Carbon dioxide and methane fluxes at the air – sea interface of Red Sea mangroves

Mangrove forests are highly productive tropical and subtropical coastal systems that provide a variety of ecosystem services, including the sequestration of carbon. While mangroves are reported to be the most intense carbon sinks among all forests, they can also support large emissions of greenhouse gases (GHGs), such as carbon dioxide (CO2) and methane (CH4), to the atmosphere. However, data derived from arid mangrove systems like the Red Sea are lacking. Here, we report net emission rates of CO2 and CH4 from mangroves along the eastern coast of the Red Sea and assess the relative role of these two gases in supporting total GHG emissions to the atmosphere. Diel CO2 and CH4 emission rates ranged from −3452 to 7500 μmol CO2 m−2 d−1 and from 0.9 to 13.3 μmol CH4 m−2 d−1 respectively. The rates reported here fall within previously reported ranges for both CO2 and CH4, but maximum CO2 and CH4 flux rates in the Red Sea are 10to 100-fold below those previously reported for mangroves elsewhere. Based on the isotopic composition of the CO2 and CH4 produced, we identified potential origins of the organic matter that support GHG emissions. In all but one mangrove stand, GHG emissions appear to be supported by organic matter from mixed sources, potentially reducing CO2 fluxes and instead enhancing CH4 production, a finding that highlights the importance of determining the origin of organic matter in GHG emissions. Methane was the main source of CO2 equivalents despite the comparatively low emission rates in most of the sampled mangroves and therefore deserves careful monitoring in this region. By further resolving GHG fluxes in arid mangroves, we will better ascertain the role of these forests in global carbon budgets.


Introduction
Mangrove forests, typically growing in the intertidal zones of tropical and subtropical coasts, are highly productive components of coastal ecosystems and adapted to high salinity and anoxic conditions associated with waterlogged sediments.Mangrove forests cover a global estimated area of 137 760 km 2 (Giri et al., 2011) and are typically constrained by temperature, with greatest biomass and species diversity in the equatorial zone (Alongi, 2012).Mangroves rank among the most threatened ecosystems in the biosphere, with losses estimated at 50 % of their global extent over the past 50 years (Alongi, 2012).These losses affect nearly all mangrove regions but the Red Sea, where mangrove coverage has increased by 12 % over the past 4 decades (Almahasheer et al., 2016).
Loss of mangrove forest represents a loss of valuable ecosystem services, including habitat and nurseries for marine species, coastal protection from erosion due to wave action, and the filtration of harmful pollutants from terrestrial sources (Alongi, 2008), as well as loss of CO 2 sink capacity.Additionally, mangroves can become a source of greenhouse gas (GHG) emissions from disturbed soil carbon stocks (Donato et al., 2011;Alongi, 2014).Hence, mangrove conservation and restoration have been proposed as important components of so-called blue carbon strategies to mitigate climate change (Duarte et al., 2013).Indeed, mangroves are reported to be the most intense carbon sinks among all forests, supporting carbon sequestration rates and organic carbon stocks by as much as 5 times higher than those in terrestrial forests (Donato et al., 2011).While mangrove forests cover less than 1 % of the total coastal ocean area, they contribute to almost 15 % of total carbon sequestration in coastal ecosys-tems (Alongi, 2012), making mangrove forests highly effective in terms of carbon sequestration per unit area.The management of mangroves to maximize CO 2 removal and subsequent storage is gaining momentum as a cost-effective strategy to mitigate climate change.
However, mangrove forests act as both carbon sinks and sources and have been reported to support large GHG emissions in the forms of CO 2 and CH 4 (Allen et al., 2007;Kristensen et al., 2008a;Chen et al., 2016).Whereas concerns are focused on GHG emissions following mangrove disturbance, estimated at 0.02-0.12Pg C yr −1 globally (Donato et al., 2011), undisturbed mangrove sediments also support GHG emissions (Purvaja and Ramesh, 2000;Kristensen et al., 2008b;Chauhan et al., 2015).Recent reports specifically highlight the importance of methane in flux estimates, as emissions of CH 4 with a higher global warming potential can offset mangrove carbon burial by as much as 20 % (Rosentreter et al., 2018b).Previous studies on GHG emission rates either focus on the soil-atmosphere interface, highlighting substantial flux ranges with mangroves reported to act as negligible (Alongi et al., 2005) to considerable sources (Livesley and Andrusiak, 2012;Chen et al., 2016), or examine net fluxes at the air-sea interface, with few studies in arid systems.Comparisons of carbon sequestration rates between mangrove stands have revealed that climatic conditions play an important role, with mangroves in the arid tropics, such as those in the Red Sea, supporting the lowest carbon sequestration rates (Almahasheer et al., 2017).Likewise, GHG emissions from mangrove forests may vary with climate, with most reported rates to-date derived from the wet tropics (Alongi et al., 2005;Chauhan et al., 2015;Chen et al., 2016).Whereas Red Sea mangroves are considered to play a minor role as CO 2 sinks, their role may be greater than portrayed by low carbon burial rates if they also support very low GHG emissions, thereby leading to a balance comparable to mangroves in the wet tropics.
Here we report air-sea emission rates of CO 2 and CH 4 , along with their carbon isotopic composition, from incubations of inundated mangrove sediment cores along the Saudi coast of the Red Sea.We assess the relative role of these two gases in supporting total GHG emissions as well as their fluctuations along the day-night cycle.

Study area
We sampled seven mangrove forests along the eastern coast of the Red Sea (Fig. 1).We collected triplicate sediment cores by inserting translucent PVC tubes (30.5 cm in height and 9.5 cm in diameter) into mangrove sediments, collecting approx.20 cm of sediment and a top seawater layer.The overlying water was regularly replaced by fresh seawater from the corresponding station in order to fill the remaining core volume and to measure CO 2 and CH 4 fluxes from underlying sediments during incubations.Mangrove sediments were sampled 5 to 10 m from the forest edge, typically in the centre of the mangrove belt.We sampled two stations (S1 and S2) in January and February 2017 and the other five mangrove stations (S3-S7) in March on board the R/V Thuwal as part of a scientific cruise.The cores collected from S1 and S2 were immediately transported to the laboratory, placed in seawater baths and enclosed in environmental growth chambers (Percival Scientific Inc., Perry, IA, USA) with 12:12 light cycles at a constant temperature of 26 • C. The sediment cores collected during the scientific cruise were transported immediately on board and placed in open aquarium tanks with running seawater in order to keep them close to in situ temperature.Salinity and temperature were routinely recorded using a conductivity, temperature, and depth sensor (EXO1, YSI inc., Yellow Springs, USA).Additionally, sediment chlorophyll a and nutrient (organic carbon and nitrogen) content was analysed from cores collected during the scientific cruise.

Sediment characteristics
The chlorophyll a content of the sediment was measured by fluorometry.The surface layer of each replicate core was collected and frozen until further analysis.Prior to chlorophyll a extraction, the sediment samples were left at room temperature to thaw.The chlorophyll a was extracted by adding 7 mL of 90 % acetone to 2 mL of sediment sample.After a 24 h incubation at 4 • C in dark conditions, the samples were centrifuged and the chlorophyll a content in the supernatant was measured on a Trilogy fluorometer.The nutrient (organic carbon and nitrogen) content of the sediment was analysed on an Organic Elemental Analyzer (Flash 2000, Thermo Fisher Scientific, Massachusetts, USA) after acidification of sediment samples.

Measurement of greenhouse gas fluxes
We measured CO 2 and CH 4 air-sea fluxes using two different techniques.The CO 2 and CH 4 fluxes from stations S1 and S2 were measured using the closed water circuit technique and the CO 2 and CH 4 fluxes from the rest of the stations sampled during the scientific cruise (S3-S7) were measured using the headspace technique.

Measurement of CO 2 and CH 4 fluxes in sediment core incubations using closed water circuit technique
We incubated mangrove sediment cores from stations S1 and S2 using a closed water circuit technique in order to measure changes in CO 2 and CH 4 concentrations.Before starting the incubation, the seawater above the sediment from each core was replaced by fresh seawater collected from the same location, avoiding disturbance of the sediment.Then, the seawater from the core was recirculated by a peristaltic pump in an enclosed water circuit through a membrane equilibrator (Liqui-cel mini module, 3M, Minnesota, USA).This set-up enables the equilibration of gases in dissolution with an enclosed air circuit.The air from the enclosed air circuit was then passed through a desiccant column (calcium sulfate, WA Hammond Drierite Co., LTD, Ohio, USA) and flowed into a cavity ring-down spectrometer (CRDS; Picarro Inc., Santa Clara, CA, USA) to continuously measure the CO 2 and CH 4 concentrations.We ran the incubations for at least 30 min under light (200 µmol photons m −2 s −1 ) and dark conditions.The concentration of CO 2 in the water circuit (µmol mL −1 ) was calculated by Eq. (1): where H cp is the Henry constant (mol mL −1 atm −1 ) calculated using the R marelac package (Soetaert et al., 2016), [HP_CO 2 ] is the given concentration of CO 2 (ppm), and pH 2 0 is the water vapour pressure (atm).
The CO 2 fluxes were calculated from the change in CO 2 concentration over time during our incubations, correcting for the seawater volume present in each core.Then, the fluxes were transformed to an aerial basis (µmol m −2 h −1 ) by taking into account the core surface area.Finally, the daily fluxes (µmol m −2 d −1 ) were calculated by multiplying the CO 2 flux obtained under light conditions by the number of light hours plus the CO 2 flux obtained under dark conditions by the number of dark hours.
The CH 4 fluxes were calculated in the same manner as for the CO 2 fluxes, with the exception that the Henry constant was calculated using Eq. ( 2): where H cp is the Henry constant (mol mL −1 atm −1 ), R is the ideal gas constant (82.057338 atm mL mol −1 K −1 ), T is standard temperature (273.15K), and β is the Bunsen solubility coefficient of CH 4 , extracted from Wiesenburg and Guinasso (1979).

Measurement of CO 2 and CH 4 fluxes in sediment core incubations using the headspace technique
Mangrove sediment cores from stations S3 to S7 were incubated using a headspace technique in order to measure changes in CO 2 and CH 4 concentrations.Before starting the incubation, the seawater above the sediment from each core was replaced by fresh seawater from the running seawater system, leaving a headspace of 200 mL.Each core was sealed with a stopper equipped with a gas-tight valve serving as a headspace sampling port.The sealed core was left for 1 h before the first headspace sampling to allow equilibration between seawater and air phases.Each core was sampled with a syringe, withdrawing 15 mL of air from the equilibrated headspace.Headspace samples were periodically drawn from each sediment incubation over a 24 h incubation period.The CO 2 and CH 4 concentrations in the headspace samples along with their isotopic composition (δ 13 C-CO 2 and δ 13 C-CH 4 ) were measured with a CRDS (Picarro Inc., Santa Clara, CA, USA) connected to a small sample isotopic module extension (SSIM A0314, Picarro Inc., Santa Clara, CA, USA).We ran standards (730 ppm CO 2 , 1.9 ppm CH 4 ) before and after every three samples.The concentration of dissolved CO 2 in the seawater after equilibrium was calculated from the concentration in the equilibrated headspace (ppm) as described previously by Wilson et al. (2012) for other gases: where β is the Bunsen solubility coefficient of CO 2 (mol mL −1 atm −1 ), m a is the given concentration of CO 2 in the equilibrated headspace (ppm), and p dry is atmospheric pressure (atm) of dry air.The Bunsen solubility coefficient of CO 2 was calculated using Eq. ( 4): where H cp is the Henry constant (mol mL −1 atm −1 ) calculated using the R marelac package (Soetaert et al., 2016), R is the ideal gas constant (82.057338 atm mL mol −1 K −1 ) and T is standard temperature (273.15K).The atmospheric pressure of dry air (p dry ) was calculated using Eq. ( 5): where p wet is the atmospheric pressure of wet air corrected by the effect of multiple syringe draws from the same core, applying Boyle's law.
The initial concentration of dissolved CO 2 in seawater before equilibrium was then calculated as follows: where [CO 2 ] w is the concentration of dissolved CO 2 in the seawater after equilibrium, V w is the volume of seawater (mL) and V a is the headspace volume (mL) in the core.Finally, treating the gas as ideal, the units were converted to nM using Eq. ( 7): where R is the ideal gas constant (0.08206 atm L mol −1 K −1 ) and T is temperature (K).
The CO 2 fluxes were calculated from the change in CO 2 concentration over time during our incubations, correcting for the seawater volume present in each core.Then, the fluxes were transformed to an aerial basis (µmol m −2 d −1 ) by taking into account the core surface area.Finally, the day and night fluxes (µmol m −2 h −1 ) were calculated from the change in CO 2 concentration between consecutive samplings during day and night-time.
The CH 4 fluxes were calculated in the same manner as the CO 2 fluxes, with the exception that the Bunsen solubility coefficient of CH 4 was calculated according to Wiesenburg and Guinasso (1979).
2.4 Isotopic composition of CO 2 (δ 13 C-CO 2 ) and CH 4 (δ 13 C-CH 4 ) The isotopic signature of the CO 2 and CH 4 produced during incubations was estimated by conducting keeling plots (Pataki et al., 2003;Thom et al., 2003;Garcias-Bonet and Duarte, 2017).Briefly, the δ 13 C of the CO 2 and CH 4 produced was extracted from the intercept of the linear regression between the inverse of the gas partial pressure and the isotopic signature.The data set is available from Sea et al. (2018).

Results
The mean (±SE) diel CO 2 and CH 4 emission rates for the seven sites were 372 ± 1309 µmol CO 2 m −2 d −1 and 5.6 ± 1.6 µmol CH 4 m −2 d −1 respectively.We observed high   variability among the seven mangrove forest sites studied, with net CO 2 and CH 4 diel emission rates ranging from −3452 to 7500 µmol CO 2 m −2 d −1 and from 0.9 to 13.3 µmol CH 4 m −2 d −1 (Table 1).Mangrove sediments absorbed CO 2 during daytime and emitted CO 2 during night-time at 5 out of 7 stations, with means (±SE) of −54.6 ± 37 µmol CO 2 m −2 h −1 and 86 ± 120 µmol CO 2 m −2 h −1 (Table 1, Fig. 2).However, in three out of seven sites, heterotrophic activities outbalanced photosynthesis on a 24 h basis.At two sites, S3 and S6, we found an increase of the CO 2 emissions between day and night, contradictory to the classical daytime primary productionnight-time respiration pattern, possibly indicative of a light mediated increase in heterotrophic processes.
Methane emissions did not show circadian patterns with linear increases in CH 4 concentration in our incubations (Fig. 2) and with similar light and dark rates (0.26±0.08 and 0.21 ± 0.07 µmol CH 4 m −2 h −1 (mean ± SE) respectively; Table 1).In terms of the total GHG contribution, the mean CO 2 equivalent (CO 2 e) emission to the atmosphere was 564 ± 1284 µmol CO 2 e m −2 d −1 (mean ± SE) using the 100year time horizon global warming potential (Myhre et al., 2013).Inundated mangrove sediments were net emitters of CO 2 e in three out of seven sites (Table 1), and in five out of seven mangrove stands sampled, CH 4 was the main source of CO 2 e to the atmosphere.While no overall trend was revealed through the relationship between day and night fluxes for CO 2 and CH 4 (Fig. 3), consistencies are evident at specific mangrove stations.For example, night CO 2 emissions are clearly visible at S2, while S3 appears to emit CO 2 during daylight hours.No relationship was apparent between GHG fluxes and the densities of organic carbon or nitrogen in the sediment.There was no discernible trend between gas fluxes and chlorophyll a content in surface sediments.
The variability in GHG emission rates reported in this study could be attributed to spatial differences, as cores were taken from different parts of each forest.Indeed, previous studies report significant discrepancies in emission rates in fringe versus forest positions (Allen et al., 2007).Additionally it is possible that differences in flux rates may exist as a result of sediment disturbance from the coring process.The effects of mangrove pneumatophores and possible bioturbation from infaunal species such as burrowing crabs were not considered here yet could pose another possible source of variation in the results as the presence of these structures influences oxygenation of sediment and pore water exchange, identified as driving factors in varying CO 2 levels (Call et al., 2014;Rosentreter et al., 2018).These factors likely affect relevant redox processes and would therefore be useful to quantify in future studies.
Uniformity of day and night emission rates for CH 4 was observed in Red Sea mangrove stands, with mean (±SE) CH 4 emission rates of 0.28±0.08µmol CH 4 m −2 h −1 during the day and 0.24±0.08µmol CH 4 m −2 h −1 during night; this is consistent with previous work reporting that emission rates for CH 4 do not vary significantly during light and dark hours in mangrove forests (Allen et al., 2007).It has been suggested instead that variables such as sediment temperature are more significant in their contributions to emission rates (Allen et al., 2007(Allen et al., , 2011)).Incubated sediment cores kept at constant temperature do not reflect the range of temperatures experienced by mangrove sediments over the diurnal cycle; future studies examining GHG emissions under more realistic temperature fluctuations are needed.Seasonal studies of longer duration have reported increased emission rates during warmer seasons (Chen et al., 2016;Livesley and Andrusiak, 2012).Methane concentrations typically remain low due to anaerobic methane oxidation processes that take place near sediment surfaces (Kristensen et al., 2008a), consistent with the low CH 4 emission rates from Red Sea mangrove sediments observed here.Additionally, environments of high salinity like the Red Sea have been associated with decreased CH 4 emissions, as sulfate-reducing bacteria are thought to outcompete methanogens (Poffenbarger et al., 2011).
Methane emission rates at the air-sea interface of Red Sea mangrove sediments, although quite low, become more substantial when considered in terms of global warming potential.In this study, CH 4 was, despite the comparatively low emission rates, the main source of CO 2 e in the majority of sampled mangroves, and therefore deserves careful monitoring in this region.Reported organic carbon burial rates of Red Sea mangroves of 3.42 mmol C m −2 d −1 (Almahasheer et al., 2017) are 10 times larger than the combined average CO 2 and CH 4 emission rates reported here (0.37 mmol C m −2 d −1 ), suggesting that these mangrove sediments could act as net atmospheric carbon sinks; however, significant alkalinity and DIC exports have been identified from mangroves as well (Sippo et al., 2016), necessitating future studies which measure these exports to neighbouring habitats in order to close the carbon budget and determine the role of Red Sea mangroves in potential climate change mitigation.Currently, protection measures and further reforestation efforts are being deployed along the Red Sea, which will further expand the area of mangroves (Almhasheer et al., 2016).The rationale for conserving mangroves in the climate change context is not adequately represented by their net carbon sink capacity when undisturbed, but rather by the emissions resulting from their disturbance.Indeed, previous studies analysing anthropogenic impacts on methane emission rates from mangrove sediments have shown that the a disturbance significantly increases methane emissions (Purvaja and Ramesh, 2001;Chen et al., 2011).This provides an additional rationale to conserve and continue to expand Red Sea mangroves.
While this study provides new insights into GHG fluxes from arid mangroves, the methods used here solely measure the air-sea fluxes of dissolved gases.If CO 2 is produced from underlying sediments, it enters the water column and becomes a part of the carbonate system, with the possibility of conversion to bicarbonate (HCO − 3 ) and carbonate (CO 2− 3 ) ions; these dominating species represent over 99 % of the dissolved inorganic carbon (DIC) under current atmospheric and oceanic conditions (Zeebe and Wolf-Gladrow, 2001).Therefore, the air-sea equilibration methods used in this study do not measure DIC fluxes but only the fluxes of the dissolved CO 2− component of this larger system.Frankignoulle and Borges (2001) show that CO 2 can be measured either directly (using equilibrator techniques and spectroscopy or chromatography) or indirectly (by making calculations based on pH, total alkalinity, and DIC).The methodology presented in this study represents the former, utilizing an air-sea equilibrator connected to a CRDS to measure GHG fluxes at the air-sea interface.Research conducted by Borges et al. (2003) utilizes the indirect approach, using pH and total alkalinity measurements in Papua New Guinea to calculate DIC and CO 2(dis) (for a computational discussion see Frankignoulle and Borges, 2001).Both methods measure at the air-sea interface (Table 2) but are not directly comparable, as a full determination of the carbonate system was not carried out in the present study.Similarly, studies using equilibrator techniques that measure the dissolved CO 2 fraction of seawater to the atmosphere are influenced by the seawater carbonate system and further steps of isotopic fraction (discussed below) and are therefore not directly comparable to those studies which measure GHG fluxes from exposed mangrove sediments to the atmosphere (Table 2).

Isotopic composition of emitted gases
There were no relationships between GHG fluxes and sediment properties, such as chlorophyll a, nitrogen density, and organic carbon density, suggesting that other factors have a greater influence on GHG flux rates in this region.Since mangroves can receive large contributions of organic carbon from other sources (Newell et al., 1995), such as algal mats, seagrass and seaweed, examination of the isotopic composition of emitted carbon provides insights into the origin of the organic carbon supporting GHG fluxes in mangrove sediments; however, it should be noted that δ 13 C values reported in this study occur after several steps of isotopic fractionation and may therefore influence results.Isotope effects can cause an unequal distribution of isotopes between DIC components; for example, as CO 2 is produced from mangrove sediments and becomes part of the carbonate system (likely forming HCO − 3 after equilibration), molecules containing the heavier carbon isotope -with a higher activation energywill typically react more slowly (Zeebe and Wolf-Gladrow, 2001), promoting a higher concentration of the heavy isotope in unreacted CO 2 and a relative depletion of this heavier isotope in resulting HCO − 3 .Similarly, this preferential incorporation and movement of molecules containing lighter isotopes can affect resulting carbon isotope ratios after air-sea equilibration (with depletion of lighter isotopes in seawater as a result of fractionation).CO 2 measured in this study is subject to these processes and may not reflect the isotopic ratios of carbon originally emitted; rather, the signatures measured in this study should be seen as a proxy which reflects isotopic ratios of air-sea discrimination and biological processing (decomposition, respiration, and photosynthesis), resulting after carbon isotope fractionation.An interpretation of the results is therefore subject to this limitation.
The isotopic signature of the CO 2 (δ 13 C-CO 2 ) produced by mangrove sediments in four out of the five mangrove stands with available isotopic data was heavier (from −11.2 ± 0.6 to −15.9 ± 1.1 ‰; Table 1) than the isotopic signature of mangrove tissues, suggesting decomposition of organic matter from mixed sources (Kennedy et al., 2010).Specifically, the isotopic signature of the mangroves found in the central Red Sea has been recently reported as δ 13 C leaves = −26.98±0.15‰, δ 13 C stems = −25.75±0.16 and δ 13 C roots = −24.90±0.17‰ for mangrove leaves, stems and roots, while the mean isotopic signature of other pri- mary producers in the central Red Sea has been reported as δ 13 C seaweed = −12.8± 0.5 and δ 13 C seagrass = −8.2± 0.2 ‰ for seaweed and seagrass tissues respectively (Almahasheer et al., 2017).However, in one mangrove stand (S3) the δ 13 C-CO 2 was much lighter (−25.72 ± 0.21 ‰), potentially indicating mangrove tissues.Thus, according to the isotopic signature, the CO 2 produced in mangrove sediments would be supported by mangrove biomass in only one mangrove stand out of the five sampled sites with available isotopic data.Moreover, the mean isotopic signature of the CH 4 produced in mangrove sediments (δ 13 C-CH 4 = −80.6 ‰) tentatively confirms its biogenic origin, which normally ranges from −40 to −80 ‰, depending on the isotopic signature of the organic compounds being biologically decomposed (Reeburgh, 2014).The lowest δ 13 C-CH 4 was detected in S3, coinciding with the lowest δ 13 C-CO 2 value, suggesting that the organic matter being decomposed by methanogens likely came from mangrove tissues as well.
Interestingly, the mangrove with the lightest δ 13 C-CO 2 and δ 13 C-CH 4 (S3) showed the lowest daily CO 2 flux (−1524±686 µmol CO 2 m −2 d −1 ) but the highest CH 4 emission rate (13.3 ± 9.5 µmol CH 4 m −2 d −1 ) compared to the fluxes detected in the rest of mangrove stands with available isotopic data.Part of the variability in the CO 2 (R 2 = 0.42) and CH 4 (R 2 = 0.40) emission rate seems to be explained by the origin of the organic matter being decomposed, estimated here as δ 13 C-CO 2 and δ 13 C-CH 4 .Organic matter with lighter isotopic composition could enhance CO 2 emissions, whereas organic matter with heavier isotopic composition could enhance CH 4 emissions (Fig. 5), possibly suggesting a different preferential use of organic matter by different microbial groups in mangrove sediments.Future studies exploring this idea with further considerations of carbon isotope fractionation would help solidify the role of the origin of organic carbon stored in mangrove sediments on their GHG emissions.

Conclusions
This study is first in reporting CO 2 and CH 4 fluxes from Red Sea mangrove sediments, contributing to the scant data on arid mangrove systems (Atwood et al., 2017;Almahasheer et al., 2017), essential to establishing a solid baseline on GHG emissions for future studies.Results show that maximum CO 2 and CH 4 flux rates from Red Sea mangrove sediments are well below those reported elsewhere, and that, even when considered in terms of CO 2 equivalents, carbon burial rates largely outweigh GHG emission rates at the airsea interface in this region.This study also highlights the importance of determining the source of organic matter in GHG flux studies, as emissions appear to be supported by organic matter from mixed sources in the majority of studied mangroves, potentially enhancing CH 4 production over CO 2 fluxes in this system.Seasonal variation should be considered in future studies on GHG emissions by Red Sea mangroves to better determine annual emission rates from this system, which reach some of the warmest temperatures experienced by mangrove forests worldwide.Similarly, a wider spatial coverage within the mangrove forest should be considered to confidently determine net GHG fluxes that can be upscaled to the entire stock of Red Sea mangroves.
Methods presented in this study include the use of an air-sea equilibrator connected to a CRDS to measure GHG fluxes at the air-sea interface, measuring the dissolved CO 2− component of the larger seawater carbonate system.This methodology is one of many used to measure GHG flux rates; establishing a unified sampling technique at both the soil-atmosphere and air-seawater interface will aid future researchers in determining total carbon budgets and accurately informing policymakers of their findings.In combination with a consideration of isotope effects, a full determination of the carbonate system will be beneficial in future studies to further resolve GHG fluxes in arid mangroves, allowing us to better ascertain the role of these forests in global carbon budgets.

Figure 1 .
Figure 1.Mangrove stands sampled along the Saudi coast of the Red Sea.Numbers indicate positions of sampling sites from this study.S1 and S2: King Abdullah University of Science and Technology, S3: Duba, S4 and S5: Al Wahj, and S6 and S7: Farasan Banks.
a gr sed −1

Figure 2 .
Figure 2. Change in CO 2 (a) and CH 4 (b) concentrations over time in triplicated mangrove sediment cores from mangrove stations S3-S7.Shaded areas represent night-time and each replicate is coded by different symbols.

Figure 3 .
Figure 3. Relationship between day and night fluxes for CO 2 (a) and CH 4 (b) at all mangrove stations.

Figure 4 .
Figure 4. Keeling plots for mangrove stations S3-S7, showing the linear regression of the inverse of CO 2 concentration (a) and CH 4 concentration (b) versus δ 13 C-CO 2 and δ 13 C-CH 4 .Y intercepts were used to estimate the isotopic signatures of produced gases.

Figure 5 .
Figure 5. Relation between (a) the carbon isotopic signature of the produced CO 2 (δ 13 C-CO 2 ) and CO 2 fluxes and (b) carbon isotopic signature of the produced CH 4 (δ 13 C-CH 4 ) and the CH 4 fluxes in Red Sea mangroves.Error bars indicate standard error of the mean.

Table 1 .
(Myhre et al., 2013)e gas fluxes and sediment characteristics from studied mangrove forests.CH 4 fluxes in brackets represent CO 2 equivalents in terms of global warming potential for a time horizon of 100 years (GWP 100 ), taking into account climate-carbon feedback as suggested by the AR5 of IPCC(Myhre et al., 2013).Data represent the mean ± SEM and nd means no data available.
CO 2 and CH 4 emissions reported in this study show that Red Sea mangroves can act as a source of GHG to the atmosphere.Values reported from this study fall within previously reported ranges for both CH 4 and CO 2 , but maximum CH 4 and CO 2 flux rates in the Red Sea are up to 100 fold below those reported elsewhere.Compiled global values for GHG fluxes range from −16.9 to 629.2 mmol CO 2 m −2 d −1 and −2.1 to 25,974 µmol CH 4 m −2 d −1 , with mean (± SE) maximum emission rates averaging 202.3±48 mmol m −2 d −1 and 4783.6 ± 2783 µmol m −2 d −1 for CO 2 and CH 4 respectively (Table

Table 2 .
Comparison of GHG fluxes from global mangrove forests and Red Sea mangroves.Literature values converted from reported form for comparison purposes.Measurements made at the 1 soil-atmosphere interface, 2 air-sea interface with DIC calculation methods, and 3 air-sea interface with equilibration methods.