Air-sea gas exchange in a seagrass ecosystem — results from a 3 He/SF 6 tracer release experiment

. Seagrass meadows are one of the most productive ecosystems in the world and could play a role inhelp mitigateing the increase of atmospheric CO 2 from human activities. However, Uunderstanding their the role of seagrasses in the global carbon cycle requires knowledge of air-sea CO 2 fluxes and hence knowledge of the gas transfer velocity. In this study, gas transfer velocityies and its controlling processes were examined in a seagrass ecosystem in south Florida. Gas transfer velocity 10 was were determined using the 3 He and SF 6 dual tracer technique in a seagrass ecosystem in south Florida, Florida Bay, near Bob Allen Keys (25.02663°N, 80.68137°W) between 1 and 8 April 2015. The observed gas transfer velocity normalized for CO 2 in freshwater at 20° C, k (600), was 4.8 ± 1.8 cm h -1 , which was . The resulting gas transfer velocities were lower than previous experiments in the coastal and open oceans at the same wind speeds. Therefore, using published that calculated from published wind speed/gas exchange parameterizations would overpredict gas transfer velocities and CO 2 fluxes in this area. 15 The deviation in k (600) from other coastal and offshore regions was only weakly correlated with tidal motion and air-sea temperature difference, implying that wind must therefore still beis the dominant factor driving gas exchange. The lower gas transfer velocity was most likely due to wave attenuation by seagrass and limited wind fetch in this the study area. A new wind speed/gas exchange parameterization is proposed ( 𝑘(600) = 0.143𝑢 102 ) , which might be applicable to other seagrass ecosystems and wind fetch limited environments.


Introduction
Seagrass meadows are one of the most productive ecosystems in the world and stock as much as 4. 2-8.4 PgC in their soils (Fourqurean et al., 2012). Because some of the the organic carbon produced via photosynthesisby seagrasses are refractory and accumulate easily sinks to the bottom on the seafloor and some of the organic carbon stays in the ocean as a refractory 25 matter, seagrass meadows are expected to be blue carbon sinks that can help mitigate the increase of anthropogenic CO2.
Seagrasses are estimated to bury 45−190 g C m -2 yr -1 , a significantly higher rate compared to terrestrial forests (0.7−13.1 g C m -2 yr -1 ; Mcleod et al., 2011;Duarte et al., 2005). However, Rrecently, the role of seagrasses in the global carbon cycle has been revisited, as CO2 carbon emissions from seagrasses CaCO3 production were found to be large (Howard et al., 2017;Van 2 dam et al., 2021;Schorn et al., 2021). Howard et al. (2017) examined the stock of organic and inorganic carbon in the soil of 30 seagrass meadows in Florida Bay and southeastern Brazil, and found that the soils in both regions have more inorganic carbon than organic carbon;. tThey suggested that both regions and are sources of CO2 to the atmosphere by assuming 0.6 mol of CO2 is produced when 1 mol of CaCO3 is produced. Schorn et al. (2021) also reported that the seagrasses in the Mediterranean Sea emit 106 μmol m -2 d -1 methane, mainly from their leaves.
Knowledge of the gas transfer velocity (k) is needed to understand the role of seagrass ecosystems in the global carbon 35 cycle, since air-sea CO2 flux is a function of k and the air-sea difference in the partial pressure of CO2 (pCO2). There are several methods to determine k in the field. The 3 He/SF6 dual tracer technique, which we employed in this study, is a mass balance technique that involves injecting these tracers into the ocean and determining k by measuring the change in the ratio of the two gases with time. The direct flux techniques, such as the eddy covariance method, measure the CO2 flux in the air and CO2 concentration both in the sea and air to derive k (McGillis et al. 2001). The k can has also been estimated from the heat transfer 40 velocity of heat by assuming that the gas and heat transfer velocities are related by their diffusivities;. howeverHowever, the estimated gas transfer velocity from heat, kH, havehas been found to overestimate the actual k (e.g., Atmane et al., 2004).
Because k is difficult to measure, it is often parameterized using easily and widely measured parameters such as wind speed. In deep offshore regions, wind is known to predict the gas transfer velocity well since wind creates waves and currents, which control turbulence and bubbles at the sea surface . Ho et al. (2018a) examined k in the Kaneohe 45 Bay in Hawai'i and showed that k can be estimated well by wind speed where the depth is deeper than 10 m. On the other hand, in shallow regions, other parameters become important as well (e.g., 2018b).  showed that k could be estimated well by wind speed and current speed in a shallow tidal estuary in south Florida, because the current enhances bottom-generated turbulence. Ho et al. (2018b) examined k in an emergent wetland where the depth < 1 m, and showed that k can be parameterized fromby heat flux, rain rate, and current velocity there. In the case of rain, rain rate is 50 included in the parameterization because rainfall increases subsurface turbulence and k (Ho et al., 1997a(Ho et al., , 2000. In Florida Bay, k has been estimated from commonly used wind speed/gas exchange parameterizations. Zhang and Fischer (2014) determined the air-sea CO2 flux to be 3.93 ± 0.91 mol m -2 yr -1 in Florida Bay; they used the wind speed/gas exchange parameterization determined from bomb-produced 14 C inventory in the ocean by Wanninkhof (1992). Van Dam et al. (2020) estimated k by using heat as a proxy (kH) in Florida Bay and found that kH wais lower compared with k derived 55 from published wind speed/gas exchange parameterizations when wind shear is relatively strong, even though kH is known to overpredict k. This finding suggests that previous wind speed/gas exchange parameterizations are not unsuitable for the seagrass-dominated area and a specific parameterization for these fetch-limited environments is needed. In the study presented here, we use a 3 He/SF6 tracer release experiment to determine k in a shallow seagrass-dominated environment to understand processes that control k and to derive a parameterization for this environment. 60 3 2 Methods

Study site
The 3 He/SF6 tracer release experiments werewas conducted between 1 and 8 April 2015 near Bob Allen Keys in Florida Bay (Fig. 1). Florida Bay is located in the southernmost part of Florida, USA. It is situated between the Everglades marsh and 65 the Florida Keys in the southernmost part of Florida, USA,, and covers approximately 2,000 km 2 . In this bay, the average depth is less than 3.5 m, and the vertical extent of seagrasses is between 0.08 and 0.2 m (Sogard et al., 1989). Thalassia testudinum and Laurencia are the dominatedominant seagrass and macroalgae, respectively, in the benthic communities, with an average standing crop of 63.6 and 8.9 g dry weight m -2 , respectively (Zieman et al., 1989). Seagrass density varies across the bay, and its standing crop is 0−20 g dry weight m -2 in the summer around our study area (bottom figure in Fig. 1) (Zieman 70 et al., 1989). The growth of seagrasses in Florida Bay show is seasonality, and theirwith larger standing cropcrops becomes larger in spring and summer and smallerthan in in fall and winter (Zieman et al., 1999). The phytoplankton community is dominated by cyanobacteria, diatoms, and dinoflagellates (Philips and Badylak, 1996), with. frequent Ccyanobacteria blooms occur frequently in the central north region of the bay due to nutrient input from the land (Philips et al., 1999;Lavrentyev et al., 1998). Wind is persistently blows from southeast to northwest during summer and from north to south during winter (Wang 75 et al., 1994). CurrentThe current speed is about 0.02−0.14 m s -1 (Wang, 1998), and the tidal amplitude is 0.1−0.4 m (Wang et al., 1994). The 3 He/SF6 tracer release experiments were conducted between 1 and 8 April 2015 near Bob Allen Keys (Fig. 1).

Tracer injection and underway SF6 measurementmeasurement
We injected 3 He and SF6 at a ratio of 1:340 into the water at the study location (25.0107°N, 80.692°W; black crossgreen 80 star in Fig. 1) on 1 April 2015 for 1 minute via a length of diffuser tubing. After injection, we performed underway SF6 measurements usingused an underway SF6 analysis system (Ho et al., 2002) that to measured surface water SF6 concentrations in the surface water every ~45 s. The system is composed of a gas extraction unit, which continuously removes SF6 from the water for measurement using a membrane contactor, and an analytical unit. The gas extraction unit continuously removes SF6 from the water for measurement using a membrane contactor. The other unit is composed of a gas chromatograph equipped 85 with an electron capture detector (GC/ECD). Based on previous experiments, tThe system has a detection limit of 1 × 10 -14 mol L -1 and an analytical precision of ±1% (Ho et al., 2018a). A personal computer showed the SF6 concentration displayed the SF6 concentrations in real time, which. This provided a spatial distribution of the SF6 patch, which guided the boat navigation and spatial survey.
Around Near the center of the SF6 patch between, 1 and 8 April 2015, we conducted 26 total stations for depth profiles 90 measurements of temperature and salinity with a conductivity, temperature, and depth (CTD) sonde and we took discrete 3 He and SF6 samples at a subset of these stations April 1 and 8 2015 (greenyellow triangles in Fig. 1). ing (see next paragraphbelow).

4
In total, 2.3 Discrete 3 He and SF6 measurements: 95 Wwe collected 16 3 He samples (~40 mL each) and 84 discrete SF6 samples. 3 He samples were taken at 26 stations in copper tubes mounted in aluminum channels and sealed at the ends with stainless steel clamps. between April 1 and 8 2015 (green triangles in Fig. 1). In the shore-based laboratory at the end of the experiment, 3 He and other gases were extracted from the water in the copper tubes and , transferred to flame-sealed glass ampoules, and . We measured 3 He concentration measured using a He isotope mass spectrometer (Ludin et al., 1998). 84 discrete SF6 samples were taken at the same stations (yellow 100 triangles in Fig. 1) using 50-mL glass syringes and submerged in water in a cooler until measurement back on shore at the end of each day. SF6 was extracted by a headspace technique and measured on a GC/ECD as described by Wanninkhof et al. (1987). We used the mean 3 He and SF6 concentration for each day to determine k, so there are six 3 He/SF6 data points between 3 and 8 April 3 and 8 (Fig. 2f). 105 2.34 Measurements of wind, temperature, salinity and tideenvironmental variables We measured wind speed, wind direction, and air temperature at ~5 m above sea level every 10 s using a sonic anemometer (Vaisala WMT700) near Bob Allen Keys (25.02663°N, 80.68137°W; blue dot in Fig. 1). The air temperature was averaged every 1 h to calculate the air-sea temperature difference (sea temperature minus air temperature). Additional wind speeds measured using a sonic anemometer (Vaisala WXT532) at ~3 m above the sea level at 25.07209°N, 80.73511°W (pink square 110 in Fig. 1, 7.4 km away from the blue dot) between 2015 and 2019 were obtained from U.S. National Park Service (NPS) (https://www.ndbc.noaa.gov/) Everglades National Park to compare k derived from this study and k estimated from published parameterizations.
Hourly tidal amplitude, water sea surface temperature, and salinity data from the same site (blue dot in Fig. 1) between 2015 and 2019 were obtained from NPSEverglades National Park (https://www.ndbc.noaa.gov/). The tidal amplitude was 120 measured using a digital shaft encoder (WaterLog H331), and sea surface. Water temperature and salinity were measured using multiparameter sondes (Hydrolab Quanta until 5 March 2019; OTT-Hydromet OTT-PLS-C thereafter). Additional wind speeds measured using a sonic anemometer (Vaisala WXT532) at ~3 m above the sea level at 25.07209°N, 80.73511°W (pink square in Fig. 1, 7.4 km away from the blue dot) between 2015 and 2019 were obtained from NPS to compare k derived from this study and k estimated from published parameterizations. 125 Wind speed data were extrapolated to 10 m above the sea level using the equation below (Amorocho and DeVries, 1980): where is the wind speed at height , κc is the von Kármán constant (0.41), 10 is the surface drag coefficient of wind at 10 m height (1.3×10 -3 ) (Stauffer, 1980).

Underway pCO2 Measurements 130
We measured the pCO2 along the boat track (red linedots in Fig. 1) using an underway system based on the design of Ho et al. (1997b) and incorporating the suggestions from Pierrot et al. (2009). Water was pumped through a thermosalinograph (TSG) into a showerhead equilibrator, and a high precisionhigh-precision thermistor measured the temperature. The gas was dried by Nafion and Mg(ClO4)2 dryers, and was continuously circulated through a non-dispersive infrared (NDIR; LI-COR LI-840A) analyzer. We stopped the flow during measurement and vented the NDIR cell to the atmosphere. The interval 135 between measurements was 41 s. Atmospheric air was taken from an inlet at the bow of the boat through a length of aluminum/plastic composite tubing (Dekabon), and was diverted into the NDIR analyzer at specific times (every ~72 min).
We calibrated the analyzer at regular time intervals (~72 min) with a 511 ppm CO2 standard calibrated with a primary standard from NOAA/ESRL/GMD and a CO2-free reference gas (UHP N2 passed through soda lime to remove CO2). In total, 1,261 and 13 xCO2 data were taken from the water and air, respectively. With measured mole fraction of CO2 (xCO2), barometric 140 pressure (P), and water vapor pressure at water surface temperature (Vp), we calculated the water and atmospheric pCO2 by applying the following expression (DOE, 1994): pCO2 = (P−Vp) × xCO2. pCO2 values were corrected for temperature shifts in the sample from the intake point (i.e., as measured by the TSG) to the pCO2 system using an empirical equation proposed by Takahashi et al. (1993). Fugacity of CO2 (fCO2) was calculated by fCO2= × CO 2 , where is an activity coefficient calculated from a formula in Wanninkhof and Thoning (1993). Additional fCO2 data were obtained from National Oceanic 145 and Atmospheric Administration (NOAA) Pacific Marine Environmental Laboratory (https://www.pmel.noaa.gov/) at 24.90°N, 80.62°W (cyan diamond in Fig. 1, 15 km away from the blue dot). CO2 flux between air and water was calculated with solubility (K0) and fCO2 using the equation below: where the K0 was calculated from the measured temperature and salinity (Weiss, 1974). 150 6 Fig. 1 Map of the study area. Green starThe black cross is the 3 He and SF6 injection location; red linesdots are the boat track where underway measurement was conducted for pCO2 and SF6; the blue dot is the location where wind velocity, air temperature, water temperature, salinity, and tidal amplitude were measured; the greenyellow triangle isare the stations where discrete samples for 3 He and SF6 were taken; the pink square is where additional wind velocity was measured; and cyanaqua 155 diamond indicates where additional fCO2 were measured. Note that water temperature, salinity, tidal amplitude, and additional wind velocity were obtained from E NPSverglades National Park, and additional fCO2 data were obtained from NOAA Pacific Marine Environmental Laboratory. Map data are generated by MATLAB geobasemap "darkwater" and downloaded from the Fish and Wildlife Research Institute (https://myfwc.com/research/) and NOAA Office for Coastal Management (https://www.noaa.gov/) 160 2.46. Gas transfer velocity calculationmeasurement and 3 He/SF6 ratio modeling The two tracers, 3 He and SF6, were injected together into the mixed layer at a constant ratio, and the ratio of 3 He/SF6 was measured over time as described above. The technique relies on the well-tested assumption that patch dilution, such as by horizontal mixing, affects the individual tracer concentrations but does not alter the 3 He/SF6 ratio; the only process that changes 165 the 3 He/SF6 ratio is air-sea gas exchange. The gas transfer velocity for 3 He, He 3 , can be determined as follows : where Sc SF 6 and Sc He 3 are the Schmidt numbers (i.e., the kinematic viscosity of water divided by diffusion coefficient of the gas in water) for SF6 and 3 He, respectively (see section 2.57). h is the measured water depth in Florida Bay, adjusted for tidal 170 variation. 3 Heexc is the 3 He in excess of solubility equilibrium with the atmosphere (used interchangeably with 3 He here). The gas transfer velocity measured during this experiment is normalized to k(600), where 600 corresponds to Sc number of CO2 in freshwater at 20°C: (4)

Calculation of Sc number 175
In the literature, Sc is often calculated from a compilation by Wanninkhof (2014). However, because the salinity in Florida Bay is 40, which is higher than the range provided by Wanninkhof (2014), we have re-calculated Sc for an extended range here. In our calculation, the kinematic viscosity for fresh water and seawater are derived using equations given by Sharqawy et al. (2010). Molecular diffusion coefficients of various gasses for freshwater were calculated using empirical equations derived from previous studies (Jähne et al., 1987;Wilke and Chang, 1955;Hayduk and Laudie, 1974;King and Saltzman, 180 1995;Saltzman et al., 1993;Zheng et al., 1998;De Bruyn and Saltzman, 1997). While the effect of temperature on molecular diffusion coefficient is well investigated, the effect of salinity has been the subject of fewer studies. SF6, methyl bromide (CH3Br), and trichlorofluoromethane (CFC-11) do not have significant differences in diffusion coefficients between fresh 8 water and a 35 g L -1 sodium chloride (NaCl) solution (King and Saltzman, 1995;De Bruyn and Saltzman, 1997;Zheng et al., 1998). However, diffusion coefficients for methane (CH4), dichlorodifluoromethane (CFC-12), and He in seawater are 4−7% 185 less than the coefficients in freshwater (Jähne et al., 1987;Saltzman et al., 1993;Zheng et al., 1998). To represent the dependence of molecular diffusion coefficients on salinity for gasses except for SF6, CH3Br and CFC-11, we linearly inter/extrapolated the molecular diffusion coefficients for various salinities by assuming that the diffusion coefficients decrease by 6% when the salinity is 35, compared with freshwater (Jähne et al., 1987;Wanninkhof, 2014). Molecular diffusion coefficients for a salinity of 40 are about 7% smaller compared to the coefficients for freshwater based on this assumption. 190 Least-squares fourth-order polynomial fit, including the effect of salinity, was produced to predict the Sc numbers at various temperatures and salinities (Table 1).

Modeling the decrease of 3 He/SF6 ratio
The decrease of the 3 He/SF6tracer ratio was compared to the decrease predicted by published wind speed/gas exchange parameterizations to assess the validity of these parameterizations for the study area. Under the assumption that air-sea gas 195 exchange is the only process that alters the 3 He/SF6 ratio in the water, the change in 3 He/SF6 ratio during this experiment can be modeled by an analytical solution to equation (3) where Robs n and Rmod n are the observed and modeled 3 He/SF6 tracer ratios, respectively, and N is the number of stations sampled after the initial sampling (5 for table 2 and 2 for Fig. 5e). The ability of commonly used parameterizations, including 205 the quadratic relationships of Wanninkhof (1992), Nightingale et al. (2000), and Ho et al. (2006), the exponential relationship of Raymond and Cole (2001), and the hybrid parameterization of Wanninkhof et al. (2009) to predict k in Florida Bay was evaluated by examining the cvRMSE. Equation (6) was also used to derive the optimal coefficients (A) for a quadratic (k = Au10m 2 ) parameterization by minimizing the cvRMSE. We regarded A with minimum cvRMSE as the best coefficient for parameterization. 210

Calculation of Sc number
In the literature, Sc is often calculated from a compilation by Wanninkhof (2014). However, because the salinity in Florida Bay is 40, which iswas higher than the range provided by Wanninkhof (2014),, we have re-calculated Sc for an extended 9 salinity range here. In our calculation, tThe kinematic viscositiesy for fresh water and seawater arewere deriveddetermined 215 using equations given by Sharqawy et al. (2010), and the . mMolecular diffusion coefficients of various gasses for freshwater were calculated using empirical equations derived from previous studies (Jähne et al., 1987;Wilke and Chang, 1955;Hayduk and Laudie, 1974;King and Saltzman, 1995;Saltzman et al., 1993;Zheng et al., 1998;De Bruyn and Saltzman, 1997). While the effect of temperature on molecular diffusion coefficient is well investigatedstudied, the effect of salinity has been the subject of fewer studies investigations. For SF6, CH3Br, and CFC-11, the diffusion coefficients in seawater are similar to those 220 in freshwater (King and Saltzman, 1995;De Bruyn and Saltzman, 1997;Zheng et al., 1998). SF6, methyl bromide (CH3Br), and trichlorofluoromethane (CFC-11) do not have significant differences in diffusion coefficients between fresh water and a 35 g L -1 sodium chloride (NaCl) solution (King and Saltzman, 1995;De Bruyn and Saltzman, 1997;Zheng et al., 1998).
However, the diffusion coefficients for methane (CH4), CFC-12, and He in seawater are 4-7% lower than those in freshwater (Jähne et al., 1987;Saltzman et al., 1993;Zheng et al., 1998). To represent the dependence of molecular diffusion coefficients 225 on salinity for gases other than SF6, CH3Br, and CFC-11, we linearly interpolated/extrapolated the molecular diffusion coefficients for different salinities by assuming that the diffusion coefficients decrease by 6% when the salinity is 35 compared to freshwater (Jähne et al., 1987;Wanninkhof, 2014). This assumption suggests that the diffusion coefficients for a salinity of 40 are approximately 7% smaller than those for freshwater. A least-squares fourth-order polynomial fit, incorporating the effect of salinity, was used to predict Sc values at various temperatures and salinities However, diffusion coefficients for 230 methane (CH4), dichlorodifluoromethane (CFC-12), and He in seawater are 4−7% less than the coefficients in freshwater (Jähne et al., 1987;Saltzman et al., 1993;Zheng et al., 1998). To represent the dependence of molecular diffusion coefficients on salinity for gasses except for SF6, CH3Br and CFC-11, we linearly inter/extrapolated the molecular diffusion coefficients for various salinities by assuming that the diffusion coefficients decrease by 6% when the salinity is 35, compared with freshwater (Jähne et al., 1987;Wanninkhof, 2014). Molecular diffusion coefficients for a salinity of 40 are about 7% smaller 235 compared to the coefficients for freshwater based on this assumption. Least-squares fourth-order polynomial fit, including the effect of salinity, was produced to predict the Sc numbers at various temperatures and salinities (Table 1).   Jähne et al. (1987); Ar, O2, N2, N2O, and CCl4 fit from Wilke and Chang (1955) adapted by Hayduk and Laudie (1974); SF6 measured by King and Saltzman (1995); DMS measured by Saltzman et al. (1993); CFC-11 and CFC-12 measured by Zheng et al. (1998); CH3Br measured by De Bruyn and Saltzman 245 (1997). Sc numbers for temperature of 20°C and salinity of 35 become larger than Sc numbers for temperature of 20°C and salinity of 0 by 4.7-4.8% for SF6, CFC-11 and CH3Br and 10.8-12.8% for other gasses, respectively. Note that the fits are based on simple assumptions (see section 2.5), and the dependence of Sc on salinity needs to be investigated further in the future.

Environmental parameters 12
During the experiment, wind direction was predominately from the east, and wind speeds increased towards the latter part of the study period (Fig. 2a). The mean and the standard deviation of the wind speed during the study period was 5.5 ± 2.0 m s -1 (range=0.12-12 m s -1 ). Mean wWater temperature showed a diurnal pattern with a mean and standard deviation of 26.3 ± 255 1.3°C (Fig. 2b). The diurnal pattern of the air temperature was weak, as with athe mean and standard deviation were of 25.1 ± 0.6°C. The air-sea temperature difference showed diurnal cycles, which waswere mainly driven by the diurnal cycle of the sea temperature, consistent with observations by Van Dam et al. (2020). Salinity was remained consistent throughout the study period (41 ± 0.1) (; not shown). The tide consistently showed semidiurnal cycles with an amplitude of ≤ 0.2 m throughout the study period. 260 13 14 Fig. 2 Time series of (a) hourly averaged wind vector at 10 m height (m s -1 ), (b) water temperature and air temperature (°C), (c) temperature difference (water temperature minus air temperature; units: °C), (d) tidal amplitude (units: m) and (e) measured and estimated k(600) which is the gas transfer velocities normalized for 20°C freshwater CO2 at in-situ temperature and salinity 265 and (f) measured and modeled change in 3 He/SF6. Note that the wind direction is towards the north when the vector is towards the left. The time zone is local time. The figure legend for (e) is the same as that in (f). The numbers in (f) indicate the periods corresponding to the x-axis in Fig. 5.

Gas transfer velocity in Florida Bay and assessment of published parameterization 270
The measured k(600) was 4.8 ± 1.8 cm h -1 (mean ± s.d.) (Figs. 2e and 3), which was lower than previous studies conducted in coastal and open oceans at the same wind speed (Fig. 3 of Ho & Wanninkhof, 2016). A new parameterization was produced based on results from this experiment by minimizing the cvRMSE of A • 10 2 , where A is a coefficient (Fig. 4):

Fig. 4
The relationship between cvRMSE and the coefficient A in the equation k(600)=A u10 2 . Three vertical lines indicate the 280 coefficients derived from this study, as well as those of Ho et al. (2006) and Wanninkhof (1992) from left to right. Note that k(660) in Wanninkhof (1992) was converted to k(600) by assuming that they scale as Sc to the power of −1/2.
The cvRMSE between the measured 3 He/SF6 and this new parameterization, equation (7), was 8.6%, while the cvRMSEs calculated from previously published wind speed/gas exchange parameterizations were more than 70% (Table 2). The 285 coefficient of 0.143 was 46% and 56% lower than the k(600) of 0.266 and 0.325 from Ho et al. (2006) and Wanninkhof (1992), respectively (Fig. 3). The result of previous studies which that used the parameterization of Wanninkhof (1992) in Florida Bay was modified in section 3.32. The estimated k(600) for CO2 at in-situ temperature and salinity derived from equation (7) was 6.35.5±3.03 cm h -1 , while all the published parameterizations estimated over 9.010 cm h -1 on average between 3 and 8 April 2015 (Table 2 and Fig. 2e). k for CO2 at in-situ temperature and salinity between 2015 and 2019 were also calculated using 290 the equation (7) and the previously published parameterizations (Table 3). Annual averaged k ranged between 3.7-4.3 cm h -1 in Florida Bay between 2015 and 2019, while published parameterization would yieldsyield values of 6.9-11.6 cm h -1 .
The deviations of between observed 3 He/SF6 and modeled 3 He/SF6 , which is derived from published parameterizations, become larger with time, as shown in ( Figure 2f). This means indicates that the published parameterizations overpredict k in Florida Bay, which is consistent with the result findings of Van Dam et al. (2020). 295 Table 2. Gas transfer velocities determined from this study and published parameterization.

References Parameterization Mean k(600) (cm h -1 ) cvRMSE
This study k(600)= 0.143u10 2 5.5±3.0 8.6% Wanninkhof (1992) k (660) The observed k(600) was 4.8 ± 1.8 cm h -1 (average ± standard deviation). Note that k(660) is was converted to k(600) by assuming that the scale by Sc to the power of -1/2.  The standard deviation of Raymond and Cole (2001) werewas large in 2017 since wind speed was as high as 27.5 m s -1 , and k was as high as 6.1×10 3 cm h -1 . 305 The deviations between observed and modeled 3 He/SF6, which is derived from published parameterizations, become larger with time ( Figure 2f). This indicates that the published parameterizations overpredict k in Florida Bay, which is consistent with the findings of Van Dam et al. (2020).
Van Dam et al. (2020) estimated the air-sea gas transfer velocityk using heat as a proxy (kH) in Florida Bay. They found 310 that kH was lower than k calculated from published parameterization even though kH is known to overpredict k. They suggested that the stratification due to temperature restricts air-sea gas exchange since the deviation between kH and k from commonlyused parameterization was large when the air-sea temperature difference was large. To investigate the relationship between environmental parameters and the deviation between measured and estimated air-sea gas exchange, we examined the relationship between temperature difference and the deviation between observation and the models by calculating cvRMSE 315 separately in four periods (Fig. 5). We found no clear relationship between the deviation and air-sea temperature difference.
The deviation observed in Van Dam et al. (2020) might be due to the fact that kH contains the air-sea temperature difference in its equation (equation 7 in Van Dam et al. 2020); kH becomes smaller when the air-sea temperature difference is large and vice versa. The deviation between observation and models was generally larger when wind speed was higher. cvRMSE became the largest values for all parameterizations in Period 4 when the mean wind speed was 7.3 m s -1 . 320 The new wind speed/gas exchange parameterization predicts the observed change in 3 He/SF6 well ( Fig. 2f and Table 2), suggesting that wind is the dominant factor controlling gas exchange in this area. In Florida bay, waves are damped by seagrasses (Prager and Halley, 1999), which might could be one of the causes of lower k in this study. There is also the possibility that limited wind fetch in this region led to relatively weak waves and turbulence compared to other regions, contributing to lower k. Wind fetch is limited in this region, since the wind mostly blows from east to west, and the Florida 325 Keys restricts the water exchange between the bay and the Atlantic Ocean ( Fig. 1 and Fig. 2a). There was almost no rainfall to affect k during the study period. The tidal amplitude was small (~0.1 m) (Fig. 2d), suggesting that the bottom-generated turbulence was weak. The cvRMSE, (f) mean wind speed (m s -1 ) and air-sea temperature difference (°C) during the period of 1-4. The x-axis represents the periods in Fig. 2f. 335

Implications for biogeochemistry
Although the experiment was conducted over a short period of 8 days, our new parameterization, equation (7), fit the observations well; This implies that equation (7) can be applied even in different seasons and years if the wind speed is in the range of 0.12-12 m s -1 and seagrass conditions are similar (dominant seagrass of Thalassia testudinum has 63.6 (range=0-340 215) g dry weight m -2 standing crop in Florida Bay (Zieman et al., 1989)). The parameterization determined in this study should be applicable to other seagrass ecosystems as well, since seagrass ecosystems are typically in coastal regions. In these environments, waves are damped by seagrasses and limited fetch. This wind speed/gas exchange parameterization proposed here might be applicable not only in seagrass ecosystems but also in other wind-fetch limited areas. To assess the applicability of this new parameterization in other inland ecosystems, additional 3 He/SF6 dual tracer experiments will need to be conducted. 345 Specifically, measuring the seagrass density and conducting dual-tracer experiments simultaneously is needed to relate the k and vegetation distribution.
The observed daytime pCO2water and pCO2air were 2248 ± 126 and 3913 ± 3 μatm, respectively (Fig. 6a). The pCO2water of 2248 ± 126 μatm was in the range shown by Zhang and Fischer (2014), who examined the pCO2water in the whole basin of the Florida Bay from 2006 to 2012, and showed that pCO2water minimuma was ~200 μatm in April (Fig. 3

of Zhang and 350
Fischer 2014). Since the observed pCO2water was lower than pCO2air, CO2 goes from the air to the sea during the daytime in the observation period (between 3 and 8 April 2015). The calculated daytime CO2 flux using the measured pCO2 difference and modeled k in this study (Black solid line in Fig. 2e) was -5.3 ± 3.0 mmol m -2 day -1 (negative value denotes CO2 flux from the air to the water) (Fig. 6b). The CO2 flux varied both within a day and between days mainly due to the variability in k (Note that k(600) in Fig. 2e is filtered with 25 minutes running average). Diurnal fCO2water amplitude at the NOAA station (cyan 355 diamond in Fig. 1) between 3 and 8 April 2015 was as small as 25-53 μatm, and so we calculated daily CO2 flux by assuming CO2 difference between air and water during the night is the mean daytime CO2 difference. The calculated daily CO2 flux using the measured pCO2 difference and modeled k in this study (Black solid line in Fig. 2e) was -7.04.4 ± 3.52.6 mmol m -2 day -1 (negative value means CO2 goes from the air to the sea) (Fig. 6b)., which was higher than daytime CO2 flux because wind speed was higher during the night. Although we did not conduct pCO2 measurement during the night and so the calculated 360 value is biased toward daytime, the daily averaged pCO2water and CO2 flux during the whole observation period would still be lower than pCO2air and negative, respectively, considering that the observed pCO2 was as low as 228 μatm and the CO2 flux at the NOAA station (aqua diamond in Fig. 1) was always negative with diurnal fCO2water amplitude of 25-53 μatm between April 3 and 8, 2015. 365 22 Fig. 6 Time series of (a) measured pCO2water (blue dots) and pCO2air (cyan dots) (units: μatm), (b) calculated CO2 flux (units: mmole m -2 day -1 ). Negative CO2 flux indicates that the sea is a sink of CO2. The time zone is local time.
Contrary to our experimental period, the aAnnual averaged CO2 flux is, however, known to be from the water to the air in 370 Florida Bay (e.g., Zhang and Fischer, 2014;Van dam et al., 2021). The pCO2 and CO2 flux in Florida Bay are suggested to have seasonality due to cyanobacteria blooms (Zhang and Fischer, 2014). The seasonality of seagrasses may also contribute to the seasonality of pCO2 and CO2 flux, as its productivity also shows seasonality (higher in spring and summer and lower in fall and winter) (Zieman et al., 1999). Zhang and Fischer (2014) measured the pCO2water for the whole area of the Florida Bay and estimated the CO2 flux in Florida Bay to be 3.93 ± 0.91 mol m -2 yr -1 using the parameterization of Wanninkhof (1992); 375 23 we recalculated the CO2 flux to be 1.73 ± 0.40 mol m -2 yr -1 by multiplying 0.44 (e.g., 1 minus 0.56; see section 3.2). By conducting atmospheric eddy covariance measurements near the Bob Allen Keys (blue dot in Fig. 1), Van Dam et al. (2021) showed that the CO2 flux in Florida Bay wasis 6.1-7.0 mol m −2 year −1 , which is significantly higher than the corrected value of 1.73 ± 0.40 mol m -2 yr -1 in Zhang and Fischer (2014). Although the reason is not clear, primary production by phytoplankton and seagrasses might be lower when Van Dam et al. (2021) conducted their observation (2019-2020), resulting in higher CO2 380 flux from sea to air, since there is no negative mean CO2 flux in spring when they conducted their measurements (Fig. 1a in Van Dam et al., 2021). Van Dam et al. (2021) also calculated the excess CO2, which is the CO2 concentration difference between water and air to achieve the annual CO2 flux of 6.1-7.0 mol m −2 year −1 , in Florida Bay to be between 5.2 and 6.0 μmol kg -1 , using a mean k of 11.7 cm h -1 ; we recalculated the excess CO2 to be between 14 and 16 μmol kg -1 using the k of 4.3 cm h -1 , which is parameterized from thise current study (Table 3). The recalculated excess CO2 almost doubles their calculation 385 of Van Dam et al. (2021)5.2-6.0 μmol kg -1 and hence requires significantly more CO2 input.

Summary
Air-sea gas exchange was investigated in a seagrass ecosystem in South Florida, USA, using the 3 He and SF6 dual tracer 390 technique. The gas transfer velocity was lower than that in other coastal areas and open oceans, and commonly-used wind speed/gas exchange parameterizations tend to overpredict the gas transfer velocitiesy, especially when wind speeds wereas relatively high (> 7 m s -1 )strong. A new wind speed/gas exchange parameterization was proposed ( (600) = 0.143 10 2 ), which was able to predict fitted well to the observed gas transfer velocities significantly better than existing parameterizationsexchange. This result suggests that wind is the dominant factor controlling gas exchange in the studied 395 seagrass ecosystem, but the lower gas transfer velocity at a given wind speed was due to limited wind fetch in the study area and wave attenuation by seagrass. To assess the wider applicability of the proposed wind speed/gas exchange parameterization, more tracer release experiments are needed at similar inland ecosystems.

Data availability
The data used for this article is found at https://doi.org/10.5281/zenodo.6730934. Click "Version Florida 10.5281/zenodo.7087773" in the right column.

Author contributions
DH conceived, designed, and conducted the experiment. RD performed the data analysis. 405