Large CO2 seeps and hydrates field in the Indian Ocean (Mayotte Island)

doi:10.21203/rs.3.rs-4720763/v1

Download PDF

Physical Sciences - Article

Large CO₂ seeps and hydrates field in the Indian Ocean (Mayotte Island)

https://doi.org/10.21203/rs.3.rs-4720763/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

About 80% of Earth volcanic activity occurs underwater, releasing deep carbon to submarine environments and impacting Earth’s climate over geological timescales. The CO₂ emitted during submarine eruptions and/or hydrothermal degassing creates local ocean acidification, affecting the seawater carbonate equilibrium and oceanic ecosystems at large regional scales. Here, we report for the first time the existence of a major CO₂ hydrates field at the seafloor offshore Mayotte Island (Indian Ocean) associated with liquid CO₂ venting, following the submarine eruption that occurred in 2018. Using detailed acoustic surveys and in situ Raman spectroscopy, we reveal multiple hydrate mounds and seep zones distributed over an area of 0.06 km². We show that the gas seeps are mainly composed of CO₂, with minor contributions of CH₄ and H₂, with noble gas ratios and stable and radio-carbon isotopes clearly demonstrating their magmatic origin. Estimates of the CO₂ emitted over the entire area represent about 0.5% of the global magmatic carbon flux. Our discovery also suggests that CO₂ hydrates may potentially be stable at the seafloor at the right pressure-temperature conditions, bringing new prospects into CO₂ sequestration and decarbonization pathways in the ocean, in particular regarding kinetics of hydrates dissolution and environmental impacts.

Earth and environmental sciences/Ocean sciences/Marine chemistry

Earth and environmental sciences/Solid Earth sciences/Volcanology

Volcanism is the main pathway to release carbon stored in the Earth’s interior. In the preindustrial era, volcanic outgassing represented up to 90% Earth’s surface carbon emission¹. Current estimates suggest that most global volcanic outgassing occurs in submarine intraplate volcanic regions through diffuse emissions, the rest being distributed between mid-ocean ridges and subduction zones². Although current anthropogenic fluxes overshadow volcanic ones, the long-term fluxes of carbon from solid Earth to the atmosphere have been dominated by volcanic sources over most of Earth’s history with significant climatic impacts over geological timescales³. Massive and rapid volcanic CO₂ emissions to the atmosphere have led to severe global warming, ocean anoxia and acidification with lethal consequences for terrestrial and marine life⁴.

Our understanding of the magnitude of CO₂ fluxes emitted from volcanic and magmatic active regions on Earth continues to evolve⁵. Contribution of submarine volcanic sources to the global budget, including direct and diffuse emissions might be largely underestimated due to lack of direct observation, poor fluxes quantification and remote and challenging site locations^6,7. The eruption and creation of Fani Maoré volcano offshore Mayotte Island is, to date, the largest and deepest (3500 − 2700 m) deep-sea eruption to be studied both during and after the eruption⁸. As observed during most submarine volcanic eruptions, the Fani Maoré eruption released important quantities of magmatic volatiles^9,10 among which CO₂, H₂, and noble gases¹¹ along with metals and metalloids^11,12 with ambiguous impacts on adjacent and peripheral ecosystems in the deep ocean^11,13–15.

Very few occurrences of liquid CO₂ venting have been observed at the seafloor and, to our knowledge, there has not been any direct observation of CO₂-hydrates sitting on the seafloor in the deep ocean^16–18. CO₂ hydrates formed upon contact between seawater and CO₂-rich fluid bubbles have been reported in the Okinawa Trough as small horn-shaped pipes (10 cm) quickly washed away¹⁸, while most other observations have been inferred as embedded in sediment and/or associated with methane hydrates. While current industrial solutions for carbon dioxide sequestration via gas hydrates at the seafloor are being investigated^19,20, major questions remain regarding their short and long-term stability and potential impacts on surrounding seawater ecosystems, especially in case of CO₂ leakage²⁰. Direct evidence in the field is critically needed to provide ground truthing about formation kinetics of CO₂ hydrates, efficiency of hydrate caps or films against CO₂ dissolution and resilience of impacted ecosystems²¹.

Here we present the results of the GeoFLAMME sea expedition onboard RV Pourquoi Pas? offshore Mayotte Island (Comores Archipelago) following the Fani Maoré submarine eruption. Combining acoustic and water column data, sampling,video footage and in situ Raman spectroscopy, we provide evidence of a major submarine CO₂ hydrate field located in the Horseshoe area, 10 km east of Mayotte and 40 km west of the Fani Maoré volcanic edifice, associated with the release of mantle-derived liquid CO₂ likely driven by the eruption. We show that the generated CO₂ fluxes are equivalent to those from the most active volcanoes on Earth and result in significant local acidification, with impact on local benthic and pelagic ecosystems.

The Horseshoe edifice is a 4 km-wide collapsed volcanic cone. A major collapse resulted in the formation of a 2 km-wide depression at its center that opens to the north. Its crest is marked by a sharp and well-defined U-shaped limit: the Horseshoe’s rim²². The Horseshoe area lies on the submarine flank of Mayotte in the depth range of 1200–1600 m, aligned along a west-northwest–east-southeast-trending volcanic ridge, with Fani Maoré at its easternmost tip²². The seafloor in the Horseshoe area is characterized by pumice mixed bioclastic-volcanoclastic content, including fresh phonolitic lava and bomb rims, confirming the volcanic origin and past activity of the edifice^23,24.

Since 2019, water column acoustic surveys using vessel hull multibeam echosounder were performed within the REVOSIMA monitoring program, allowing for the extensive mapping of the Horseshoe area, and the identification of vigorous acoustic plumes. Indeed, we discovered two active fluid emission sites in the Horseshoe area in May 2019, and by May 2021, that number had reached 15. These focused fluid flows cover an area of 0.063 km² over the ≈ 40 km² of the Horseshoe structure. They are mainly distributed over seven major zones, most around the Horseshoe rim, but also within the depression and at the eastern outside corner. (Fig. 1).

In spring 2021, the remotely operated vehicle (ROV) Victor 6000 revealed that each active site hosted a few to a hundred distinct fluid outlets (Table 1). Droplets rising from the seafloor were seen both with and without a milky skin. In addition, ROV images showed extensive presence of white milky patches and mounds either directly laying on the seafloor or inserted into crevices (Fig. 2). During an extensive survey in active site B0 (Fig. 1C), we counted at least 128 white mounds of different sizes and shapes, ranging from a few centimeters to 5 m high and up to ~ 1 m wide (SI video 1). Most grew in an almost columnar-like structure. The biggest mounds presented multiple liquid droplet streams rising from their bases and/or summits into the water column (Fig. 3G-J). On the Horseshoe rim structure, and along the ridge, steady stream and/or pulsing burst of droplets discharges were observed. Droplets escaped from open faults and cracks within the cliffs, and directly from holes in the gravel on the seafloor at the center depression of the Horseshoe area (Fig. 2). Filamentous microbial mats coated with oxidized iron compounds were observed in zone A, mostly between sites A1 and A2 (Fig. 2H). Flow rates were highly variable from one vent to another (Extended Data Table 1, Supplementary Table 1).

Table 1

Composition, isotopic signature and fluxes of CO₂ seeps in the Horseshoe area
A. Gas fluxes and composition in the Horseshoe area
Site		C4	C4	B0	B0
Samples		GFL-PL778-09-PGZ01 #1	GFL-PL778-09-FLU1	GFL-PL779-10-PGZ03-#4	GFL-PL783-14-PGZ01-#12
Date of sampling		08/05/2021	08/05/2021	11/05/2021	21/05/2021
Type of sampler		PEGAZ	Gas tight syringe	PEGAZ	PEGAZ
Latitude		12°49,8827	12°49,8827	12°49,8275	12°49,8751
Longitude		45°23,0572	45°23,0572	45°22,7816	45°22,8430
Depth	m	1293	1293	1369	1357
Temperature recorded during sampling	°C	4–8	4–8	8.52	4.8–9.05
CO₂	%	98.89	98.04	94.70	98.93
CH₄	%	0.849	0.594	0.462	0.8
H₂	%	0.0012	0.0009	0.116	0.111
N₂	%	0.232	1.11	3.73	0.139
O₂	%	0.025	0.255	0.99	0.017
Rc/Ra	†	6.9	-	6.5	6.4
δ¹³C-CO₂	‰ vs PDB	-4.25	-	-4.5	-4.6
Δ¹⁴C-CO₂	‰	-995	-990	-998	-995
†R as ³He/⁴He in a sample; Ra as ³He/⁴He in air - see 31

B. Flux estimates
Type of flow rates	Measured CO₂ flow rates (avg ± std, n)	Number of fluid outlets in site B0	Estimated CO₂ fluxes * (avg ± std)
	ml s^− 1		tC y^− 1
weak	5.7 ± 3.7 (3)	23	1.22 ± 1.01–10⁴
medium	18.6 ± 7.7 (12)	119	2.79 ± 1.82 10⁵
high	65.2 ± 32.3 (7)	17	1.40 ± 0.99 10⁵
All types extrapolated over the entire Horseshoe Area			3.84 ± 2.59 10⁵
* with density of liquid CO₂ of 0.98 kg m^− 3
C. Comparison with other systems
Volcanic systems	Ref.	Estimated CO₂ fluxes (avg ± std)	Contribution of Horseshoe
		GtC y^− 1	%
Horseshoe area	this study	3.84 ± 2.59 10^− 4
Magmatic inputs
MORB systems	²	0.016	2.4
All Magmatic C fluxes	^2,44	0.072–0.079	0.49–0.53
Volcanic inputs
El Hierro submarine volcano	⁴⁵	2.19 10^− 4	253.4
MOR systems	^1,44	0.0264–0.097	0.2–0.8
All volcanic C fluxes	¹	0.64	0.06

The liquid droplets were recovered and stored under pressure using the Pegaz device²⁵ for onshore analysis by gas chromatography²⁶. They consist mainly of CO₂, 97.6 ± 2.0% v/v, with CH₄ as a minor secondary component, 0.7 ± 0.2% v/v. Similar to the liquid CO₂ previously observed at Champagne vent (NW Eifuku)¹⁶, the droplets stuck to the ROV like clumps of grapes, and did not coalesce into larger droplets. The white patchy and mound structures consisted of a layer of milky-skin liquid CO₂ droplets. The film coating the droplets is likely CO₂ hydrates, which form from the interaction of CO₂ and water within the CO₂ hydrate stability zone conditions. The horseshoe area is well inside the stability conditions for both pure liquid CO₂ in pure water (387 m at 4°C) and pure CO₂ hydrates (200 m depth at 4°C). Presence of methane (or other chemicals) dissolved into the liquid CO₂ as well as the salinity may shift these melting points to slightly shallower depths. The mounds appear to be a more advanced stage of the patchy structures since they exhibit a hard core, which we hypothesize originates from the growth of the thin hydrate layer surrounding the liquid CO₂ droplet into bulk hydrate crystals.

Site B0 (Fig. 1) is one of the two seeping sites early detected in 2019 and the more extended in surface area. We therefore selected it for a detailed investigation of hydrates and CO₂ fountain morphology. CO₂ droplets presented different sizes and shapes: single droplets, droplets covered with a hydrate film, slender droplet formation or even tubular hydrates (Fig. 3, Supplementary Video 1), similar to what was observed at the JADE site in the Okinawa Trough¹⁸ and varied according to flowrate, seep diameter and temperature, among other factors²⁷. Ambient seawater temperature was ~ 4.3 ⁰C compared to the fluid coming out of the seafloor of ~ 9 ⁰C (Fig. 3E). All temperature measurements were made with the temperature probe in direct contact with the hydrate without drilling any holes. Temperature was not homogeneous along the hydrate surface at both mounds studied. Hydrate formation is an exothermic process, so active hydrate formation may locally increase the temperature of the hydrate surface.

We used in situ Raman spectroscopy to confirm the presence of CO₂ in a clathrate environment (not only under liquid form)^28,29 and to characterize the chemical composition of four mounds of different sizes chosen over a broad area to assess differences in the chemical composition of fluid emissions in the Horseshoe area. Chemical selectivity of gas hydrates is well known and their presence could modify the chemical composition of fluids reaching the water column. We measured temperature and Raman spectra at three different heights and inside a hole drilled in the mound using the ROV arms. The two spectra in Fig. 3E show the Raman signature of the Horseshoe seawater (shaded spectrum) and the white mounds (orange line). The two very strong vibrational bands C and D (Fermi resonance) indicate the presence of CO₂ in the white mounds. Corresponding to the OH bending mode of water, the vibrational band F is very intense in seawater but very weak in the solid sample. The difference in shape of the vibrational bands H and I (symmetric and asymmetric OH stretches) indicates that water molecules do not have the same structure in both samples, being compatible with the presence of water under a clathrate structure for the white mound sample. In situ Raman spectrum supports our visual observations: the white mounds are formed by liquid CO₂ in equilibrium with CO₂ hydrates and there is no evidence of methane or other gases entrapped in the gas hydrates, although presence of methane traces cannot be excluded given the sensitivity of the in situ Raman spectrometer. This is to our knowledge the first observation of natural CO₂ hydrates mounds on the seafloor.

Emitted fluids in the Horseshoe area generated small turbidity anomalies in the water column but high concentrations of ³He, CH₄, CO₂ and H₂ and large decreases in Eh, pH and alkalinity compared to ambient seawater (Fig. 1, Extended Fig. 1, Supplementary Video 2). Fluid composition indicates that magmatic/hydrothermal gases are composed primarily of CO₂ and CH₄ with small contribution of H₂ (Table 1A), consistent with emissions at most other volcanoes^9,30–32 and in subaerial volcanic gasses from Mayotte Petite Terre³³. The isotopic composition of CO₂ indicates a typical mantle³⁴ signature (δ¹³C-CO₂ = -3.7 ± 0.2‰, Δ¹⁴C_CO2 < -990‰). Since no molten lava is currently emitted at the seafloor in the Horseshoe area, H₂ likely originates from a magmatic/hydrothermal source³⁵ formed at depth along with CO₂ and dissolved within. Emitted fluids show very little variability with ³He/⁴He ratios (corrected for air contamination and expressed as Rc/Ra) of 6.62 ± 0.28 Ra, comparable to the values reported from gas emissions on Mayotte Petite Terre³³ and mantle xenoliths of Grande Comore³⁶. This suggests a common magmatic source at depth likely aligned along an old fracture zone oriented N130 with potential secondary storage in gas saturated rocks located below the Horseshoe area³⁷.

Very few data are available for deep-sea habitats in the area, but this part of the Mozambique Channel is considered as a biodiversity hotspot³⁸. To assess the impact of the CO₂ seeps on local and regional benthic ecosystems, we mapped the sessile megafaunal organisms present in the Horseshoe area during two ROV dives. Altogether, at least 23 morphotypes of cnidarians (class Anthozoa) were reported belonging to at least 5 orders (Extended Data Table 2). Identification from images in a newly explored area severely limits our ability to assess the actual biodiversity³⁹. but our first observations are consistent with previous findings at shallower depths along the eastern slopes of Mayotte Island³⁹. Anthozoans are abundant in and out of the Horseshoe area while sponges, belonging to Desmospongia and Hexactinellidae, are mainly found in the northern part and along the eastern ridge outside of the Horseshoe area (Fig. 4).

Most abundant morphotypes belonging to Octocorallia are identified as Alcyonacea gorgonian-like individuals which represented about almost half of all observed morphotypes (Extended Data Table 2). Few Actinaria are spotted close to the B and G active seeping zones (Fig. 4; Extended Data Fig. 2–4). Most individuals are found on the rim and cliffs of the Horseshoe structure, the base and centers being almost devoid of megafauna. We observe a significantly higher proportion of dead anthozoans close to the seep sites compared to “background” areas (Kruskal-Wallis chi-squared = 4.86, df = 1, p-value = 0.02749). The highest record of dead corals occurs in active seeping zones A and C, while zones B and G are almost devoid of sessile fauna. The little fauna visible corresponds to unidentified individuals covered by layers of sediment or microbial mats and hardly visible white fragments (supplementary material). Although individual species response may vary^40,41, the emitted fluids and resulting ~ 1 pH unit decrease observed in the water column within the Horseshoe area have indisputable impacts on the local deep-sea ecosystems, in particular for Scleratinian species. Deleterious effect of ocean acidification on cold-water corals but also adjacent planktonic, and even terrestrial, ecosystems have been demonstrated^4,42,43 and are therefore to be expected offshore Mayotte. Coral death related to a decrease in calcification rate can lead to major shifts in benthic communities⁴¹, and eventually a loss in biodiversity. Mechanisms underlying changes in deep-sea ecosystem composition related to ocean acidification and submarine CO₂ release remain to be elucidated, and the Horseshoe area offers the ability to study potential remediation solutions.

Quantitative estimates of the liquid CO₂ emissions on site B0 based on ROV video survey show that CO₂ fluxes represent 1.0 ± 0.10^− 4 GtC y^− 1. Extrapolated to the entire Horseshoe area by hypothesizing that all the other 15 sites distributed over the seven major active zones display similar fluxes and seep densities, this leads to 3.84 ± 2.59 10^− 4 GtC y^− 1 (Table 1B, Extended Data Table 3). Considering that the Horseshoe area is an active open window into local mantle C outputs³³, from a global perspective these fluxes amount about 2% of all Mid Ocean Ridges Basalts (MORB) and 0.5% of all magmatic fluxes^2,44. While this may represent only 0.06 ± 0.04% of all volcanic fluxes¹ (Table 1C), several aspects need to be considered. 1) This is about twice the estimated fluxes emitted during the El Hierro eruption in 2011⁴⁵. 2) The number of seep sites have been steadily increasing since 2019 and those fluxes therefore represent continuous ones rather than short-lived eruption-based emissions. 3) Consequently, we hypothesize that CO₂ hydrates have been forming ever since. These estimates do not take into account the standing reservoir stored as CO₂ hydrates, which alone at site B0 may represent 5 tC based on the 128 hydrates observed occurrences, with a median volume of 0.5m³ and a density⁴⁶ of 1.1. Aside from its contribution to the regional carbon cycle and non-negligible C inputs, the Horseshoe area and its CO₂ gas hydrate field constitutes an unprecedented opportunity to study CO₂ sequestration pathways through gas hydrate formation. Such a large occurrence of naturally forming CO₂ hydrates may allow further studies regarding short- and long-term stability, kinetics of hydrates formation and dissolution in seawater²⁰. In addition, monitoring liquid CO₂ vents may provide critical information regarding the environmental impacts of CO₂ leakage, in particular the ability of biological species to develop and adapt to low and variable pH conditions, and provide possible remediation solutions, or at least assess ecosystem vulnerability in natural conditions. High storage capacity and hydrate cap formation preventing CO₂ leaks make CO₂ hydrates attractive vehicles for deep ocean carbon sequestration^47,48. However, observations of Horseshoe hydrates clearly indicate concomitant hydrate formation and CO₂ seeping, questioning the efficiency and safety of potential hydrate self-sealing process for preventing leakage as previously suggested in deep-sea sediment carbon sequestration studies^49,50. A long-term seafloor observatory currently under construction in the Horseshoe area may constitute a first step to ground truth theoretical considerations regarding CO₂ hydrate decarbonization pathway. In addition, our discovery emphasizes the critical need to better understand ocean subseafloor processes, both in terms of fluid pathways, permeability and potential of CO₂ degassing, and in terms of relationships between tectonic/seismic and magmatic/thermal activity that might trigger and drive such CO₂ degassing events, locally affecting the ocean carbon cycle at multiple levels.

Water column acoustic data were collected with a dedicated acquisition protocol by vessel-hull mounted multibeam echosounders during the MAYOBS⁵¹ cruises in 2019, 2020 and 2021 (Kongsberg™ EM122 12 kHz), and during the GEOFLAMME⁵² cruise in 2021 (RESON™ 7150 24 kHz)., Data were post-processed with SonarScope (https://doi.org/10.17882/87777) and GLOBE (https://doi.org/10.17882/70460) software for the identification and location of the acoustic plumes⁵³.

A Sea-Bird Electronics™ CTD (SBE911plus) was used together with a rosette water sampler (SBE-32) equipped with 24 OTE 10L bottles for hydrographic measurements. For optical backscatter measurements a SeaPoint™ nephelometer were attached to the CTD probe. Upon CTD/rosette recovery, bottles were immediately sampled for gas, pH, silicates and total alkalinity analysis on board. pH was measured over 10ml samples using a Metrohm pH-meter, while total alkalinity measurements were performed over 10–30 ml water samples using Methrom titrimeter.

Gases analysis from hydrocast operations were performed onboard directly after sampling from unpoisoned samples. Hydrogen and Carbon dioxide concentrations were determined by headspace technique with HID detection⁵⁴ while Methane concentrations were determined by purge and trap technique coupled with GC-FID detection⁵⁵.

During the GEOFLAMME cruise, gas seeping from the vents (consisting mainly of liquid CO₂) was collected using the PEGAZ gas-bubble sampler²⁵. In order to estimate seep fluid discharge and associated fluxes, 3D reconstruction of the seafloor based on ROV images, along with seep fluid counting and sampling were performed over zone B, specifically on site B0, one of the eldest and most vigorous sites. Immediately after recovery, the cylinder (50ml) of the PEGAZ sampler was positioned on a titanium cell equipped with a high pressure sensor and connected to a gas extractor for subsampling²⁶. Once the vacuum has been achieved throughout the system, the Pegaz cylinder was first opened to the cell to evaluate the initial pressure. It was afterwards gradually opened to the gas extractor in order to expand, dry and subsample gases into vacuumed stainless-steel canisters of 50 to 1000 ml capacity equipped with gas-tight valves. Note that at ambient atmospheric pressure and temperature, liquid CO₂ decompresses quickly in the gas extractor and changes phase to a gas. The residual gas remaining in the extraction line was injected on board into an SRA Instruments ® micro-chromatograph for gas analysis²⁶. Aliquots of extracted gases were also recovered in copper tubes for onshore analysis of helium and δ¹³C_CO2 isotopes at INGV^30,33. To allow for a reasonable (and safe) CO₂ decompression in the extraction line, only one or two droplets of gases were collected with the PEGAZ sampler.

Radiocarbon analyzes were carried out on CO₂ droplets collected from the seeps around the hydrate field. The gas was collected, purified then converted to graphite⁵⁶ before being measured at the Artemis LMC14 AMS facility⁵⁷.

In situ Raman spectroscopy on hydrates mounds.

Visual recognition and identification of the white mounds was performed using the ROV video transects GFL-PL783-14 and GFL-PL785-16. Raman spectra were recorded using a “custom-made” spectrometer for in situ measurements named Ramses and mounted on the ROV^58,59 It is equipped with a Horiba Jobin Yvon axial spectrometer and can perform real-time Raman spectroscopy on liquids, gases, and solids at depths up to 4800 m. Housed in Titanium, it features a 600 gr/mm grating for a spectral resolution of 10 cm^− 1 and is equipped with two lasers (532 nm and 691 nm) for different sample types. It is coupled to an Andor DU440 CCD sensor.

Fluid flow rates estimates

Three ROV dives (GFL-ROV-PL776-07, GFL-ROV-PL778-09; GFL-ROV-PL785-16) were visually inspected to map, count and classify seep outlets of every active site in the entire Horseshoe area. A dedicated survey was performed in site B0 to quantify flow rates: a small funnel (530ml total volume) with volumetric graduation marks was used to measure flow rates on each event noted. The funnel was also deployed on site C1, D1, E0 and G0. Observed seeps were then classified into groups as a function of hydrate presence or absence and of liquid CO₂ flow rates low flow rate < 10 ml s^− 1, medium flow rates between 10 and 40 ml s^− 1 and high flow rates > 40 ml s^− 1. We calculated the number of seeps per category, average and standard deviation of the flow rate for each category. Density of seeps on site B0 was then calculated for each flow rate category and extrapolated over the entire Horseshoe area by assuming the same density repartition at all seeping sites than in site B0. Error estimate was set to 50% to account for site disparities. All annotated events and associated description and flux information are provided as supplementary material.

Megafauna mapping

Recognition and identification of megafauna was performed using the ROV videos acquired during GFL-PL07, GFL-PL16, and transects dedicated to the 3D reconstruction of site B. The Victor6000 has multiple cameras but because video acquisition was not set up for habitat mapping, the downward-looking camera could not be exploited properly. For this study we thus used the main ROV camera, and both the downward-looking and main cameras for site B. Images were analyzed using the ADELIE video annotation software using 3 criteria “sessile fauna category” (including dead Anthozoa), “organisms’ density” and “substrata type”. Because of the uncertainty in the identification of gorgonian- and coral-like morphotypes, these observations were grouped under the ‘undetermined Anthozoa’ modality (Extended Data Table 2). The definition of dead Anthozoa was based on the absence of color but also on the presence of living organisms around. For example, white corals observed among coloured corals or gorgonians were not considered as dead. Conversely, isolated colourless corals surrounded by broken sessile organisms covered by bacterial mats were annotated as ‘dead’. During the annotation process, associated metadata (timecode, image name, geographical coordinates) were recorded with ADELIE software and compilation was performed on ArcGIS software. The opportunistic nature of this dataset did not allow for the acquisition of quantitative data and each faunal record was thus processed as one observation. To assess the impact of liquid CO₂ on faunal distribution, the mean proportion of each taxa, and ‘dead’ vs living corals, between active sites and background areas was compared using a Kruskall-Wallis test.

Competing interests

The authors declare no competing interests

Author contributions

C. Cathalot, E. Rinnert, C. Scalabrin, O. Fandino, N. Feuillet, designed the work, lead the acquisition, analysis and interpretation of all data along with drafting the ms. H. Ondreas, acquired and interpreted CO₂ fluxes data. T.Giunta, O.Rouxel, M. Mastin., V. Chavagnac, A. L. Rizzo, J.P.Donval, V. Guyader, M. Manoux acquired, analyzed and interpreted chemical and isotopic fluid data. C.Rabouille, J.P. Dumoulin., B.Bombled acquired and interpreted ¹⁴C data. S. Walker provided the MapR material and data interpretation. M. Tardivel, E. Prado, M. El Rawke acquired and interpreted in situ RAMAN data. G. Page and M. Matabos analyzed and interpreted all biological data. All authors contributed significantly to the ms and approved the submitted version.

Acknowledgements

We thank all the participants from the GEOFLAMME and MAYOBS sea expeditions, for their amazing work and support. We also thank the captains and crew members from R/V Pourquoi Pas?, along with the AUV and ROV Victor 6000 team (GENAVIR) crew members. We thank Delphine Pierre and Anne-Sophie Alix for their help in generating the Horseshoe map. This research was supported by the Mayotte volcanological and seismological monitoring network (REVOSIMA, doi: 10.18715/MAYOTTE.REVOSIMA), by the French Oceanographic Fleet, by TotalEnergies (TOTAL ENERGIES FR000063751 //876 IFREMER 20/1001730) and by ISblue project, Interdisciplinary graduate school for the blue planet (ANR-17-EURE-0015) and co-funded by a grant from the French government under the program "Investissements d'Avenir" embedded in France 2030. We are finally grateful to Mariano Tantillo and Igor Oliveri, who carried out helium and carbon isotopic analyses at INGV Palermo laboratories.

Burton, M., Sawyer, G. & Granieri, D. Deep Carbon Emissions from Volcanoes. Reviews in Mineralogy and Geochemistry 75, 323–354, doi:10.2138/rmg.2013.75.11 (2013).
Plank, T. & Manning, C. E. Subducting carbon. Nature 574, 343–352, doi:10.1038/s41586-019-1643-z (2019).
Tolstoy, M. Mid-ocean ridge eruptions as a climate valve. Geophysical Research Letters 42, 1346–1351, doi:10.1002/2014gl063015 (2015).
Cui, Y., Li, M., van Soelen, E. E., Peterse, F. & Kürschner, W. M. Massive and rapid predominantly volcanic CO < sub > 2</sub > emission during the end-Permian mass extinction. Proceedings of the National Academy of Sciences 118, e2014701118, doi:doi:10.1073/pnas.2014701118 (2021).
Lizarralde, D., Soule, S. A., Seewald, J. S. & Proskurowski, G. Carbon release by off-axis magmatism in a young sedimented spreading centre. Nature Geoscience 4, 50–54, doi:10.1038/ngeo1006 (2011).
Werner, C. et al. in Deep Carbon: Past to Present (eds Beth N. Orcutt, Isabelle Daniel, & Rajdeep Dasgupta) 188–236 (Cambridge University Press, 2019).
Orcutt, B. N., Daniel, I., Dasgupta, R., Crist, D. T. & Edmonds, M. in Deep Carbon: Past to Present (eds Beth N. Orcutt, Isabelle Daniel, & Rajdeep Dasgupta) 1–3 (Cambridge University Press, 2019).
Feuillet, N. et al. Birth of a large volcanic edifice offshore Mayotte via lithosphere-scale dyke intrusion. Nature Geoscience 14, 787–795, doi:10.1038/s41561-021-00809-x (2021).
Resing, J. A. et al. Active submarine eruption of boninite in the northeastern Lau Basin. Nature Geoscience 4, 799–806, doi:10.1038/ngeo1275 (2011).
Stucker, V. K., de Ronde, C. E. J., Laurence, K. J. & Phillips, A. M. Rare Time Series of Hydrothermal Fluids for a Submarine Volcano: 14 Years of Vent Fluid Compositions for Brothers Volcano, Kermadec Arc New Zealand. Economic Geology, doi:10.5382/econgeo.4922 (2022).
Mastin, M. et al. Strong geochemical anomalies following active submarine eruption offshore Mayotte. Chemical Geology 640, 121739, doi:https://doi.org/10.1016/j.chemgeo.2023.121739 (2023).
Rubin, K. Degassing of metals and metalloids from erupting seamount and mid-ocean ridge volcanoes: Observations and predictions. Geochim. Cosmochim. Acta 61, 3525–3542, doi:https://doi.org/10.1016/S0016-7037(97)00179-8 (1997).
Gómez-Letona, M., Arístegui, J., Ramos, A. G., Montero, M. F. & Coca, J. Lack of impact of the El Hierro (Canary Islands) submarine volcanic eruption on the local phytoplankton community. Scientific Reports 8, 4667, doi:10.1038/s41598-018-22967-6 (2018).
Mittal, T. & Delbridge, B. Detection of the 2012 Havre submarine eruption plume using Argo floats and its implications for ocean dynamics. Earth and Planetary Science Letters 511, 105–116, doi:https://doi.org/10.1016/j.epsl.2019.01.035 (2019).
Santana-Casiano, J. M. et al. The natural ocean acidification and fertilization event caused by the submarine eruption of El Hierro. Scientific Reports 3, 1140, doi:10.1038/srep01140 (2013).
Lupton, J. et al. Submarine venting of liquid carbon dioxide on a Mariana Arc volcano. Geochemistry, Geophysics, Geosystems 7, Q08007, doi:10.1029/2005GC001152 (2006).
Lupton, J. et al. Venting of a separate CO2-rich gas phase from submarine arc volcanoes: Examples from the Mariana and Tonga-Kermadec arcs. Journal of Geophysical Research: Solid Earth 113, B08S12, doi:10.1029/2007JB005467 (2008).
Sakai, H. et al. Venting of Carbon Dioxide-Rich Fluid and Hydrate Formation in Mid-Okinawa Trough Backarc Basin. Science 248, 1093–1096 (1990).
Park, Y. et al. Sequestering carbon dioxide into complex structures of naturally occurring gas hydrates. Proceedings of the National Academy of Sciences 103, 12690–12694, doi:doi:10.1073/pnas.0602251103 (2006).
Zheng, J., Chong, Z. R., Qureshi, M. F. & Linga, P. Carbon Dioxide Sequestration via Gas Hydrates: A Potential Pathway toward Decarbonization. Energy & Fuels 34, 10529–10546, doi:10.1021/acs.energyfuels.0c02309 (2020).
Shindo, Y., Lund, P. C., Fujioka, Y. & Komiyama, H. Kinetics of formation of CO2 hydrate. Energy Conversion and Management 34, 1073–1079, doi:https://doi.org/10.1016/0196-8904(93)90055-F (1993).
Puzenat, V. et al. Volcano-tectonic structures of Mayotte’s upper submarine slope: insights from high-resolution bathymetry and in-situ imagery from a deep-towed camera. C. R. Geosci. Online first (2022), 1–24, doi:10.5802/crgeos.175 (2022).
Thivet, S. et al. Volatiles of the active Mayotte volcanic chain: STA & EGA-MS analysis of volcanic products. Chemical Geology 618, 121297, doi:https://doi.org/10.1016/j.chemgeo.2022.121297 (2023).
Paquet, F. et al. in ASF 2022 - XVIIIème Congrès Français de Sédimentologie (Brest, France, 2022).
Ruffine, L. et al. Gas Seepage along the Edge of the Aquitaine Shelf (France): Origin and Local Fluxes. Geofluids 2017, 4240818, doi:10.1155/2017/4240818 (2017).
Charlou, J. L. et al. Physical and chemical characterization of gas hydrates and associated methane plumes in the Congo–Angola Basin. Chemical Geology 205, 405–425, doi:https://doi.org/10.1016/j.chemgeo.2003.12.033 (2004).
Hirai, S. & Sanda, H. Injection and boiling of liquid CO2 with a hydrate coating. Energy 30, 2275–2283, doi:https://doi.org/10.1016/j.energy.2003.10.020 (2005).
Li, L. et al. In Situ Raman Spectral Characteristics of Carbon Dioxide in a Deep-Sea Simulator of Extreme Environments Reaching 300 ℃ and 30 MPa. Applied Spectroscopy 72, 48–59, doi:10.1177/0003702817722820 (2018).
Zhang, X. et al. In Situ Raman Detection of Gas Hydrates Exposed on the Seafloor of the South China Sea. Geochemistry, Geophysics, Geosystems 18, 3700–3713, doi:https://doi.org/10.1002/2017GC006987 (2017).
Rizzo, A. L. et al. Geochemistry of CO2-Rich Gases Venting From Submarine Volcanism: The Case of Kolumbo (Hellenic Volcanic Arc, Greece). Frontiers in Earth Science 7, doi:10.3389/feart.2019.00060 (2019).
Resing, J. A. et al. Chemistry of hydrothermal plumes above submarine volcanoes of the Mariana Arc. Geochemistry, Geophysics, Geosystems 10, doi:10.1029/2008gc002141 (2009).
Baumberger, T. et al. Dissolved Gas and Metal Composition of Hydrothermal Plumes From a 2008 Submarine Eruption on the Northeast Lau Spreading Center. Frontiers in Marine Science 7, doi:10.3389/fmars.2020.00171 (2020).
Liuzzo, M. et al. Gas Geochemistry at Grande Comore and Mayotte Volcanic Islands (Comoros Archipelago), Indian Ocean. Geochemistry, Geophysics, Geosystems 22, e2021GC009870, doi:https://doi.org/10.1029/2021GC009870 (2021).
Mason, E., Edmonds, M. & Turchyn, A. V. Remobilization of crustal carbon may dominate volcanic arc emissions. Science 357, 290–294, doi:doi:10.1126/science.aan5049 (2017).
Worman, S. L., Pratson, L. F., Karson, J. A. & Schlesinger, W. H. Abiotic hydrogen (H < sub > 2) sources and sinks near the Mid-Ocean Ridge (MOR) with implications for the subseafloor biosphere. Proceedings of the National Academy of Sciences 117, 13283–13293, doi:doi:10.1073/pnas.2002619117 (2020).
Bordenca, C. V. et al. Geochemical evidence for a lithospheric origin of the Comoros Archipelago (Indian Ocean) as revealed by ultramafic mantle xenoliths from La Grille volcano. Lithos 462–463, 107406, doi:https://doi.org/10.1016/j.lithos.2023.107406 (2023).
Foix, O. et al. Offshore Mayotte volcanic plumbing revealed by local passive tomography. Journal of Volcanology and Geothermal Research 420, 107395, doi:https://doi.org/10.1016/j.jvolgeores.2021.107395 (2021).
Obura, D. The Diversity and Biogeography of Western Indian Ocean Reef-Building Corals. PLOS ONE 7, e45013, doi:10.1371/journal.pone.0045013 (2012).
Hanafi-Portier, M. et al. When Imagery and Physical Sampling Work Together: Toward an Integrative Methodology of Deep-Sea Image-Based Megafauna Identification. Frontiers in Marine Science 8, doi:10.3389/fmars.2021.749078 (2021).
Form, A. U. & Riebesell, U. Acclimation to ocean acidification during long-term CO2 exposure in the cold-water coral Lophelia pertusa. Global Change Biology 18, 843–853, doi:10.1111/j.1365-2486.2011.02583.x (2012).
Comeau, S. et al. pH variability at volcanic CO2 seeps regulates coral calcifying fluid chemistry. Global Change Biology 28, 2751–2763, doi:https://doi.org/10.1111/gcb.16093 (2022).
Maier, C., Hegeman, J., Weinbauer, M. G. & Gattuso, J. P. Calcification of the cold-water coral Lophelia pertusa under ambient and reduced pH. Biogeosciences 6, 1671–1680 (2009).
Fraile-Nuez, E. et al. The submarine volcano eruption at the island of El Hierro: physical-chemical perturbation and biological response. Scientific reports 2, 486–486, doi:10.1038/srep00486 (2012).
Marty, B. & Tolstikhin, I. N. CO2 fluxes from mid-ocean ridges, arcs and plumes. Chemical Geology 145, 233–248, doi:https://doi.org/10.1016/S0009-2541(97)00145-9 (1998).
Santana-Casiano, J. M. et al. Significant discharge of CO2 from hydrothermalism associated with the submarine volcano of El Hierro Island. Scientific Reports 6, 25686, doi:10.1038/srep25686 https://www.nature.com/articles/srep25686#supplementary-information (2016).
Aya, I., Yamane, K. & Nariai, H. Solubility of CO2 and density of CO2 hydrate at 30 MPa. Energy 22, 263–271, doi:https://doi.org/10.1016/S0360-5442(96)00093-X (1997).
Zhao, G. et al. Effects of hydrate cap on leakage prevention and capacity improvement of sub-seabed CO2 sequestration. Chemical Engineering Journal 450, 138493, doi:https://doi.org/10.1016/j.cej.2022.138493 (2022).
Fahed Qureshi, M. et al. Laboratory demonstration of the stability of CO2 hydrates in deep-oceanic sediments. Chemical Engineering Journal 432, 134290, doi:https://doi.org/10.1016/j.cej.2021.134290 (2022).
Teng, Y. & Zhang, D. Long-term viability of carbon sequestration in deep-sea sediments. Science Advances 4, eaao6588, doi:10.1126/sciadv.aao6588.
Brewer, P. G., Friederich, G., Peltzer, E. T. & Orr, F. M. Direct Experiments on the Ocean Disposal of Fossil Fuel CO2. Science 284, 943–945, doi:10.1126/science.284.5416.943 (1999).
Rinnert, E. et al. (2019).
Cathalot, C., Rinnert, E. & Feuillet, N. (R/V Pourquoi Pas?, 2021).
Scalabrin, C. Site d'émissions de fluides, Mayotte, zone Fer à Cheval (C. Scalabrin, 2022). (Ifremer GEO-OCEAN, 2023).
Donval, J. P. & Guyader, V. Analysis of hydrogen and methane in seawater by “Headspace” method: Determination at trace level with an automatic headspace sampler. Talanta 162, 408–414, doi:http://dx.doi.org/10.1016/j.talanta.2016.10.034 (2017).
Charlou, J.-L. & Donval, J.-P. Hydrothermal methane venting between 12°N and 26°N along the Mid-Atlantic Ridge. Journal of Geophysical Research: Solid Earth 98, 9625–9642, doi:10.1029/92JB02047 (1993).
Dumoulin, J. P. et al. Status Report on Sample Preparation Protocols Developed at the LMC14 Laboratory, Saclay, France: From Sample Collection to 14C AMS Measurement. Radiocarbon 59, 713–726, doi:10.1017/RDC.2016.116 (2017).
Moreau, C. et al. ARTEMIS, THE 14C AMS FACILITY OF THE LMC14 NATIONAL LABORATORY: A STATUS REPORT ON QUALITY CONTROL AND MICROSAMPLE PROCEDURES. Radiocarbon 62, 1755–1770, doi:10.1017/RDC.2020.73 (2020).
Péron, O., Rinnert, E., Colas, F., Lehaitre, M. & Compère, C. First Steps of in situ Surface-Enhanced Raman Scattering during Shipboard Experiments. Applied Spectroscopy 64, 1086–1093, doi:10.1366/000370210792973505 (2010).
Ruffine, L. et al. Multidisciplinary investigation on cold seeps with vigorous gas emissions in the Sea of Marmara (MarsiteCruise): Strategy for site detection and sampling and first scientific outcome. Deep Sea Research Part II: Topical Studies in Oceanography 153, 36–47, doi:https://doi.org/10.1016/j.dsr2.2018.03.006 (2018).

There is NO Competing Interest.

SupplementaryinfoTable1.xlsx
Supplementary Table 1
ExtendedData.docx
Extended Data Tables 1-3
ExtendedDataFig1.png
Extended Data Figure 1
ExtendedDataFig2.pdf
Extended Data Figure 2
ExtendedDataFig3.pdf
Extended Data Figure 3
ExtendedDataFig4.pdf
Extended Data Figure 4
SupplementaryinfovideoEh.wmv
Supplementary Video 1
SupplementaryinfovideositeB.mp4
Supplementary Video 2

Download PDF

Version 1

posted

You are reading this latest preprint version

Large CO₂ seeps and hydrates field in the Indian Ocean (Mayotte Island)

Status:

Version 1

Abstract

Figures

Introduction

The Horseshoe edifice: a large liquid CO2 venting and hydrate field area

Origin and ecological impact of the magmatic volatiles

CO2 budget and natural C sequestration laboratory

Methods

Declarations

Competing interests

Author contributions

Acknowledgements

References

Additional Declarations

Supplementary Files

Status:

Version 1