Methane and carbon dioxide dynamics in a subtropical estuary over a diel cycle: Insights from automated in situ radioactive and stable isotope measurements

Marine Chemistry 168 (2015) 69–79

Contents lists available at ScienceDirect

Marine Chemistry journal homepage: www.elsevier.com/locate/marchem

Methane and carbon dioxide dynamics in a subtropical estuary over a diel cycle: Insights from automated in situ radioactive and stable isotope measurements Damien T. Maher a,⁎, Kirsten Cowley b, Isaac R. Santos a, Paul Macklin a, Bradley D. Eyre a a b

Centre for Coastal Biogeochemistry, School of Environment, Science and Engineering, Southern Cross University, PO Box 157, Lismore, NSW 2480, Australia Department of Environment & Geography, Macquarie University, NSW 2109, Australia

a r t i c l e

i n f o

Article history: Received 1 September 2014 Received in revised form 22 October 2014 Accepted 31 October 2014 Available online 8 November 2014 Keywords: Air–water gas exchange Greenhouse gas Carbon stable isotopes Floodplain Coastal acid sulfate soils Wetland

a b s t r a c t Estuaries have high rates of primary production and respiration and can be hotspots for carbon dioxide and methane enriched submarine groundwater discharge. Here, we report high resolution pCO2, CH4, δ13C–CO2, δ13C–CH4 and radon (222Rn, a natural groundwater tracer) observations along North Creek estuary, Australia (S28°48′, E153°34′) during four spatial surveys over a diurnal cycle in January 2013 (summer). There were distinct tidal and diurnal differences in estuarine pCO2 and CH4, which lead to tidal differences of 3.6 fold and 5 fold in the estimated CO2 and CH4 diffusive water to air fluxes respectively, and up to a 2.4 fold difference in diurnal flux estimates of CH4. Carbon stable isotopes revealed tidal and diurnal differences in the source δ13C value of CO2 and CH4, and minor CH4 oxidation within the estuary. The CO2 outgassing rates based on the spatial surveys were different than the outgassing derived from three fixed time series stations along the estuary. There was agreement between the methods in the lower and upper estuary where pCO2 had a relatively low range over the study (~ 600 μatm and 3000 μatm respectively). However, in the mangrove surrounded mid estuary where pCO2 ranged from ~1450 to 11,000 μatm over a tidal cycle, fluxes estimated by the survey method were ~30% of the time series estimates. This study highlights the importance of considering tidal and diurnal variability when estimating the flux of CO2 and CH4 from estuaries, and discusses how a combination of diurnal (productivity/respiration) and tidal (groundwater/mixing) processes may drive surface water pCO2 and CH4 over short-term time scales. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Atmospheric concentrations of CO2 have increased by approximately 40% since the industrial revolution and, concurrently, atmospheric CH4 concentrations have more than doubled since pre-industrial times (IPCC, 2007). There is still a paucity of data from a number of biomes to adequately constrain the global CO2 and CH4 cycles. Despite the relatively small surface area, global estuarine CO2 degassing is estimated to be as large as the CO2 uptake by the continental shelf (Borges, 2005; Borges et al., 2005; Cai, 2011), but large uncertainties remain in estimates of the global estuarine CO2 flux (Laruelle et al., 2010; Borges and Abril, 2011; Cai, 2011; Maher and Eyre, 2012). Even fewer estimates are available on the global estuarine CH4 flux (Borges and Abril, 2011; Oritz-Llorente and Alvarez-Cobelas, 2012). Difficulties in calculating global emissions of CO2 and CH4 from estuaries include the lack of empirical data on controlling factors, the high degree of spatial and

⁎ Corresponding author. E-mail address: [email protected] (D.T. Maher).

http://dx.doi.org/10.1016/j.marchem.2014.10.017 0304-4203/© 2014 Elsevier B.V. All rights reserved.

temporal variability in these systems, and the lack of data in subtropical and tropical systems. Previous studies have found high temporal and spatial variability in pCO2 and CH4 concentrations within estuarine systems (Sansone et al., 1999; Middelburg et al., 2002; Chen et al., 2008; Macklin et al.,2014). However it has generally been observed that estuarine systems are net sources of CO2 and CH4 to the atmosphere. Typically pCO2 and dissolved CH4 concentrations are measured during a spatial survey or at fixed time series stations, however these sampling strategies may not adequately measure the high degree of both spatial and temporal variability often found in estuarine systems. Fixed time series measurements using automated instrumentation (e.g. Atkins et al., 2013) capture temporal variability extremely well, with the resolution of measurements typically at the scale of minutes. However the measurements are restricted to a single location, preventing assessment of spatial variation. Survey methods using continuous underway measurements or discrete sampling offer the advantage of determining spatial variation, and are therefore the most commonly utilized method for estimating estuary-wide fluxes. However single surveys do not capture variability induced by diel and tidal cycles, which may be significant. For example

70

D.T. Maher et al. / Marine Chemistry 168 (2015) 69–79

CO2 and CH4 concentrations in mangrove creeks and marshes can vary by up to an order of magnitude over a tidal cycle (e.g. Borges et al., 2003; Wang and Cai, 2004; Bouillon et al., 2007; Gatland et al., 2014), and significant diel variability in pCO2 and CH4 concentrations driven by photosynthesis and respiration has been observed in estuarine macrophyte beds (Maher et al., 2013b; Saderne et al., 2013). Therefore, careful consideration of diel and tidal effects upon pCO2 and CH4 concentrations need to be assessed when calculating whole system fluxes. A combination of factors may drive the diurnal and tidal variation of CO2 and CH4 fluxes in estuaries. Estuaries can have high productivity and respiration rates (Kemp et al., 1997; McGlathery et al., 2001; Middelburg and Herman, 2007; Maher and Eyre, 2012) which influence pCO2 on diurnal scales through the uptake (photosynthesis) and release (respiration) of CO2 during day and night periods, respectively. Diel cycles may also influence surface water CH4 concentrations, with previous studies finding light-induced inhibition of CH4 oxidation rates in freshwater lakes (Murase and Sugimoto, 2005; Tang et al., 2014). In addition, productivity and respiration driven cycles of sediment oxygen dynamics may lead to oscillation in aerobic methane oxidation and methane production rates (King et al., 1990; Maher et al., 2013b), and methylated substrates produced by phytoplankton have been shown to sustain aerobic CH4 production in the ocean (Karl et al., 2008). The actions of tides can deliver groundwater enriched in solutes through tidal pumping, directly influencing surface water pCO2 and CH4 (Borges et al., 2003; Dulaiova et al., 2006; Bouillon et al., 2007; Kristensen et al., 2008; Dorsett et al., 2011; Atkins et al., 2013). In addition, groundwater can deliver dissolved organic matter (Santos et al., 2008; Bauer and Bianchi, 2011) and nutrients (Gleeson et al., 2013; Charbonnier et al., 2013; Santos et al., 2013; Makings et al., 2014) to surface waters, which may indirectly influence pCO2 through enhanced autotrophic and/or heterotrophic metabolism. Therefore calculations of estuarine air–water CO2 and CH4 fluxes determined by surveys of surface water pCO2 and CH4 that do not capture tidal and diurnal variability may produce inaccurate estimates. Determining the major drivers of estuarine surface water pCO2 and CH4 involves disentangling the various tidal and diel processes. Recent development of cavity ring-down spectroscopy (CRDS) for high resolution in situ measurements of δ13C–CO2 and δ13C–CH4 may help constrain the source δ13C value of the dissolved CO2 and CH4 pools (Bass et al., 2013; Maher et al., 2013b; Gatland et al., 2014), thereby giving insights into the processes controlling surface water pCO2 and CH4 concentrations. For example, respiration will produce CO2 with a similar δ13C to the organic matter source (Lin and Ehleringer, 1997), and the dominant CH4 production pathways produce CH4 with differing δ13C values (Whiticar et al., 1986). In addition, development of high precision continuous measurement of in situ 222Rn (a natural groundwater tracer with a half-life of 3.84 days) now enables qualitative and quantitative assessment of groundwater input to estuaries (Dulaiova et al., 2005; Peterson et al., 2010). By combining simultaneous measurements of surface water 222Rn, pCO2, CH4, and δ13C–CO2 and δ13C–CH4 insights into the contribution of biological (e.g. primary production/ respiration) versus hydrological (e.g. groundwater and surface runoff) drivers of surface water CO2 and CH4 concentrations can be gained. We hypothesize that estimates of CO2 and CH4 fluxes from estuaries will vary significantly over diurnal and tidal cycles, due to tidal pumping and ecosystem metabolism. We tested this hypothesis by undertaking four surveys over a 24 hour period along a small shallow estuary making continuous underway measurements of 222Rn, pCO2, pCH4, δ13C–CO2 and δ13C–CH4 along with other physico-chemical parameters (wind speed, temperature, dissolved oxygen, chromophoric dissolved organic matter, chlorophyll a, and alkalinity). In addition three fixed time series stations were simultaneously deployed along the length of the estuary measuring pCO2 and 222Rn to compare water to air flux estimates and temporal variability in concentrations between the survey and the fixed time series method.

2. Methods 2.1. Study site North Creek is a small subtropical estuary located on the North Coast of New South Wales, Australia (Fig. 1). The estuary is shallow (average depth ~ 2 m) and well mixed with a tidal range of ~ 2 m. The region has a mean annual rainfall of 1863 mm with the wetter months occurring over summer and autumn. Intense rain events associated with tropical low-pressure systems contribute a majority of this rain to the catchment (Eyre and Pont, 2003). The hydrology of North Creek has been modified significantly in the last 100 years, through extensive drainage of wetland areas, which have been converted to agricultural land. An extensive drainage network including floodgates was constructed to facilitate drainage and prevent tidal inundation (King, 2003). Groundwater has been found to account for up to ~76% of surface waters of the upper reaches of North Creek (Atkins et al., 2013), but earlier investigations did not assess the mid and lower estuarine sections. For the purposes of this study, the estuary was divided into three zones (Fig. 1). The first zone, the lower estuary, comprises of large marine influenced tidal flats. Land usage is largely residential, with the city of Ballina (population ~ 8000) surrounding the lower estuary. Riparian vegetation is largely absent however there are small patches of remnant mangrove, salt marsh and river-flat forest communities. The mid-reaches (hereafter termed mid-estuary) is located within Ballina Nature Reserve, with an intact riparian zone of mangrove and swamp sclerophyll forests (King, 2003). The upper reach (hereafter termed upper estuary) is highly modified. It has been straightened to provide drainage for agriculture (predominantly sugar cane cultivation) and is generally devoid of riparian vegetation. 2.2. Spatial survey Four spatial surveys were conducted over two consecutive tidal cycles starting on the 15th of January 2013. A boat was driven at 5– 6 km h− 1 along North Creek with surveys starting during the two low (pre-dusk and pre-dawn) and two high tides (~ midday and ~ midnight) to investigate tidal and diurnal influences on pCO2 and CH4 concentrations. Water was pumped continuously from a depth of ~ 0.5 m at a flow rate of ~ 3 L min−1 using a Rule™ bilge pump to a showerhead equilibrator, and 222Rn, CO2 and CH4 were measured in a dried closed gas loop using the system described in Santos et al. (2012). 222Rn was measured at 10 minute intervals using a RAD7 radon detector. The RAD7 uses a silicon semi-conductor to count the positively charged 218Po and 214Po daughters. Dissolved 222Rn concentrations were then calculated using the temperature and salinity function of solubility (Schubert et al., 2012). pCO2, pCH4, δ13C–CO2 and δ13C–CH4 were measured from the same gas stream using a cavity ring down spectrometer (CRDS, Picarro G2201-i) at ~ 1 second intervals and then averaged over 1 minute intervals (Maher et al., 2013b). The xCO2 and xCH4 (i.e. dry CO2 and CH4 concentrations measured by the CRDS) were converted to in situ pCO2 and CH4 using standards equations (Pierrot et al., 2009) with pCH4 converted to concentrations using the solubility coefficients of Yamamoto et al. (1976). Four calibration gases (two CO2 and two CH4) were analyzed prior to each survey. CO2 standards had a concentration of 306 ppm (δ13C– CO2 = −14.4‰) and 2017 ppm (δ13C–CO2 = −17.1‰) with the δ13C values referenced using an Isotope Ratio Mass Spectrometer (IRMS) which was referenced to NIST certified standards. CH4 standards had a concentration of 3 ppm (δ13C–CH4 = − 45.9‰) and 200 ppm (δ13C– CH4 = −40.1‰). The δ13C–CH4 values of the standards were obtained by running the standards through a factory calibrated CRDS, and were later referenced against certified standards obtained from Isometric

D.T. Maher et al. / Marine Chemistry 168 (2015) 69–79

71

Fig. 1. Location of study area including discrete survey sampling locations (black circles), time series locations (white crosses) and estuary section classification (color coded) (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.).

Instruments (Tiso 1, δ13C–CH4 = −38.3 ± 0.2‰ and Liso1, δ13C–CH4 = −66.5 ± 0.2‰). Accuracy of CO2 concentrations was better than 1 ppm and δ13C–CO2 accuracy for a 5 min average was better than 0.5‰. CH4 concentration accuracy was better than 80 ppb. δ13C–CH4 accuracy for a 5 min averaging interval was better than 0.6‰ for both Tiso1 and Liso1 standards. Additional physico-chemical water quality parameters including dissolved oxygen (DO), salinity, pH, chromophoric dissolved organic

matter (CDOM, expressed as Quinine Sulfate Equivalents — QSE), and chlorophylla were measured at one minute intervals throughout the surveys using a Hydrolab DS5 sonde and Wetlabs Triplet flurometer. Instantaneous wind speed was measured using an onboard sonic anemometer (Airmar PB200). DO concentrations were converted to apparent oxygen utilization (AOU): AOU = [O2⁎] − [O2]

72

D.T. Maher et al. / Marine Chemistry 168 (2015) 69–79

Fig. 2. Solute distribution during the four estuarine surveys (D HT = day high tide, D LT = day low tide, N HT = night high tide, N LT = night low tide). A. salinity, B. apparent oxygen utilization (AOU), C. chlorophyll a, D. chromophoric dissolved organic matter (CDOM), E. 222Rn concentration, F. windspeed, G. dissolved inorganic carbon (DIC), H. alkalinity.

where [O2] is the measured DO, and [O2⁎] is the oxygen concentration at 100% saturation, calculated as a function of temperature and salinity (Benson and Krause, 1984). Discrete samples for alkalinity (TAlk) analysis were collected at 7 equidistant spaced stations along the estuary (~ every 2 km; Fig. 1) during each survey. Samples were filtered through 0.7 μm GFF filters into 30 mL polycarbonate vials, and stored at 4 °C until analysis within 2 days. TAlk was determined by Gran Titration using an automated titration system (Metrohm Titrando), using standardized 0.01 M HCl and referenced to Dickson Certified Reference Material (Batch 122). Dissolved inorganic carbon (DIC) was calculated from pCO2 and TAlk using the CO2SYS program (Lewis and Wallace, 1998) using the carbonic acid disassociation constants of Mehrbach et al. (1973) as refitted by Dickson and Millero (1987). Excess DIC is defined as the amount of DIC transferred to the atmosphere as CO2 after complete water–air equilibration (Abril et al., 2000) and was calculated as the difference between in situ DIC and equilibrium DIC both of which were calculated using CO2SYS using TAlk and pCO2, and assuming atmospheric pCO2 of 400 μatm.

2.3. Time series Three time series stations, measuring pCO2 and 222Rn and physicochemical water quality parameters were installed within the three sections of the estuary (Fig. 1) between 12.07 pm on the 14th of January and 12.42 pm on the 16th of January 2013. 222Rn was sampled at 30 minutes intervals using a RAD7 radon detector. Differential nondispersive infrared (NDIR) gas analyzers were used to measure CO2 at one-minute intervals (Licor 7000 at the downstream and midstream sites, and Licor Li-820 at the upstream site). The downstream Licor was calibrated prior to deployment using 0 ppm, 306 ppm and 2017 ppm standards (Coregas, Australia). The mid- and upper estuary Licor instruments were calibrated with 0 ppm, 2017 ppm and 10,000 ppm standards (Coregas, Australia). A submersible pump (Rule™ bilge pump) was used to deliver water to a showerhead equilibrator from which a closed air loop was continuously pumped through the Rad7 and NDIR analyzers (Santos et al., 2012). Temperature, dissolved oxygen, salinity, and pH, were also monitored at 10 minute intervals at these sites using calibrated Hydrolab DS5x sondes.

D.T. Maher et al. / Marine Chemistry 168 (2015) 69–79

73

Table 1 Key estuarine characteristics during the four spatial surveys. Fluxes are integrated over the 6 hour period centered on the survey. D HT = day high tide survey, D LT = day low tide survey, N HT = night high tide survey, N LT = night low tide survey.

D HT D LT N HT N LT

Estuary volume (m3)

Ave depth (m)

Ave k600 (m d−1)

Total CH4 (mol)

Total excess DIC (mol)

Total AOU (mol)

222

Rn inventory (dpm)

Total atmospheric CO2 flux (mol)

Total atmospheric CH4 flux (mol)

δ13C–CO2 Source (‰)

δ13C–CH4 Source (‰)

6,043,801 2,993,108 5,471,821 2,613,022

3.17 1.57 2.87 1.38

2.75 1.70 0.98 1.14

44 48 72 89

36,1219 48,8579 43,7376 52,0909

12,584 43,581 84,977 13,4646

1.81 × 109 3.01 × 109 3.33 × 109 N.D.

6139 22,151 7003 18,957

2 11 5 16

−20.44 (±0.09) −22.16 (±0.08) −20.39 (±0.12) −22.22 (±0.09)

−59.59 (±0.12) −60.08 (±0.16) −65.17 (±0.08) −63.76 (±0.17)

where k is the gas transfer velocity of CO2 or CH4 (cm h−1), α is the solubility coefficient of CO2 (Weiss, 1974) or CH4 (Yamamoto et al., 1976), and ΔC is the difference between pCO2 and pCH4 in the air and surface water (μatm). Positive numbers indicate a flux from the water to the atmosphere while negative numbers indicate atmosphere to water exchange. Gas transfer velocities of sparingly soluble gases such as CO2 and CH4 are dependent on aqueous boundary layer turbulence (Broecker and Peng, 1982). Determining gas transfer velocities when calculating CO2 and CH4 fluxes can be problematic due to the high degree of heterogeneity in environmental conditions, which is particularly true for shallow tidal systems such as estuaries (Raymond and Cole, 2001). CO2 and CH4 fluxes have been calculated using the gas transfer parameterization of Borges et al. (2004b):

This empirical relationship which was established for the Scheldt estuary was used due to the similar morphological characteristics of the Scheldt estuary and North Creek estuary (Atkins et al., 2013). It is important to note, however, that the relationship between wind speed and k600 in estuaries is likely to be site specific (Borges et al., 2004a). In situ Schmidt numbers of CO2 and CH4 were calculated using the equations of Wanninkhof (1992) from in situ temperature and salinity assuming a linear relationship of the Schmidt number and salinity between salinities of 0 and 35. To calculate estuary-wide fluxes from the spatial surveys, flux rates from each individual point were interpolated using the spline with barrier methods in ARC GIS (Maher and Eyre, 2012). To calculate estuary-wide fluxes from the time series data, we assumed that the measured pCO2 at each of the time series stations was representative of the entire section they were situated within (as defined in Fig. 1). We used wind speed data from the surveys to calculate time specific rates, as described for the surveys, and scaled the rates to the entire section based on section specific areas. The one minute fluxes from each section were integrated over the same time period that the surveys were undertaken, and divided by the total estuarine area and number of days to compare estuary-wide fluxes based on surveys and time series methods.

k600 = 5.141u0.758 10

2.5. Estuarine inventory

where k600 is the gas transfer velocity (cm h−1) normalized to a Schmidt number of 600, and u10 is the wind speed at a height of 10 m (m s−1).

Hydrographic survey data from the New South Wales Department of Environment and Heritage was used to produce a 10 m × 10 m grid

2.4. Flux calculations Air–water gas exchange rates are controlled by gas solubility, the concentration gradient between the water and air phase and the gas transfer velocity: F = kαΔC

Fig. 3. Distribution of pCO2, dissolved CH4, δ13C–CO2, and δ13C–CH4 during the four estuarine surveys.(D HT = day high tide, D LT = day low tide, N HT = night high tide, N LT = night low tide).

74

D.T. Maher et al. / Marine Chemistry 168 (2015) 69–79

resolution (± 0.1 m) digital elevation model (DEM). Estuarine CH4 loads for each survey (i.e. the total mass of dissolved CH4 within the entire estuary), were calculated by integrating an interpolated GIS raster of CH4 concentration constructed using ARC GIS, with the bathymetric DEM (corrected for tidal height during each survey). The same mass loading method was used to calculate estuarine inventory of 222Rn, AOU and excess DIC. Due to the shallow depth (average of 2.3 m) and generally narrow channel (Fig. 1), homogenous vertical and cross creek distribution in concentrations were assumed. The tidal height was assumed to be constant throughout each survey. Surveys were initiated 1 h prior to high and low tides, and surveys took ~2 h to complete. This resulted in a maximum depth variation of ~12 cm over the duration of individual surveys, which leads to an error of between 4% and 9% for the four surveys based on the average estuary depth during each survey. 2.6. Isotopic characterization of CO2 and CH4 sources To estimate the isotope value of the CH4 and CO2 sources during each survey we used the linear model developed for atmospheric CO2 isotopic source characterization in the form of: δobsCobs = δsCobs − Cbg(δbg − δs) where C and δ are concentration and δ13C, respectively, and the subscripts obs, bg and s refer to the observed, background and source values (Miller and Tans, 2003), respectively. By plotting δobsCobs (y) against Cobs (x) (Miller–Tans plots) the slope of the line is equal to δs.

There was a distinct offset in the spatial distribution of δ13C–CO2 between high tide and low tide surveys in the lower and mid-estuaries, with low tide surveys having δ13C–CO2 values ~ 2‰ depleted for a given distance from the mouth (Fig. 3C). δ13C–CO2 values converged at the most upstream location for all tides at ~−19‰. Spatial distribution of δ13C–CH4 values differed between each survey, with least depleted values during the D HT survey and most depleted values in the lower half of the estuary during the N LT survey. (Fig. 3D). During the high tide surveys δ13C–CH4 generally showed an upstream depletion in δ13C value, while during the low tide surveys, δ13C–CH4, values decreased from the lower to mid-estuary then increased from the mid to upper estuary. Salinity was relatively constant at the lower estuary time series station (Fig. 4A), and displayed a clear tidal trend at the mid estuary (Fig. 4B) and upper estuary stations (Fig. 4C). Time series station pCO2 and 222Rn generally followed a tidal trend at all three time series stations. pCO2 ranged from ~ 500–1200 μatm, 1400–12,000 μatm and 4000–6000 μatm at the lower (Fig. 4A), mid (Fig. 4B) and upper estuary (Fig. 4C) stations, respectively. Highest 222Rn concentrations (up to ~4200 dpm m−3) were measured in the mid-estuary (Fig. 4B) and the lowest concentrations were measured in the lower estuary (Fig. 4A). The highest 222Rn concentrations and pCO2 values at the mid estuary station exceeded the highest survey values by ~2 fold. There was a distinct difference in overall estimated diffusive CO2 and CH4 fluxes between high and low tides. The integrated estuary-wide CO2 flux during the D HT survey was 19.32 mmol CO2 m−2 d−1 while the D LT was 69.70 mmol CO2 m−2 d−1 (Fig. 5). The difference between

3. Results Salinity ranged between ~ 36 and 10 throughout the study and was similar during the day high tide (D HT) survey and night high tide (N HT) survey (Fig. 2A). Both low tide surveys also had similar salinity distribution (Fig. 2A). AOU ranged from ~− 5 to 110 μmol kg−1 (Fig. 2B), and showed a general upstream increase during each survey. During both high tides, AOU increased to ~ 110 μmol kg−1 within the mid estuary and remained at this level in the upper estuary. During both low tides, AOU increased from the lower estuary to the midestuary and then decreased in the upper estuary where AOU declined to ~10 μmol kg−1 during the day low tide (D LT). Chlorophyll a concentrations increased in the upstream direction under all tidal regimes and during both low tide surveys chlorophyll a was ~2 fold higher within the upper estuary (Fig. 2C). CDOM had a general upstream increase in concentrations however there were distinctively higher mid estuary concentrations, and lower upper estuary concentrations during each of the low tide surveys (Fig. 2D). 222Rn showed a general upstream increase, however the D LT survey shows a mid-estuary maximum, with a decrease upstream (Fig. 2E). Due to instrument failure, no 222 Rn data is available for the night low tide (N LT) survey. Wind speed was generally higher during the two day time surveys and generally lower in the upstream direction (Fig. 2F). Alkalinity and DIC ranged from 590–2583 μmol kg−1 and 727–2499 μmol kg−1 respectively, and showed a similar trend along the estuary (Fig. 2G, H). pCO2 showed a general upstream increase. However during both low tides, pCO2 showed a decrease in the upstream direction of the upper estuary (Fig. 3A). Night time pCO2 was generally higher for each tide than the respective day time pCO2, particularly within the mid estuary. pCO2 in the upper estuary during both high tides were similar (~6000 μatm). CH4 concentrations were highest during the N LT survey, with concentrations reaching the highest concentration at 12 km upstream at ~ 150 nmol before decreasing further upstream (Fig. 3B). Concentrations during the N HT were higher than during the D HT. Except for the N LT, CH4 concentrations showed a general increase in the upstream direction, with the highest concentration recorded in the upper estuary.

Fig. 4. Temporal variability of salinity, pCO2 and 222Rn in the three time series stations. A. lower estuary, B. mid-estuary and C. upper estuary sites. The * on each graph denotes the time when surveys passed the time series sites, gray shaded areas represent night time.

D.T. Maher et al. / Marine Chemistry 168 (2015) 69–79

Fig. 5. Integrated average estuary-wide water to air fluxes of CO2 and CH4 for the four surveys. D HT = day high tide, D LT = day low tide, N HT = night high tide, N LT = night low tide.

N HT and N LT showed a similar trend with a 2.7 fold increase in CO2 flux at low tide. Estuary-wide fluxes of CH4 showed an even greater difference between tides with the integrated estuary D HT flux of 7 μmol CH4 m− 2 d− 1 while the D LT flux was 36 μmol CH4 m− 2 d− 1 (Fig. 5). CH4 flux during the N HT survey of 17 μmol CH4 m−2 d−1 and the N LT flux was 51 μmol CH4 m−2 d−1CH4. The estuarine inventory showed distinct differences in excess DIC, AOU, CH4 and 222Rn between each survey (Table 1). Despite a smaller volume of water in the estuary during low tide, inventories were generally higher for all parameters during the low tide surveys. Night time AOU, excess DIC, CH4 and 222Rn inventories were higher than those during the day time for the same tide. The δ13C–CH4 value of the source of CH4 to the estuary ranged from − 59.56‰ to − 65.17‰ over the four surveys, with values more depleted during the night time surveys (Table 1). δ13C–CO2 source values showed no diel difference, but there was a distinct tidal difference with D HT and D LT values of −20.44‰ and − 22.16‰ respectively and N HT and N LT values of − 20.39‰ and − 22.22‰ respectively (Table 1). The estuary integrated water to air flux for each tidal regime (6 hour period) varied from 2 to 16 mol for CH4 and 6139 to 22151 mol of CO2 (Table 1). 4. Discussion Estuaries are important in terms of the global carbon cycle. However, there are few studies on diurnal and tidal variation of CO2 and CH4 concentrations and water to air fluxes at the whole estuary scale. This is likely due to the challenging logistics required. Undertaking four surveys over a 24 hour period, while at the same time having three fixed time series stations along the length of the estuary may preclude our strategy from being used in many estuaries. However, the insights gained from this study are likely to be relevant to other shallow productive estuaries, and/or estuaries influenced by groundwater inputs, and may help guide future sampling strategies. There were clear tidal and diurnal differences in CO2 and CH4 concentrations, δ13C values and estimated diffusive water to air fluxes. Further, there were differences in the estimates of whole estuary water to air CO2 fluxes based on fixed time series and survey methods. Diurnal variability in pCO2 and CH4 can be attributed to four processes; 1) temperature driven differences in gas solubility/thermodynamics, 2) convection in the water column and/or sediment water interface, 3) diel primary production/respiration and methane production/ oxidation cycles and 4) diurnal variability in wind speed and associated evasion rates. Wind speed was generally higher during the day in the lower estuary (Fig. 2F). However in the mid and upper estuaries (where pCO2 and CH4 where highest), wind speeds where generally similar for the D LT, N HT and N LT surveys, although somewhat higher for the D HT survey. Average k600 values based on the wind speed parameterization of Borges et al. (2004b) for each survey were 2.75,

75

1.70, 0.98 and 1.14 m d−1 for the D HT, D LT, N HT and N LT surveys respectively. This equates to a loss of 22, 28, 9 and 21% of the CO2 and CH4 inventories in excess of atmospheric equilibrium for the 6 hour period between the surveys for the D HT, D LT, N HT and N LT surveys, respectively. These differences could explain some of the diurnal variability in surface water pCO2 and CH4 through increased evasion during the day when winds are usually stronger. While we have only used wind speed to determine k600, studies have found that currents in shallow estuarine systems can contribute significantly to surface turbulence and therefore gas transfer velocity (Zappa et al., 2003; Borges et al., 2004b; Ho et al., 2014). For example, Ho et al. (2014) found that by neglecting to include bottom generated turbulence, k600 was underestimated by between 40 and 70% in the Shark River estuary. Atkins et al. (2013) found that current driven evasion dominated water to air fluxes of CO2 and 222Rn in the upper North Creek estuary. Current driven surface turbulence is controlled by velocity, depth and bottom roughness (O'connor and Dobbins, 1958; Zappa et al., 2003), and therefore would not be subjected to diurnal variability. As such, it is likely that the differences associated with diurnal wind changes are overestimated, due to the lack of current driven surface turbulence parameterization. Further, integrated estuary-wide water to air fluxes of CO2 and CH4 showed the opposite diurnal trend to k600 values, with higher or similar estuary integrated fluxes at night (Table 1). Temperature affects pCO2 through changes in solubility and thermodynamic shifts in the carbonate system, with warmer water having a higher partial pressure of the gas at the same concentration (Takahashi et al., 1993). Previous studies have found that temperature can drive diurnal changes in pCO2 in the open ocean (Goyet and Peltzer, 1997; Bates et al., 1998) and oligotrophic areas of the South China Sea (Dai et al., 2009), However, diurnal temperature variation (i.e. warmer water temperatures during the day), would lead to higher pCO2 during the day and lower pCO2 during the night, the opposite to what was found during this study. Therefore we can exclude temperature driven changes in solubility as the source of diurnal variability in surface water pCO2 dynamics during our study. Diurnal primary production and respiration cycles lead to a lower pCO2 during the day, driven by CO2 uptake by primary producers, and higher pCO2 during the night through in situ respiration. The effect of ecosystem metabolism on estuarine CH4 dynamics is less clear. However, enhanced oxygen production during the day, and depletion during the night may lead to oscillations in sediment oxygen penetration, influencing CH4 production and oxidation rates (King et al., 1990; Maher et al., 2013b). Ecosystem primary production and respiration has been found to control pCO2 diurnal variability in the coastal waters of Tampa Bay (Yates et al., 2007), coral reefs (Gattuso et al., 1993; Cyronak et al., 2014) and in macrophyte beds (Saderne et al., 2013). The tight coupling between pCO2 and AOU in the mid- and lower estuaries (i.e. AOU b ~90 μmol kg−1; Fig. 6C), suggests that in situ biological processes exerts some control over pCO2 in this section of the estuary. In addition, a significant relationship between pCO2 and CDOM, and also CH4 and CDOM (Fig. 6H) may be indicative of dissolved organic matter (and associated respiration and/or photodegradation) contributing to in situ pCO2 and CH4 dynamics. Interestingly in the upper estuary, there was a sharp decline in AOU (Fig. 2B) during the two low tide surveys, and the pCO2–AOU relationship displayed discontinuity, with AOU decreasing but pCO2 remaining relatively constant (Fig. 6C). Further, highest chlorophyll a concentrations throughout the study occur at the upstream limit during both of the low tide surveys (Fig. 2C). This suggests that primary production is higher in the upper estuary beyond the boundary of the study area, and during low tide this water flows into the upper estuary section of the study area. This water is high in oxygen (i.e. low AOU), however pCO2 remains relatively high. This indicates that a process other than in situ primary production and respiration are driving pCO2 dynamics throughout some parts of the estuary. While nitrification can be a

76

D.T. Maher et al. / Marine Chemistry 168 (2015) 69–79

significant oxygen consumption process in eutrophic estuaries (Gazeau et al., 2005) and through H+ production, nitrification concomitantly increases pCO2, (Frankignoulle et al., 1996), our previous investigations in North Creek have found relatively low DIN concentrations (up to 4 μM, unpublished data) which suggests this may not be a major driver of AOU or pCO2 in North Creek. In addition to diurnal variability, during our study, pCO2, CH4, δ13C– CO2 and δ13C–CH4 distribution had marked tidal differences, particularly in the mid-estuary surrounded by mangroves (Fig. 3). There were also clear differences in the δ13C source values and estuarine inventories of CO2 and CH4 (Table 1). Few studies have assessed the spatial dynamics of pCO2 over different tides, and to our knowledge there have been no studies on the tidal influence on in situ CH4 concentration, δ13C–CO2 and δ13C–CH4 distribution along an estuary. Jiang et al. (2008) found average annual low tide estuary integrated pCO2 was ~ 50% higher

than high tide values for 3 estuaries along the Georgia Coast (USA). Wang and Cai (2004) also found differences between high and low tide pCO2 in the nearby Duplin River (Georgia USA), with the greatest difference occurring during the summer. Tidal variability in estuarine CH4 concentrations have been previously assessed at fixed time series stations (Dulaiova et al., 2006; Bouillon et al., 2007; Deborde et al., 2010; Linto et al., 2014) which have found tidal variability in concentrations with higher concentrations during low tide, as found during our study. Tidal differences in surface water pCO2 and CH4 concentrations are driven by a number of processes, including inputs and mixing of freshwater and oceanic water, inputs of CO2 and CH4 enriched groundwater, lateral inputs from marshes, mangroves and intertidal areas, and stimulation of in situ primary production and/or respiration by inputs of organic matter and/or nutrients from any of these processes. There was

Fig. 6. Relationship between: 222Rn and pCO2 (A.), 222Rn and CH4 (B.), AOU and pCO2 (C.), AOU and CH4 (D.), Chlorophyll a and pCO2 (E.), Chlorophyll a and CH4 (F.), CDOM and pCO2 (G.), CDOM and CH4 (H.) during the four estuarine surveys.

D.T. Maher et al. / Marine Chemistry 168 (2015) 69–79

a significant positive exponential relationship between 222Rn (an unambiguous groundwater tracer in North Creek, Atkins et al., 2013) and both pCO2 (all surveys, R2 = 0.88 n = 53, p b 0.001) and CH4 (all surveys, R2 = 0.88, n = 53, p b 0.001) (Fig. 6A and B) indicating that groundwater/porewater exchange is closely coupled to surface water pCO2 and CH4 dynamics (see below). There was also a significant positive linear relationship between CDOM and 222Rn (R2 = 0.82, n = 53, p b 0.001), which may simply be a covariance between groundwater input and CDOM, or, CDOM may be derived from groundwater. The relationship between CDOM and both CH4 and pCO2 described earlier, may therefore be due to groundwater derived inputs of dissolved organic matter as observed in other systems (Santos et al., 2009; Kim et al., 2012; Maher et al., 2013a). Temperature driven convection of porewater–surface water exchange may explain diurnal variability in pCO2 and CH4 dynamics. Rocha (2000) found that temperature driven convection of porewater in permeable intertidal estuarine sediments may lead to a rapid exchange of porewater with surface waters, and the process can be orders of magnitude more important than diffusive fluxes. As porewaters are highly enriched in pCO2 and CH4 (Cable et al., 1996; Atkins et al., 2013), diurnal temperature driven porewater convection could explain the diurnal patterns found during this study. Interestingly, 222Rn also showed diurnal variability, with concentrations and inventory higher during the N HT survey than the D HT survey (Fig. 2E, Table 1), which implies increased groundwater/porewater inputs during the night. High night time 222Rn is consistent with observations in a coral reef lagoon, where the faster cooling of surface waters than underlying porewater at night was suggested to create a density inversion at the sediment-water interface that could explain a 15% increase in night time groundwater discharge (Santos et al., 2011). Further, the estuarine inventory of CH4, excess DIC, 222Rn and the total estuarine AOU show higher values during the night than during the day (Table 1). While primary production and respiration could explain diurnal CH4 and pCO2 dynamics, clearly the increased 222Rn inventory during the N HT survey is not linked to biological processes, and implies that temperature driven sediment porewater convection may be a local driver of surface water chemistry. Atkins et al. (2013) found that groundwater was the major driver of surface water pCO2 in the upper section of the North Creek estuary where agricultural drains cut through the shallow aquifer. In this study there was a strong relationship between pCO2 and CH4 and 222 Rn throughout the first three surveys (Fig. 6A and B; 222Rn unavailable during the fourth survey), suggesting that surface water pCO2 and CH4 is coupled to groundwater/porewater inputs either directly (i.e. inputs of CO2 and CH4 enriched groundwater) or indirectly (e.g. through groundwater inputs fuelling ecosystem metabolic responses that in turn influence surface water pCO2 and CH4). Whether CH4 is oxidized or fluxed to the atmosphere from an estuary has important implications considering CH4 has a global warming potential (GWP) ~100 times greater than CO2 (the product of CH4 oxidation) over a 20 year time frame (Shindell et al., 2009). The importance of oxidation as a CH4 sink varies considerably among estuaries, and even within an estuary. For example Sansone et al. (1999) found evidence of CH4 oxidation in the Columbia River estuary (USA), but not in four other estuaries and Upstill-Goddard et al. (2000) found that CH4 oxidation rates were an order of magnitude lower than atmospheric flux in North Sea estuaries. Abril and Iversen (2002) found high rates of CH4 oxidation at a brackish water station, but minimal rates in the lower estuary of Randers Fjord (Denmark) and Abril et al. (2007) found enhanced CH4 oxidation at the estuarine turbidity maximum zone in the Gironde Estuary in France. The survey δ13C–CH4 values during this study suggest that minimal oxidation of CH4 occurred along the estuary (Fig. 3D). During oxidation, δ13C–CH4 values become more positive due to preferential oxidation of the more negative (“lighter”) CH4 pool by methanotrophs (Whiticar, 1999). However, lower estuary δ13C–CH4 are lower than atmospheric equilibrium (approximately −47‰; Quay

77

et al., 1999) and the δ13C–CH4 source values (Table 1) are in the range expected for the acetate reduction pathway (~ − 65 to − 50‰; Whiticar et al., 1986). Integrated estuary-wide average fluxes for the surveys were 19 to 70 mmol m−2 d−1 for CO2 and 7 to 51 μmol m−2 d−1 for CH4 (Fig. 5). These fluxes fall within the range of fluxes reported for estuaries (Borges and Abril, 2011). Typically, estuarine pCO2 and CH4 are surveyed once (most likely during the day), and fluxes are based on these single surveys. Alternatively, pCO2 and CH4 may be measured at a fixed time series station, with flux estimates scaled up to the surrounding area based on those single location concentrations (e.g. Bouillon et al., 2007; Atkins et al., 2013). Using either of these methods in isolation, and not accounting for diurnal, tidal and spatial variability may lead to significant under or over estimation of total water to air fluxes. While the "four survey" method presented here may seem most likely to provide accurate estimates, a comparison with fluxes based on the time series data suggest that this may not be the case for all sections of the estuary (Fig. 7). The large discrepancy between the CO2 flux estimates from the time series and survey data for the mid estuarine section was due to the surveys not passing through this section of the estuary during the times of highest pCO2 (see * locations in Fig. 4). This section of the estuary had a pCO2 range of 1400 to 12,000 μatm, however the highest pCO2 measured during the survey was only ~6000 μatm. The peak in pCO2 during the time series occurred 2–3 hours after low tide, after the survey had passed through that section of the estuary. The same is observed for 222Rn, where the highest survey concentrations were ~ 2200 dpm m−3, but at the mid-estuary time series station concentrations reached ~4000 dpm m−3. Therefore a combination of time series and spatial surveys may be required to adequately constrain the variability in estuarine pCO2 and CH4 fluxes. 5. Conclusion While it was beyond the scope of this study to quantitatively assess all the drivers of diurnal and tidal variability in surface water pCO2 and CH4 dynamics, insights from the high temporal and spatial resolution time series and survey experiments suggest that a complex interaction of biological (productivity/respiration), hydrological (groundwater inputs and mixing of riverine and oceanic waters) and physical (temperature driven porewater/groundwater convection and wind driven evasion) processes occurred. One of the major limitations in scaling up estuarine water to air greenhouse gas fluxes to the global scale is the paucity of information on diurnal dynamics (Cai, 2011). In this study we show that this limitation should not be taken lightly, at least for small shallow productive estuaries that receive some groundwater input. By neglecting to account for tidal and diurnal differences in surface water pCO2 and CH4, flux estimates can be significantly over or

Fig. 7. Comparison of water to air fluxes of CO2 estimated from the four spatial surveys with estimates from the three fixed time series locations.

78

D.T. Maher et al. / Marine Chemistry 168 (2015) 69–79

under estimated. Research into whether similar diurnal and tidal differences occur in larger estuaries needs to be undertaken to decrease uncertainty in estuarine CO2 and CH4 fluxes.

Acknowledgments We acknowledge funding from the Australian Research Council (DP120101645, LE120100156, LP110200975, LP100200732, and DE150100581). KC acknowledges the SCU GeoScience Summer School program for her participation in this study. DTM is funded through an ARC DECRA Fellowship (DE150100581).

References Abril, G., Iversen, N., 2002. Methane dynamics in a shallow non-tidal estuary (Randers Fjord, Denmark). Mar. Ecol. Prog. Ser. 230, 171–181. Abril, G., Etcheber, H., Borges, A., Frankignoulle, M., 2000. Excess atmospheric carbon dioxide transported by rivers into the Scheldt estuary. Earth Planet. Sci. 330, 761–768. Abril, G., Commarieu, M.-V., Guérin, F., 2007. Enhanced methane oxidation in an estuarine turbidity maximum. Limnol. Oceanogr. 52 (1), 470–475. Atkins, M.L., Santos, I.R., Ruiz-Halpern, S., Maher, D.T., 2013. Carbon dioxide dynamics driven by groundwater discharge in a coastal floodplain creek. J. Hydrol. 493, 30–42. Bass, A.M., Munksgaard, N.C., Leblanc, M., Tweed, S., Bird, M.I., 2013. Contrasting carbon export dynamics of human impacted and pristine tropical catchments in response to a short-lived discharge event. Hydrol. Process. 28, 1835–1843. Bates, N.R., Takahashi, T., Chipman, D.W., Knap, A.H., 1998. Variability of pCO2 on diel to seasonal timescales in the Sargasso Sea near Bermuda. J. Geophys. Res. 103 (C8), 15,567–15,585. Bauer, J.E., Bianchi, T.S., 2011. Dissolved organic carbon\ cycling and transformation. In: Wolanski, E., McLusky, D.S. (Eds.), Treatise on Estuarine and Coastal Science. Academic Press, pp. 7–67. Benson, B.B., Krause, D., 1984. The concentration and isotopic fractionation of oxygen dissolved in freshwater and seawater in equilibrium with the atmosphere. Limnol. Oceanogr. 29, 620–632. Borges, A.V., 2005. Do we have enough pieces of the jigsaw to integrate CO2 fluxes in the coastal ocean? Estuaries 28, 1–25. Borges, A.V., Abril, G., 2011. Carbon dioxide and methane dynamics in estuaries. In: Wolanski, E., McLusky, D.S. (Eds.), Treatise on estuarine and coastal science. Academic Press, pp. 119–161. Borges, A.V., Djenidi, S., Lacroix, G., Théate, J., Delille, B., Frankignoulle, M., 2003. Atmospheric CO2 flux from mangrove surrounding waters. Geophys. Res. Lett. 30. Borges, A.V., Delille, B., Schiettecatte, L.-S., Gazeau, E., Abril, G., Frankignoulle, M., 2004a. Gas transfer velocities of CO2 in three European estuaries (Randers Fjord, Scheldt, and Thames). Limnol. Oceanogr. 49, 1630–1641. Borges, A.V., Vanderborght, J.-P., Schiettecatte, L.-S., Gazeau, F., Ferrón-Smith, S., Delille, B., Frankignoulle, M., 2004b. Variability of the gas transfer velocity of CO2 in a macrotidal estuary (The Scheldt). Estuaries 27, 595–605. Borges, A.V., Delille, B., Frankignoulle, M., 2005. Budgeting sinks and sources of CO2 in the coastal ocean: diversity of ecosystem counts. Geophys. Res. Lett. 32. Bouillon, S., Middelburg, J.J., Dehairs, F., Borges, A.V., Abril, G., Flindt, M.R., Ulomi, S., Kristensen, E., 2007. Importance of intertidal sediment processes and porewater exchange on the water column biogeochemistry in a pristine mangrove creek (Ras Dege, Tanzania). Biogeosciences 4, 311–322. Broecker, W.S., Peng, T.-H., 1982. Tracers in the sea. Eldigio Press. Cable, J.E., Bugna, G.C., Burnett, W., Chanton, J.P., 1996. Application of 222Rn and CH4 for assessment of groundwater discharge to the coastal ocean. Limnol. Oceanogr. 41, 1347–1353. Cai, W.-J., 2011. Estuarine and coastal ocean carbon paradox; CO2 sinks or sites of terrestrial carbon incineration. Ann. Rev. Mar. Sci. 3, 123–145. Charbonnier, C., Anschutz, P., Poirier, D., Bujan, S., Lecroart, P., 2013. Aerobic respiration in a high-energy sandy beach. Mar. Chem. 155, 10–21. Chen, C.-T.A., Zhai, W., Dai, M., 2008. Riverine input and air–sea CO2 exchanges near the Changjiang (Yangtze River) Estuary: status quo and implication on possible future changes in metabolic status. Cont. Shelf Res. 28, 1476–1482. Cyronak, T.J., Santos, I.R., Erler, D.V., Maher, D.T., Eyre, B.D., 2014. Drivers of pCO2 variability in two contrasting coral reef lagoons: the influence of submarine groundwater discharge. Glob. Biogeochem. Cycles 28, 398–414. Dai, M., Lu, Z., Zhai, W., Chen, B., Cao, Z., Zhou, K., Cai, W.-J., Chen, C.-T.A., 2009. Diurnal variations of surface seawater pCO2 in contrasting coastal environments. Limnol. Oceanogr. 54, 735–745. Deborde, J., Anschutz, P., Guérin, F., Poirier, D., Marty, D., Boucher, G., Thouzeau, G., Canton, M., Abril, G., 2010. Methane sources, sinks and fluxes in a temperate tidal lagoon: the Arcachon lagoon (SW France). Estuar. Coast. Shelf Sci. 89, 256–266. Dickson, A.G., Millero, F.J., 1987. A comparison of the equilibrium constants for the dissociation of carbonic acid in seawater media. Deep-Sea Res. Part A Oceanogr. Res. Pap. 34, 1733–1743. Dorsett, A., Cherrier, J., Martin, J.B., Cable, J.E., 2011. Assessing hydrologic and biogeochemical controls on pore-water dissolved inorganic carbon cycling in a subterranean estuary: a 14C and 13C mass balance approach. Mar. Chem. 127 (1–4), 76–89.

Dulaiova, H., Peterson, R., Burnett, W., Lane-Smith, D., 2005. A multi-detector continuous monitor for assessment of 222Rn in the coastal ocean. J. Radioanal. Nucl. Chem. 263, 361–365. Dulaiova, H., Burnett, W.C., Chanton, J.P., Moore, W.S., Bokuniewicz, H.J., Charette, M.A., Sholkovitz, E., 2006. Assessment of groundwater discharges into West Neck Bay, New York, via natural tracers. Cont. Shelf Res. 26, 1971–1983. Eyre, B.D., Pont, D., 2003. Intra- and inter-annual variability in the different forms of diffuse nitrogen and phosphorus delivered to seven sub-tropical east Australian estuaries. Estuar. Coast. Shelf Sci. 57, 137. Frankignoulle, M., Bourge, I., Wollast, R., 1996. Atmospheric CO2 fluxes in a highly polluted estuary (the Scheldt). Limnol. Oceanogr. 41 (2), 365–369. Gatland, J.R., Santos, I.R., Maher, D.T., Duncan, T.M., Erler, D.V., 2014. Carbon dioxide and methane emissions from an artificially drained coastal wetland during a flood: implications for wetland global warming potential. J. Geophys. Res. Biogeosci. 19 (8), 1698–1716. Gattuso, J.-P., Pichon, M., Delesalle, B., Frankignoulle, M., 1993. Community metabolism and air–sea CO2 fluxes in a coral reef ecosystem (Moorea, French Polynesia). Mar. Ecol. Prog. Ser. 96, 259–267. Gazeau, F., Gattuso, J.P., Middelburg, J.J., Brion, N., Schiettecatte, L.S., Frankignoulle, M., Borges, A.V., 2005. Planktonic and whole system metabolism in a nutrient-rich estuary (the Scheldt estuary). Estuaries 28 (6), 868–883. Gleeson, J., Santos, I.R., Maher, D.T., Golsby-Smith, L., 2013. Groundwater–surface water exchange in a mangrove tidal creek: evidence from natural geochemical tracers and implications for nutrient budgets. Mar. Chem. 156, 27–37. Goyet, C., Peltzer, E.T., 1997. Variation of CO2 partial pressure in surface seawater in the equatorial Pacific Ocean. Deep-Sea Res. I 44, 1611–1625. Ho, D.T., Ferrón, S., Engel, V.C., Larsen, L.G., Barr, J.G., 2014. Air–water gas exchange and CO2 flux in a mangrove-dominated estuary. Geophys. Res. Lett. 41, 108–113. IPCC (Intergovernmental Panel on Climate Change), 2007. Climate change 2007: The Physical Science Basis, Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L. (Eds.), Cambridge University Press, Cambridge United Kingdom and New York, NY, USA (996 pp.). Jiang, L.-Q., Cai, W.-J., Wang, Y., 2008. A comparative study of carbon dioxide degassing in river- and marine-dominated estuaries. Limnol. Oceanogr. 53, 2603–2615. Karl, D.M., Beversdorf, L., Björkman, K.M., Church, M.J., Martinez, A., Delong, E.F., 2008. Aerobic production of methane in the sea. Nat. Geosci. 1, 473–478. Kemp, W.M., Smith, E.M., Marvin-Dipasquale, M., Boynton, W.R., 1997. Organic carbon balance and net ecosystem metabolism in Chesapeake Bay. Mar. Ecol. Prog. Ser. 150, 229–248. Kim, T.-H., Waska, H., Kwon, E., Suryaputra, I.G.N., Kim, G., 2012. Production, degradation, and flux of dissolved organic matter in the subterranean estuary of a large tidal flat. Mar. Chem. 142–144, 1–10. King, L., 2003. Ballina nature reserve plan of management. In: National Parks and Wildlife Service (Ed.). King, G.M., Roslev, P., Skovgaard, H., 1990. Distribution and rate of methane oxidation in sediments of the Florida Everglades. Appl. Environ. Microbiol. 56, 2902–2911. Kristensen, E., Flindt, M.R., Ulomi, S., Borges, A.V., Abril, G., Bouillon, S., 2008. Emission of CO2 and CH4 to the atmosphere by sediments and open waters in two Tanzanian mangrove forests. Mar. Ecol. Prog. Ser. 370, 53–67. Laruelle, G.G., Dürr, H.H., Slomp, C.P., Borges, A.V., 2010. Evaluation of sinks and sources of CO2 in the global ocean using spatially-explicit typology of estuaries and shelves. Geophys. Res. Lett. 37, L15607. Lewis, E., Wallace, D.W.R., 1998. Program developed for CO2 system calculations. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, US Department of Energy. Lin, G., Ehleringer, J.R., 1997. Carbon isotopic fractionation does not occur during dark respiration in C3 and C4 plants. Plant Physiol. 114, 391–394. Linto, N., Barnes, J., Ramachandran, R., Divia, J., Ramachandran, P., Upstill-Goddard, R.C., 2014. Carbon dioxide and methane emissions from mangrove-associated waters of the Andaman Islands, Bay of Bengal. Estuar. Coasts 37, 381–398. Macklin, P.A., Maher, D.T., Santos, I.R., 2014. Estuarine canal estate waters: Hotspots of CO2 outgassing driven by enhanced groundwater discharge? Mar. Chem. 167, 82–92. Maher, D., Eyre, B.D., 2012. Carbon budgets for three autotrophic Australian estuaries: implications for global estimates of the coastal air–water CO2 flux. Glob. Biogeochem. Cycles 26, GB1032. Maher, D.T., Santos, I.R., Golsby-Smith, L., Gleeson, J., Eyre, B.D., 2013a. Groundwaterderived dissolved inorganic and organic carbon exports from a mangrove tidal creek: the missing mangrove carbon sink? Limnol. Oceanogr. 58, 475–488. Maher, D.T., Santos, I.R., Leuven, J.R.F.W., Oakes, J.M., E. D. V., Eyre, B.D., 2013b. Novel use of cavity ring-down spectrometry to investigate aquatic carbon cycling from microbial to ecosystem scales. Environ. Sci. Technol. 47, 12938–12945. Makings, U., Santos, I.R., Maher, D.T., Golsby-Smith, L., Eyre, B.D., 2014. Importance of budgets for estimating the input of groundwater-derived nutrients to an eutrophic tidal river and estuary. Estuar. Coast. Shelf Sci. 143, 65–76. McGlathery, K.J., Anderson, I.C., Tyler, A.C., 2001. Magnitude and variability of benthic and pelagic metabolism in a temperate coastal lagoon. Mar. Ecol. Prog. Ser. 216, 1–15. Mehrbach, C., Culberson, C.H., Hawley, J.E., Pytkowicz, R.N., 1973. Measurement of the apparent dissociation constants of carbonic acid in seawater at atmospheric pressure. Limnol. Oceanogr. 18, 897–907. Middelburg, J.J., Herman, P.M.J., 2007. Organic matter processing in tidal estuaries. Mar. Chem. 106 (1–2), 127–147. Middelburg, J.J., Nieuwenhuize, J., Iversen, N., Nøgh, N., De Wilde, H., Helder, W., Seifert, R., Christof, O., 2002. Methane distribution in European tidal estuaries. Biogeochemistry 59, 95–119.

D.T. Maher et al. / Marine Chemistry 168 (2015) 69–79 Miller, J.B., Tans, P.P., 2003. Calculating isotopic fractionation from atmospheric measurements at various scales. Tellus 55B, 207–214. Murase, J., Sugimoto, A., 2005. Inhibitory effect of light on methane oxidation in the pelagic water column of a mesotrophic lake (Lake Biwa, Japan). Limnol. Oceanogr. 50, 1339–1343. O'connor, D., Dobbins, W., 1958. Mechanism of reaeration in natural streams. Trans. Am. Soc. Civ. Eng. 123, 641–684. Oritz-Llorente, M.J., Alvarez-Cobelas, M., 2012. Comparison of biogenic methane emissions from unmanaged estuaries, lakes, oceans, rivers and wetlands. Atmos. Environ. 59, 328–337. Peterson, R.N., Santos, I.R., Burnett, W.C., 2010. Evaluating groundwater discharge to tidal rivers based on a Rn-222 time-series approach. Estuar. Coast. Shelf Sci. 86, 165–178. Pierrot, D., Neill, C., Sullivan, K., Castle, R., Wanninkhof, R., Lüger, H., Johannessen, T., Olsen, A.B., Feely, R.A., Cosca, C.E., 2009. Recommendations for autonomous underway pCO2 measuring systems and data-reduction routines. Deep-Sea Res. II 56, 512–522. Quay, P., Stutsman, J., Wilbur, D., Snover, A., Dlugokencky, E., Brown, T., 1999. The isotopic composition of atmospheric methane. Glob. Biogeochem. Cycles 13, 445–461. Raymond, P.A., Cole, J.J., 2001. Gas exchange in rivers and estuaries: choosing a gas transfer velocity. Estuaries 24, 312–317. Rocha, C., 2000. Density-driven convection during flooding of warm, permeable intertidal sediments: the ecological importance of the convective turnover pump. J. Sea Res. 43, 1–14. Saderne, V., Fietzek, P., Herman, P.M.J., 2013. Extreme variations of pCO2 and pH in a macrophyte meadow of the Baltic Sea in summer: evidence of the effect of photosynthesis and local upwelling. PLoS ONE 8, e62689. Sansone, F.J., Holmes, M.E., Popp, B.N., 1999. Methane stable isotopic ratios and concentrations as indicators of methane dynamics in estuaries. Glob. Biogeochem. Cycles 13, 463–474. Santos, I.R., Burnett, W., Chanton, J., Mwashote, B., Suryaputra, I.G.N.A., Dittmar, T., 2008. Nutrient biogeochemistry in a Gulf of Mexico subterranean estuary and groundwater-derived fluxes to the coastal ocean. Limnol. Oceanogr. 53, 705–718. Santos, I.R., Burnett, W.C., Dittmar, T., Suryaputra, I.G.N.A., Chanton, J., 2009. Tidal pumping drives nutrient and dissolved organic matter dynamics in a Gulf of Mexico subterranean estuary. Geochim. Cosmochim. Acta 73, 1325–1339. Santos, I.R., Glud, R.N., Maher, D., Erler, D., Eyre, B.D., 2011. Diel coral reef acidification driven by porewater advection in permeable carbonate sands, Heron Island, Great Barrier Reef. Geophys. Res. Lett. 38.

79

Santos, I.R., Maher, D.T., Eyre, B.D., 2012. Coupling automated radon and carbon dioxide measurements in coastal waters. Environ. Sci. Technol. 46, 7685–7691. Santos, I.R., De Weys, J., Tait, D.R., Eyre, B.D., 2013. The contribution of groundwater discharge to nutrient exports from a coastal catchment: post-flood seepage increases estuarine N?P ratios. Estuar. Coasts 36, 56–73. Schubert, M., Paschke, A., Lieberman, E., Burnett, W.C., 2012. Air–water partitioning of 222 Rn and its dependence on water temperature and salinity. Environ. Sci. Technol. 46, 3905–3911. Shindell, D.T., Faluvegi, G., Koch, D.M., Schmidt, G.A., Unger, N., Bauer, S.E., 2009. Improved attribution of climate forcing to emissions. Science 326, 716–718. Takahashi, T., Olafsson, J., Goddard, J.G., Chipman, D.W., Sutherland, S.C., 1993. Seasonal variation of CO2. Glob. Biogeochem. Cycles 7, 843–878. Tang, K.M., Mcginnis, D.F., Frindte, K., Brüchert, V., Grossart, H.-P., 2014. Paradox reconsidered: methane oversaturation in well-oxygenated lake waters. Limnol. Oceanogr. 59, 275–284. Upstill-Goddard, R.C., Barnes, J., Frost, T., Punshon, S., Owens, N.J.P., 2000. Methane in the southern North Sea: low-salinity inputs, estuarine removal, and atmospheric flux. Glob. Biogeochem. Cycles 14, 1205–1217. Wang, Z.A., Cai, W.-J., 2004. Carbon dioxide degassing and inorganic carbon export from a marsh-dominated estuary (the Duplin River): a marsh CO2 pump. Limnol. Oceanogr. 49, 341–354. Wanninkhof, R., 1992. Relationship between wind speed and gas exchange over the ocean. J. Geophys. Res. 97, 7373–7382. Weiss, R.F., 1974. Carbon dioxide in water and seawater: the solubility of a non-ideal gas. Mar. Chem. 2, 203–215. Whiticar, M.J., 1999. Carbon and hydrogen isotope systematics of bacterial formation and oxidation of methane. Chem. Geol. 161, 291–314. Whiticar, M.J., Faber, E., Schoell, M., 1986. Biogenic methane formation in marine and freshwater environments: CO2 reduction vs. acetate fermentation — isotope evidence. Geochim. Cosmochim. Acta 50, 693–709. Yamamoto, S., Alcauskas, J.B., Crozier, T.E., 1976. Solubility of methane in distilled water and seawater. J. Chem. Eng. Data 21, 78–80. Yates, K.K., Dufore, C., Smiley, N., Jackson, C., Halley, R.B., 2007. Diurnal variation of oxygen and carbonate system parameters in Tampa Bay and Florida Bay. Mar. Chem. 104, 110–124. Zappa, C.J., Raymond, P.A., Terray, E.A., Mcgillis, W.R., 2003. Variation in surface turbulence and gas transfer velocity over a tidal cycle in a macro-tidal estuary. Estuaries 26, 1401–1415.