In several fields of research, the Bay of Concarneau, through its geological and physiographic characteristics, constitutes a privileged site for the study of Holocene stratigraphic series that are still well preserved and continuous. Its current morphology only shows on the surface the faint trace of the ancient fluvial network of which the work of Menier et al. (2010) showed all the complexity. The first studies (e.g. Baltzer et al., 2014) carried out in the Concarneau region show that this environment has all the characteristics to better understand the processes of formation of pockmarks (craters on the seabed linked to the release of fluid). The geographical proximity of this site and the shallow depth of the seabed also offer the opportunity to propose an approach integrating time series of multidisciplinary and multi-scale measurements to better understand in situ the gas formation processes in the sediment, estimate its kinetics, identify the geochemical processes in which they are involved during their migration in the sedimentary column until their fate in the water column. The expected results would indeed make it possible to understand the dynamics of degassing and deformation of the seabed. The experimental nature of certain planned acquisitions, possible due to the originality of this system, will therefore allow us to better understand the geological processes that we study on other sites such as the Bay of Biscay and the Sea of Marmara.

To better understand the ¿fluid machine¿ of the Bay of Concarneau, the results of the SYPOCO mission concern several aspects of the geological compartment (surface and deep) and water column:

(1) Mapping the roof of the bedrock, the network of incised valleys and the thickness of the filling highlighted a dichotomy between the region to the north and south of the tip of Trévignon (Belleney, 2018). The morphology of the network of incised valleys obtained from the mapping of the bedrock is upstream (North zone), sinuous and branched in shape, which converges into a single drain in the downstream part (South zone). The reactivation of the set of faults (ESA-N120, faults-N60 and Kerforne-N150), on the periphery of the uplift of the central part of the Armoricain Massif, could explain this difference in behavior of the North and South blocks of the bay, the compartmentalization of the nature of the geological base juxtaposing with the shape of the submerged hydrographic network (figure 1). Structural heritage appears to play a major role in the partitioning of the area. Indeed, the zone of collusion of drains and maximum filling corresponds to the presence of a mica schist base which makes the transition between two compartments (North and South) of folded and faulted limestone. The difference in North/South filling (thin/thick) found around the isobath - 36 meters, would indicate at this limit a stationing of sea level at the start of the Holocene marine transgression. A change in the sedimentation regime within the paleo-valley and in the Bay of Concarneau should be considered with the evolution of the speed of sea level rise around 10,000 years. In addition, the opening of the Basse Jaune and Moutons pass (respectively at - 30 m and - 15 m) in connection with the rise in sea level, leading to a change in the hydrodynamic context (swell/tide) could explain the changes in more recent sedimentary filling such as the creation of prograding sandy sedimentary prisms in the northwest of the area.

(2) The analysis of sediment cores and very high resolution seismic data made it possible to clarify the nature of the Holocene sedimentation in the Bay of Concarneau, partly filled by a network of Pleistocene paleo-channels (Sommer-Delfolie, 2019) . Most of the sedimentary sequence is made up of silty mud sometimes rich in turrets. Among these estuarine deposits, a coarse, heavily shelly sedimentary horizon, from a few decimeters to more than a meter thick, is intercalated at the base of the last superficial sedimentary unit. Its grain size and poor sorting characterize sediments reworked by storms and probably record a succession of climatic deteriorations between 5500 and 500 cal years BP on the Breton coasts. The last centuries of sedimentation are represented by a silty unit, with a low sedimentation rate, and characterized on the surface to the right of the Kerforne paleovalley, by a large field of pockmarks. This shell bed, coupled with variations in thickness of the most recent sedimentary unit, plays a major role in the distribution and density of pockmarks on the surface (figure 2).

(3) the geotechnical and sedimentological characteristics of the pockmark field of Concarneau Bay were studied using lithology, grain size, MSCL data, obtained using oedometers and piezometers (Sombstay, 2021). Each site studied revealed different specific characteristics depending on the depth thanks to seismic analysis. No ¿chimney¿ for expulsion of fluids that could witness the migration of gas from the sediments to the water column is identified in the seismic data. The results found no obvious links between the fault network in the Eocene basin and the distribution of pockmarks (no alignment). Two hypotheses on the gas concentration level are possible: Either the gas is concentrated at the roof of unit U3 (estuarine muds) in the interface which corresponds to a coarse discordant deposit (reflector of very high amplitude corresponding to a relatively permeable and not very compressible see Sommer-Delfolie, 2019), or it is linked to the most recent unit U4 Holocene. All the data acquired showed that unit U4 is composed of offshore mud (fine compressible sediments), in which more sandy or shelly levels are interspersed, capable of storing gas pockets. The heterogeneity of this layer contributes to both the retention and release of gas at the surface. As suggested by in-situ piezometric data, surface fluid release is controlled by tidal pore pressure changes. Unit U4 which drapes the rest of the area, devoid of pockmarks, is of a different and more sandy nature.

(4) Several cores taken by Kullenberg corer and by divers were the subject of a more precise investigation to determine the presence of methane (Hubert et al, 2018). The Kullenberg SYP-KS11 and SYP-KS21 cores were recovered by zodiac immediately after collection. The first step consisted of opening windows measuring 10 cm by 8 cm in order to take samples for methane, porosity, and 4 syringes for measuring sulfate reduction rates in the case of SYP-KS21 only. Secondly, rhizons were placed approximately every 10 cm to extract the pore water. It was also decided to sample a pockmark already investigated by Dubois et al (2015) in order to study a potential temporal evolution (figure 3). The alkalinity of the pore water is measured by titration quickly after sampling in order to avoid any potential change in the solution. For the determination of sulphides, the samples were analyzed on site in continuity with the sampling, thus ensuring more reliable results. On the other hand, the analysis of methane is carried out by gas chromatography (Perichrom® Pr2100) equipped with a FID detector (Flame Ionization Detector), coupled with a static ¿headspace¿ injector (DANI® HSS 86.50) (Sarradin and Caprais , 1996). The profiles of the dip cores and SYP-KS21 are characteristic of a classic early diagenesis sequence (Libouban, 2019). The reduction of sulfates is very marked along the sedimentary column. The methane present at depth is oxidized in the SMTZ at 0.2 m. The increase in the CH4 concentration below the SMTZ demonstrates the degradation of organic matter by bacteria via the process of methanogenesis described in Schulz and Zabel (2013). Sulfur speciation, iron speciation as well as carbon, nitrogen and sulfur contents were carried out (Sanlis, 2021). The PCA shows that the organic carbon content is greater in the muddy sediments and that, perhaps, only a small quantity of methane would rise from the pockmarks and which could explain the few differences observed between the zones.

Figure 1 : Geo-structural dichotomy of the Bay of Concarneau (Belleney, 2018)

Figure 2: Interpretation of the fluid migration paths which would be at the origin of the field of pockmarks observed on the seabed of the Bay of Concarneau (Sommer-Delfolie, 2019)

Figure 3: Location of one of the geochemical measurement sites from SYPOCO SMF data (Hubert et al, 2018)