derive CO₂ flux estimates during a storm period. South Bay was a source of CO₂ during the storm. When the storm period was removed and only fair-weather data were considered, South Bay was a larger sink of CO₂ (−4.20 mmol CO₂ m⁻² d⁻¹). Our results underline the importance of evaluating air-water gas fluxes over multiple day-night cycles and during multiple weather conditions.