the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 25 Jun 2026
Quantifying air–sea CO2 fluxes above desert-fringing coral reefs in the northern Red Sea revealed by eddy covariance
Abstract. Eddy covariance (EC) measurements of air-sea CO2 exchange over desert fringing coral reefs in the Gulf of Eilat (Aqaba) (GoE), northern Red Sea, show these ecosystems are net sinks of atmospheric CO2. This result contrasts with marine productivity models and bulk formula calculations based on water chemistry that are often used methods to determine the magnitude and direction of the CO2 flux with the atmosphere over coral reefs. These studies have often concluded that coral reefs are net sources of CO2 to the atmosphere with only rare cases finding otherwise. Our EC measurements find coral reefs in the GoE may absorb around 4.5 times more carbon from the atmosphere than other marine and terrestrial ecosystems and only slightly less than some tropical rainforests. This highlights the need for further direct measurements of air-sea CO2 exchanges over coral reefs in different environmental settings so their role in the global carbon cycle can be accurately quantified.
- Preprint
(1350 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 20 Aug 2026)
-
RC1: 'Comment on egusphere-2026-3467', Anonymous Referee #1, 23 Jul 2026
reply
-
AC1: 'Reply on RC1', Hamish McGowan, 31 Jul 2026
reply
We thank the Anonymous Referee #1 for posting their comments on the 23 Jul 2026 on our article that reports direct measurements of air-sea CO2 fluxes over coral reefs in the northern Gulf of Eilat (Aqaba), northern Red Sea. Here we take the opportunity to respond to their comments including their statement, “The most critical point is that the discussion in this study relies heavily on flux observations obtained using the EC method and is not sufficiently supported by independent supplementary data, such as the pCO₂ in seawater or other carbonate system parameters.”
The eddy covariance (EC) method is acknowledged as the ‘gold standard' for direct measurement of energy, moisture, and trace exchanges between the Earth’s surface and the atmosphere. Importantly, when deployed correctly it provides representative ecosystem scale measurement of sensible and latent heat fluxes and for example, fluxes of CO2 and CH4. This is the reason why EC is the measurement method of choice for the direct measurement of CO2 fluxes by the global FLUXNET program (Baldocchi et al., 2001) as we discuss in lines 27-31 and the articles we cite therein.
Referee #1 raised the issue of EC data corrections for factors such as effects of water vapor fluxes, sensible heat fluxes, density variations, instrument response etc. These corrections are standard operating procedure for EC deployments and are now part of EC software packages such the EddyPro® Software (Li-COR Biosciences) that we used (see lines 150-151). An excellent introduction to EC is provided by Burba (2013) who we cite which also discusses these corrections in detail. In addition, the EC instrumentation we used has been successfully deployed in other marine settings however, it is important to ensure that the sapphire lens of the Li-Cor 7500 open-path gas analyzer are kept clean of dust and salt to avoid potential CO2-H2O cross-talk issues which we discuss in lines 142-148. We did this by regular washing of the instrument.
As stated by Referee #1, our open path gas analyser sampled air (at 20Hz and averaged over 30 minutes) at a height 2.5 m above the surface as shown in Fig 1b. Air arriving at this sampling height carried with it the signatures of the upwind surface (measurement footprint) as shown in Fig 1c which was determined using an EC foot-printing model – in our study the widely used Kljun model which we cite (Kljun, 2004). As shown in Fig 1c and discussed in section 2.2 we filtered out samples arriving to our EC system from outside the measurement footprint shown in Fig 1c. This gives the EC method a very significant advantage over approaches mentioned by Referee #1 such as “pCO₂ and/or DO in seawater” which we discuss in our article for example, in lines 39-50. Importantly, such approaches referred to by Referee #1 do not provide a direct measurement of air-sea CO2 exchange which EC does. They also have many inherit limitations that we discuss in lines 45 – 50 including a sampling methodology that often does not capture ecosystem scale pCO₂ and/or DO in seawater because of their point and time specific water sampling.
Regarding the comment by Referee #1 on the relative air-sea CO2 exchange over the coral reef at Eilat which we report in comparison to air-sea CO2 exchange over the adjacent ocean. The approach Referee #1 suggests of possibly subtracting the ‘background’ ocean air-sea CO2 exchange from the coral reef air-sea CO2 exchange to find what we assume they are suggesting is the net air-sea CO2 exchange directly attributable to the coral reef is fraught with uncertainty and not appropriate. As we describe in our article in lines 108-115, water under our EC measurement footprint travels alongshore at our study site and is therefore imprinted with the hydrodynamic and biogeochemical properties of the coral reef that determine the water pCO₂. Offshore the pCO₂ of the oligotrophic/mesotrophic deep water in the Gulf is determined by other factors such as for example, seasonal overturning, exchange of surface water with the Red Sea, and dark inorganic carbon fixation etc. (see Reich et al 2024, Abir et al. 2024, Biton & Gildor 2011).
Referee #1 suggests that we should provide more than ‘qualitative’ discussions on environmental factors that may influence air-sea CO2 exchange over the coral reef at our study site. We do this in lines 205-224 however, we do agree that additional research is required to unpick further the relative contributions of environmental drivers of air-sea CO2 exchange over coral reefs as we mention in our Conclusion but is beyond the scope of the current study. We would like to stress that the point of our article is to show that when direct measurement of air-sea CO2 exchange over coral reefs is undertaken at ecosystem scale, some reefs such as at Eilat behave as net CO2 sinks as increasingly reported at other reefs in studies that we cite.
Referee #1 questions the ‘novelty’ of our article and the duration of our observations. In lines 56-62 we discuss the different focus of the current article to our previously published research. We do present data on seasonal and diurnal variations in air-sea CO2 exchange over the coral reef as suggested by Referee #1 in Fig 4 & Fig 5. We agree with Referee #1 that interannual variations in air-sea CO2 exchange over the coral reef is a most worthy area of investigation which we indicate in our Conclusion, but it is beyond the scope of the current study and funding. However, we would like to draw attention to the previous study of Rey-Sánchez et al. (2017) that we discuss in our article who also used EC to directly measure air-sea CO2 exchange including over part of the coral reef at Eilat from late 1 April to 27 August of 2009 (142 days during spring and summer). Rey-Sánchez et al. (2017) also concluded the coral reef was a net sink of CO2. Collectively the study by Rey-Sánchez et al. (2017) and our study clearly show that on an annual basis the coral reefs at Eilat are net sinks of CO2. These findings align with a growing body of research, some of which we cite, that concludes some coral reefs act as net sinks of atmospheric CO2 (see the review article by Kayanne 2025 that we cite).
Response to minor comments by Referee #1
- L27: What exactly does “15 digits” refer to?
Authors response
This refers to the time and space scales of the biophysical processes controlling trace gas fluxes and originates from Figure 1 in Osmond et al (1980). We will revise our text to include the reference to the original publication of Osmond et al (1980).
- L78: The dissolution of CaCO₃ is the reverse reaction of calcification and is generally thought to act in a way that lowers pCO₂ in seawater.
Authors response
We thank Referee #1 for identifying this point and we will provide clarification on the role of dissolution of CaCO₃ in lowering pCO₂ in seawater.
- L124: Regarding the method used to calculate the carbon footprint, please describe the specific methodology used and cite appropriate references. Also, since the calculated carbon footprint may include marine areas other than coral reefs, how were coral reef areas distinguished from other marine areas?
Authors response
The EC measurement footprint was determined using the footprinting method of Kljun (2004) who we cite in lines 152-154. This included a wind direction filter that excluded samples from wind directions 90 to 360° (from over land and adjacent ocean) as stated in line 154. We appreciate that the Ocean Science community may not be overly familiar with EC footprinting and we will provide additional details on this commonly used method in a revised version of our manuscript.
- L223: Isn’t chlorophyll a concentration primarily an indicator of phytoplankton abundance in bulk seawater, rather than a direct reflection of benthic algae?
Authors response
We don’t imply that the chlorophyll-a concentration is a direct “reflection of benthic algae”. We instead refer to ‘algae in the water column over the coral reef’ in line 224.
- L233: Light intensity is thought to have a more direct impact on photosynthesis and CO₂ flux in coral reef ecosystems than chlorophyll a concentration or sea surface skin temperature. Please explain why light intensity was not included in the evaluation.
Authors response
Solar irradiance was included in the evaluation - see Fig 2b and lines 215-216. Our analysis found that solar radiation was not correlated with CO2 flux (r -0.07) (Line 215). We did not directly measure photosynthetically active radiation (PAR) which is defined as the radiative flux between 400 and 700 nm although its diurnal cycle and intensity is closely related to total solar irradiance (i.e. Meek et al 1984).
- L243: If discussing the impact of sea surface skin temperature, wouldn’t the effect mediated through changes in CO₂ solubility immediately below the sea surface be more direct and significant than the impact on biological activity?
Authors response
We thank Referee #1 for this comment. In lines 305-313 we discuss further the notable influence of sea surface skin temperature on air-sea CO2 exchange over the coral reef which has been observed elsewhere in studies we cite. This is certainly an area worthy of further investigation although we are not exactly certain of what Referee #1 is suggesting.
- L251: Please write the “2” in “CO₂” as a subscript. Since similar formatting errors are found elsewhere, please review the entire manuscript.
Authors response
We thank Referee #1 for this comment and will proceed with such corrections.
- L267: Please specify the methods, assumptions, time period, and calculation procedures used for the extrapolation.
Authors response
The conversion of EC CO2 flux to C flux is straightforward. There are no assumptions involved – please see the caption for Table in lines 284-287. Mass of C=Mass of CO2×0.2727.
- L309: Please clearly state the basis for adopting the value of 22.7°C.
Authors response
We report what our data showed, namely as we state in line 309 when the mean water skin temperature was < 22.7 °C the coral reef remained a CO2 sink throughout the diurnal cycle.
- L325: What is the basis for determining that the gas exchange coefficient is incorrect? Please specify the equations, parameters, or prior research used for comparison.
Authors response
We do not state that the gas exchange coefficient is incorrect – please read line 325. However, the gas transfer velocity of CO2 across the air-sea boundary is dependent on many factors such as water skin temperature, wave environment, bubble entrainment, wind and temperature gradients, biological films etc, and has an uncertainty of around 20% for open ocean applications (see Wanninkhof 2014 who we cite). Accordingly, non-EC approaches that rely on determining the pCO₂ in seawater relative to the air and then estimate the net air-sea CO2 flux using the CO2 gas transfer velocity inherit such uncertainty. They don’t account for the many environmental factors that control gas (CO2) transfer. EC on the other hand directly measures the turbulent net exchange of CO2 from or toward the sea surface. In doing so it accounts for the influence of water skin temperature, wave environment, bubble entrainment, wind and temperature gradients, biological films etc in near real-time, that may influence air-sea CO2 exchange.
References:
Reich, T., Belkin, N., Sisma-Ventura, G., Berman-Frank, I. and Rahav, E. (2024), Significant dark inorganic carbon fixation in the euphotic zone of an oligotrophic sea. Limnol Oceanogr, 69: 1129-1142. https://doi.org/10.1002/lno.12560
Abir, S., McGowan, H.A., Shaked, Y., Gildor, H., Efrat, M. and Lensky, N.G. (2024) Air-sea heat exchange in the desert semi-enclosed Gulf of Eilat (Aqaba). Atmospheric Chemistry and Physics, 24, 6177–6195.
Biton, E., and H. Gildor (2011), The general circulation of the Gulf of Aqaba (Gulf of Eilat) revisited: The interplay between the exchange flow through the Straits of Tiran and surface fluxes, J. Geophys. Res., 116, C08020, doi:10.1029/2010JC006860.
Burba, G.: Eddy Covariance Method for Scientific, Industrial, Agricultural, and Regulatory Applications: A Field Book on Measuring Ecosystem Gas Exchange and Areal Emission Rates. LI-COR Biosciences, Lincoln, NE, USA, 331 pp, 2013.
Osmond C.B., Bjorkman O. & Anderson D.J. (1980) Physiological Processes in Plant Ecology: Towards a Synthesis with Atriplex. Springer-Verlag, Berlin.
Kayanne, H.: Thirty years since the coral reef CO2 sink/source debate. Galaxea, J. Coral Reef Stud., 27(1),118-130, 2025.
Meek, D.W., Hatfield, J.L., Howell, T.A., Idso, S.B. and Reginato, R.J. (1984), A Generalized Relationship between Photosynthetically Active Radiation and Solar Radiation1. Agron. J., 76: 939-945. https://doi.org/10.2134/agronj1984.00021962007600060018x
Citation: https://doi.org/10.5194/egusphere-2026-3467-AC1
-
AC1: 'Reply on RC1', Hamish McGowan, 31 Jul 2026
reply
-
RC2: 'Comment on egusphere-2026-3467', Anonymous Referee #2, 03 Aug 2026
reply
This study provides an interesting dataset of directly measured CO2 exchange over a reef in the Red Sea. The results are presented thoroughly and with useful additional measurements to explore drivers of the strong carbon uptake measured over this ecosystem. However, in my opinion there is a very strong focus on atmospheric drivers and a lack of investigation of reef metabolic processes. Calcification is only mentioned sparsely, and dissolution is once incorrectly said to lower CO2. A strong focus is put on the "net flux", although the study only covers 6 months of data (with 25% of that data interpolated over gaps). The authors use a lot of their manuscript to construct a large disagreement between directly measured EC fluxes and those inferred from delta pCO2 and gas transfer parameterizations that is not really substantiated. I think a greater focus on the ecosystem processes at play at this particular reef over the time scales you are able to resolve, as opposed to "net flux" comparisons with reefs elsewhere over various time frames would greatly improve this study.
Minor comments:l.23: “Understanding net ecosystem – atmosphere CO2 exchange (Net Ecosystem Exchange (NEE)) is crucial to inform policy responses to anthropogenic global warming.”
Why? Is this the right terminology when in fact you are looking at air-sea flux?l.45: “a limitation (…)”
How are these studies limited by their focus on scleractinian corals? Studies that use bulk formulas and water chemistry measurements usually are unable to attribute changes to specific groups of organisms and instead capture the net effect of all metabolic processes (and other processes).l.68: remove “over”
l.72: “influence this observation”
Are you saying that this influences the observation of- or the actual air-sea exchange?
l.78: “Calcification and dissolution raise the pCO2”
Dissolution lowers pCO2.
l. 80: “meteorological” instead of “meteorology”
l. 90: “refugium” instead of “refugia” (or remove “a” in the previous line)
l. 158 Gap filling: Could you provide a figure that shows the flux measurement time series with and without filled gaps?193: Why would “fertilization” of coral reefs have an instantaneous effect on CO2 drawdown? What are the time scales of the relevant processes?
Table 1: The time frames of these studies are so different that I question the utility of the comparison. You also don’t mention in the text your study at Heron Reef that found fluxes exceeding those that you report here.
l. 311: “overlying” instead of “over lying”
l. 319: You report photosynthesis/respiration ratios but what about the influence of calcification? Is this not a very significant process for reef carbon biogeochemistry?
l. 321: You often refer to other studies that have concluded reefs to be sources of CO2 to the atmosphere. However, all of these studies have taken place at other reef sites than the one you investigate here, so the construction of a “disagreement” between methods seems a bit artificial here, especially since you also suggest that all other methods should not be used, and instead all estimates should be done with EC. Can you compare EC measurements with results from other methods at the same site? The Heron Island reef, for instance, is indeed a sink for atmospheric CO2 when evaluated with the MAPCO2 mooring data and standard gas transfer parameterizations at that location.
l. 366 “whether coral reefs act as a source or sink” you are simplifying the question here – why shouldn’t different reefs look different in this respect? Some could be sinks and some could be sources, and the interesting questions are the details of what is driving those differences.
Citation: https://doi.org/10.5194/egusphere-2026-3467-RC2 -
AC2: 'Reply on RC2', Hamish McGowan, 05 Aug 2026
reply
We thank Anonymous Referee #2 for posting their comments 03 Aug 2026 on our article that reports direct measurements of air-sea CO2 fluxes over coral reefs in the northern Gulf of Eilat (Aqaba), northern Red Sea. Here we take the opportunity to respond to their comments.
The purpose of our study was to measure and report direct air – sea CO2 exchanges over the coral reefs at Eilat at ecosystem scale as such observations are scarce with our study presenting the longest continuous record of such measurements. These show that the coral reefs at Eilat are net sinks of CO2 – a result supported by previous research conducted near our site more than a decade ago by Rey-Sánchez et al., (2017) over a slightly shorter observation period during summer. This stands in contrast to approaches such as marine productivity models and water chemistry that have typically been used to calculate air – sea CO2 exchanges over coral reefs which we discuss in lines 38-46 that Referee #2 seems more familiar with. As we discuss in our article such approaches are typically site and time specific and seldom provide continuous measurements of the pCO2 in the water over a coral reef at the temporal resolution of eddy covariance (EC) and not at ecosystem scale. Further, such approaches do not provide a direct measurement of air-sea CO2 flux as we discuss and they have several limitations that we discuss also in lines 321-333.
To be clear, the focus of our article is not coral reef metabolic processes which we assume Referee #2 is suggesting as a means to try and resolve why the coral reef at our study site is acting as a CO2 sink. While we also believe that developing understanding of CO2 pathways in the reef ecosystem is important, it is not the focus of the current study. Furthermore, undertaking such investigations at the same space and time scales as eddy covariance is not feasible.
Referee #2 is correct – the focus on our article is on the "net flux" of CO2 which we sample at 20 Hz and average of 30 minutes for 181 days as discussed in Materials and Methods. The sampling period includes late summer, and winter seasons allowing comparisons of diurnal and seasonal air-sea CO2 flux (see Figs 3,4 & 5). Note this represents the longest record of direct measurement of air-sea CO2 flux over a coral reef. Gaps in data as noted Referee #2 are filled using accepted robust techniques as detailed in lines 157-171 with supporting citations (i.e. Moffat et al., 2007, Falge et al., 2001; Nemitz et al., 2018). As stated in line 161 all gaps in data were less than 15 hrs except for one gap of 3.54 days. Gap filling had minimal influence on computed CO2 fluxes as discussed in lines 168-171.
Minor comments by Referee #2:
l.23: “Understanding net ecosystem – atmosphere CO2 exchange (Net Ecosystem Exchange (NEE)) is crucial to inform policy responses to anthropogenic global warming.”
Why? Is this the right terminology when in fact you are looking at air-sea flux?Authors response.
We thank Referee #2 for this comment. Understanding net ecosystem – atmosphere CO2 exchange (Net Ecosystem Exchange (NEE – the correct terminology) is crucial as it informs the calculation of carbon budgets needed for accurate greenhouse gas accounting and developing understanding of the role of different ecosystems in the carbon cycle. We can provide further discussion of this point if our article is accepted for publication.
l.45: “a limitation (…)”
How are these studies limited by their focus on scleractinian corals? Studies that use bulk formulas and water chemistry measurements usually are unable to attribute changes to specific groups of organisms and instead capture the net effect of all metabolic processes (and other processes).
Authors response.
We thank Referee #2 for raising this question. On review we accept that the word ‘limitation’ may not accurately convey our message and we will revise this sentence. Our intention was to highlight that studies that use bulk formulas and water chemistry measurements typically do not capture the diversity of coral reef ecosystem influences on water pCO2. This is because of their site and time specific sampling strategies, i.e. often only sampling 1 or 2 sites on a reef and perhaps only several times per day for a few days, compared to EC that continuously samples the turbulent exchange of air originating from over a source area extending over thousands of square meters. We will revise this sentence accordingly to provide clarification. Bulk formulas and water chemistry measurements do not account for the influence of environmental factors on air-sea CO2 exchange such as air temperature, water skin temperature, wind, waves, bubble transfer, biofilms etc which EC does account for via the direct measurement of CO2 exchange with the atmosphere.
l.68: remove “over”
Authors response.
We will make this change.
l.72: “influence this observation”
Are you saying that this influences the observation of- or the actual air-sea exchange?Authors response.
We are referring to “the actual air-sea exchange”.
l.78: “Calcification and dissolution raise the pCO2”
Dissolution lowers pCO2.Authors response.
We thank Referee #2 for highlighting this as did Referee #1, and we will revise our text if our article is accepted to remove the apparent ambiguity that is present in our article on this point.
l. 80: “meteorological” instead of “meteorology”Authors response.
We agree with Referee #2 and we will amend our text “meteorological” processes.
l. 90: “refugium” instead of “refugia” (or remove “a” in the previous line)Authors response.
We agree with Referee #2 and we will amend our text accordingly.
l. 158 Gap filling: Could you provide a figure that shows the flux measurement time series with and without filled gaps?Authors response.
Yes, that is possible although what would be gained? The convention with EC is to typically present gap filled records, if possible, with an accompanying description of the gap filling procedure as we present in lines 157-171 and a discussion of the impact on the record. In EC records where gaps of several days to weeks are filled then these are often (should be) supported with a detailed analysis of the effect(s) of gap filling (see for example Mahabbati et al, 2021). This is also true in settings where the synoptic meteorology is highly variable and experiences rapid changes as occurs with the passage of cold fronts in mid-latitude settings.
193: Why would “fertilization” of coral reefs have an instantaneous effect on CO2 drawdown? What are the time scales of the relevant processes?
Authors response.
Dusts have been shown by Blanckaert et al., (2022) who we cite to provide important trace elements to corals in the GoE. They conclude that dusts provide “nitrate and other essential bioelements, such as iron, manganese, zinc and copper, that are rapidly assimilated by the symbionts”. This increases chlorophyll concentration and photosynthesis. Note we don’t use the term or imply an instantaneous effect on CO2 drawdown. The response time is hours to days.
Table 1: The time frames of these studies are so different that I question the utility of the comparison. You also don’t mention in the text your study at Heron Reef that found fluxes exceeding those that you report here.
Authors response.
We do acknowledge the various length of the datasets we present in Table 1 and we do highlight the length of some datasets and suggest caution is needed when making the comparisons (line 274). Nonetheless, these examples severe as direct comparisons between EC datasets and our EC GoE data of NEE which we consider very important. We hope that by making such comparisons and highlighting the length of data records, particularly in marine environments that are very short that it may encourage the collection of longer direct measurement EC CO2 flux datasets. We do cite the article that presents our previous EC CO2 flux measurements at Heron Reef which are also presented in Table 1
- 311: “overlying” instead of “over lying”
Authors response.
We thank Referee #2 for highlighting the need for this correction.
- 319: You report photosynthesis/respiration ratios but what about the influence of calcification? Is this not a very significant process for reef carbon biogeochemistry?
Authors response.
We thank Referee #2 and agree with them that calcification is important. We do discuss calcification in lines 77-78, 240-241 and 316 and we can expand on this if our article is accepted. However, the main purpose of our article is to present and draw attention to the air-sea CO2 fluxes and not the biogeochemistry of the coral reef at Eilat.
- 321: You often refer to other studies that have concluded reefs to be sources of CO2 to the atmosphere. However, all of these studies have taken place at other reef sites than the one you investigate here, so the construction of a “disagreement” between methods seems a bit artificial here, especially since you also suggest that all other methods should not be used, and instead all estimates should be done with EC. Can you compare EC measurements with results from other methods at the same site? The Heron Island reef, for instance, is indeed a sink for atmospheric CO2 when evaluated with the MAPCO2 mooring data and standard gas transfer parameterizations at that location.
Authors response.
We do make a direct comparison with Rey-Sánchez et al., (2017) who also used EC at Eilat and also found that the reef was a net sink of CO2. We discuss this work in our article in the Discussion. We did make extensive inquires with other researchers at The Inter-University Institute for Marine Sciences facility at Eilat, but their research did not crossover with our data collection period. We do not say ‘other methods should not be used’, but we do draw attention to their limitations when attempting to obtain representative ecosystem air-sea CO2 flux measurements. For example, they do not provide direct air-sea flux measurements. We agree with Referee #2 about the need for concurrent studies that compare direct measurements made by EC with other methods that centre on water chemistry to calculate pCO2 of the water overlying a coral reef and then apply a gas transfer velocity to calculate air – sea CO2 flux knowing the air pCO2. We hope that our article will be the catalyst for such research. It is interesting to learn that the MAPCO2 mooring data indicates that Heron Reef on the Great Barrier Reef is a sink for atmospheric CO2 which corroborates previous EC results presented by McGowan et al (2016) who we cite noting these did also show concurrent spatial variability in air-sea CO2 fluxes at Heron Reef with the reef flat acting as a source of CO2.
- 366 “whether coral reefs act as a source or sink” you are simplifying the question here – why shouldn’t different reefs look different in this respect? Some could be sinks and some could be sources, and the interesting questions are the details of what is driving those differences.
Authors response.
We appreciate the perspective of Referee #2, but we believe in the first instance there is an urgent need to establish through direct measurement using EC as we do how many coral reefs (type, locations etc) are sinks/sources of atmospheric CO2 as the source – sink debate continues. As we state near the beginning of our article knowledge of net CO2 exchanges with the atmosphere (NEE) over different ecosystems is a global research priority, i.e. it underpins the global FLUXNET program (Baldocchi et al., 2001) which we discuss in lines 28-31 (as well as many national programs). Our review of the literature shows that there remains overwhelming opinion that coral reefs are sources of CO2 to the atmosphere. The source – sink debate was reviewed recently by Kayanne, (2025) who we cite. Alongside this research that centres on the direct measurement of air-sea CO2 fluxes over coral reefs the details of what is driving CO2 fluxes can and should be investigated as suggested by Referee #2. Our future research planned at Eilat will take this approach if we can secure funding.
References
Rey-Sánchez, A.C., Bohrer, G., Morin, T.H., Shlomo, D., Mirfenderesgi, G., Gildor, H. and Genin, A.: Evaporation and CO2 fluxes in a coastal reef: an eddy covariance approach. Ecosyst. Health Sustain., 3(10), 1392830, DOI: 10.1080/20964129.2017.1392830, 2017.
Mahabbati, A., Beringer, J., Leopold, M., McHugh, I., Cleverly, J., Isaac, P. and Izady, A.: A comparison of gap-filling algorithms for eddy covariance fluxes and their drivers. Geosci. Instrum. Methods Data Syst., 10(1), 123-140, 2021.
Kayanne, H.: Thirty years since the coral reef CO2 sink/source debate. Galaxea, J. Coral Reef Stud., 27(1),118-130, 2025.
McGowan, H.A., MacKellar, M.C., and Gray, M.A.: Direct measurements of air‐sea CO2 exchange over a coral reef. Geophys. Res. Lett., 43(9), 4602-4608, 2016.
Moffat, A.M., Papale, D., Reichstein, M., Hollinger, D.Y., Richardson, A.D., Barr, A.G., Beckstein, C., Braswell, B.H., Churkina, G., Desai, A.R., and Falge, E.: Comprehensive comparison of gap-filling techniques for eddy covariance net carbon fluxes. Agric. For. Meteorol., 147(3-4), 209-232, 2007.
Nemitz, E., Mammarella, I., Ibrom, A., Aurela, M., Burba, G.G., Dengel, S., et al.: Standardisation of eddy-covariance flux measurements of methane and nitrous oxide. Int. Agrophys., 32(4), 517-549, 2018.
Falge, E., Baldocchi, D., Olson, R., Anthoni, P., Aubinet, M., Bernhofer, C., Burba, G., Ceulemans, R., Clement, R., Dolman, H. and Granier, A.: Gap filling strategies for defensible annual sums of net ecosystem exchange. Agric. For. Meteorol., 107(1), 43-69, 2001.
Blanckaert, A. C. A., Omanović, D., Fine, M., Grover, R., and Ferrier-Pagès, C.: Desert dust deposition supplies essential bio-elements to Red Sea corals. Glob. Chang. Biol., 00, 1–19. https://doi.org/10.1111/gcb.16074, 2022.
Baldocchi, D., Falge, E., Gu, L., Olson, R., Hollinger, D., Running, S., Anthoni, P., Bernhofer, C., Davis, K., Evans, R. and Fuentes, J.: FLUXNET: A new tool to study the temporal and spatial variability of ecosystem-scale carbon dioxide, water vapor, and energy flux densities. Bull. Am. Meteorol. Soc., 82(11), 2415-2434, 2001.
Citation: https://doi.org/10.5194/egusphere-2026-3467-AC2
-
AC2: 'Reply on RC2', Hamish McGowan, 05 Aug 2026
reply
Data sets
MicroMet_Eilat_UOE_3_Sept_20_2_Mar_21 – Updated H. McGowan et al. https://doi.org/10.48610/4a1b305
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 47 | 16 | 5 | 68 | 4 | 3 |
- HTML: 47
- PDF: 16
- XML: 5
- Total: 68
- BibTeX: 4
- EndNote: 3
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This study presents a case of long-term observations of air-sea CO₂ fluxes in coastal waters adjacent to coral reefs using the EC method, and provides valuable field data. Furthermore, the study’s conclusion that coral reefs can function as sinks for atmospheric CO₂ has the potential to highlight a new value—contributing to climate change mitigation—for the conservation of coastal ecosystems.
On the other hand, we believe there are several significant issues with the current analysis and discussion. The most critical point is that the discussion in this study relies heavily on flux observations obtained using the EC method and is not sufficiently supported by independent supplementary data, such as the pCO₂ in seawater or other carbonate system parameters.
Strictly speaking, the EC method directly measures the vertical turbulent transport of atmospheric CO₂ at the observation height. When applying this method to marine environments, the air-sea CO₂ fluxes are often more than an order of magnitude smaller than those in typical terrestrial ecosystems; consequently, the effects of water vapor fluxes, sensible heat fluxes, density variations, instrument response, and various corrections are expected to be relatively significant. Therefore, drawing conclusions about the mechanisms and scale of CO₂ uptake in coral reefs based solely on results from the EC method involves considerable uncertainty. At the very least, consistency with independent observational data—such as pCO₂ and/or DO in seawater—must be demonstrated.
Furthermore, the text states that CO₂ absorption equivalent to approximately half that of the coral reef area was observed in the surrounding waters. If this result is valid, should we not subtract the background absorption flux in the surrounding waters—or at least explicitly evaluate the difference between the two—when assessing the CO₂ absorption specific to the coral reef? If the current estimates include absorption occurring in the surrounding waters, the amount of CO₂ absorption attributable to the coral reefs themselves may be overestimated.
Assessments of environmental factors that could influence fluxes—such as sea surface skin temperature—also appear to be limited to qualitative discussions at present. If the contributions of each environmental factor are to be discussed, quantitative evaluations using appropriate statistical methods, such as multivariate regression, are necessary. For example, the study should demonstrate the extent to which wind speed, water temperature, air temperature, humidity, radiation, turbulence intensity, sea surface skin temperature, tidal level, and changes in wind direction and footprint contribute to the observed variations in CO₂ flux.
Furthermore, while this study presents the long-term observation period as one of its novel aspects, it is unclear what new insights were obtained by extending the observation period that were not available in previous studies. For example, it is necessary to explicitly state the results that could only be detected through long-term observations, such as seasonal variations, interannual variations, responses to extreme weather events, diurnal variations, or the temporal stability of flux differences between coral reef areas and surrounding waters. We do not believe that a long observation period alone constitutes sufficient novelty relative to previous research.
In light of the above points, we should exercise caution in concluding, based on the current analysis, that coral reefs are a sink for atmospheric CO₂, and in quantitatively discussing their absorption rates and controlling factors. Unless the authors provide clear and quantitative answers to the points raised above, it will be difficult to accept this paper at this stage.
Minor comments