Search

SUMMARYThe brutal onset of seismicity offshore Mayotte island North of the Mozambique Channel, Indian Ocean, that occurred in May 2018 caught the population, authorities and scientific community off guard. Around 20 potentially felt earthquakes were recorded in the first 5 d, up to magnitude Mw 5.9. The scientific community had little pre-existing knowledge of the seismic activity in the region due to poor seismic network coverage. During 2018 and 2019, the MAYOBS/REVOSIMA seismology group was progressively built between four French research institutions to improve instrumentation and data sets to monitor what we know now as an on-going exceptional submarine basaltic eruption. After the addition of 3 medium-band stations on Mayotte island and 1 on Grande Glorieuse island in early 2019, the data recovered from the Ocean Bottom Seismometers were regularly processed by the group to improve the location of the earthquakes detected daily by the land network. We first built a new local 1-D velocity model and established specific data processing procedures. The local 1.66 low VP/VS ratio we estimated is compatible with a volcanic island context. We manually picked about 125 000 P and S phases on land and sea bottom stations to locate more than 5000 events between February 2019 and May 2020. The earthquakes outline two separate seismic clusters offshore that we named Proximal and Distal. The Proximal cluster, located 10 km offshore Mayotte eastern coastlines, is 20–50 km deep and has a cylindrical shape. The Distal cluster start 5 km to the east of the Proximal cluster and extends below Mayotte's new volcanic edifice, from 50 to 25 km depth. The two clusters appear seismically separated, however our data set is insufficient to firmly demonstrate this.

Submarine channels are the main conduits and intermediate stores for sediment transport into the deep sea, including organics, pollutants, and microplastics. Key drivers of morphological change in channels are upstream-migrating knickpoints whose initiation has typically been linked to episodic processes such as avulsion, bend cutoff, and tectonics. The initiation of knickpoints in submarine channels has never been described, and questions remain about their evolution. Sedimentary and flow processes enabling the maintenance of such features in non-lithified substrates are also poorly documented. Repeated high-resolution multibeam bathymetry between 2012 and 2018 in the Capbreton submarine canyon (southeastern Bay of Biscay, offshore France) demonstrates that knickpoints can initiate autogenically at meander bends over annual to multi-annual time scales. Partial channel clogging at tight bends is shown to predate the development of new knickpoints. We describe this initiation process and show a detailed morphological evolution of knickpoints over time. The gradients of knickpoint headwalls are sustained and can grow over time as they migrate through headward erosion. This morphology, associated plunge pools, and/or development of enhanced downstream erosion are linked herein to the formation and maintenance of hydraulic jumps. These insights of autogenically driven, temporally high-frequency knickpoints reveal that cut-and-fill cycles with depths of multiple meters can be the norm in submarine systems.

International audience; Volcanic eruptions shape Earth’s surface and provide a window into deep Earth processes. How the primary asthenosphericmelts form, pond and ascend through the lithosphere is, however, still poorly understood. Since 10 May 2018, magmaticactivity has occurred offshore eastern Mayotte (North Mozambique channel), associated with large surface displacements,very-low-frequency earthquakes and exceptionally deep earthquake swarms. Here we present geophysical and marine datafrom the MAYOBS1 cruise, which reveal that by May 2019, this activity formed an 820-m-tall, ~5 km³ volcanic edifice on the seafloor.This is the largest active submarine eruption ever documented. Seismic and deformation data indicate that deep (>55 kmdepth) magma reservoirs were rapidly drained through dykes that intruded the entire lithosphere and that pre-existing subverticalfaults in the mantle were reactivated beneath an ancient caldera structure. We locate the new volcanic edifice at the tip ofa 50-km-long ridge composed of many other recent edifices and lava flows. This volcanic ridge is an extensional feature insidea wide transtensional boundary that transfers strain between the East African and Madagascar rifts. We propose that the massiveeruption originated from hot asthenosphere at the base of a thick, old, damaged lithosphere.

Seabed telecommunication cables can be damaged or broken by powerful seafloor flows of sediment (called turbidity currents), which may runout for hundreds of kilometres into the deep ocean. These flows have the potential to affect multiple cables near-simultaneously over very large areas, so it is more challenging to reroute traffic or repair the cables. However, cable-breaking turbidity currents that runout into the deep ocean were poorly understood, and thus hard to predict, as there were no detailed measurements from these flows in action. Here we present the first detailed measurements from such cable-breaking flows, using moored-sensors along the Congo Submarine Canyon offshore West Africa. These turbidity currents include the furthest travelled sediment flow (of any type) yet measured in action on Earth. The SAT-3 (South Atlantic 3) and WACs (West Africa Cable System) cables were broken on 14-16th January 2020 by a turbidity current that accelerated from 5 to 8 m/s, as it travelled for > 1,130 km from river estuary to deep-sea, although a branch of the WACs cable located closer to shore survived. The SAT-3 cable was broken again on 9th March 2020 due to a second turbidity current, this time slowing data transfer during regional coronavirus (COVID2019) lockdown. These cables had not experienced faults due to natural causes in the previous 19 years. The two cable-breaking flows are associated with a major flood along the Congo River, which produced the 3 highest discharge (72,000m3 ) recorded at Kinshasa since the early 1960s, and this flood peak reached the river mouth on ~30th December 2019. However, the cable-breaking turbidity currents occurred 2-10 weeks after the flood peak and coincided with unusually large spring tides. Thus, the large cable-breaking flows in 2020 are caused by a combination of a major river flood and tides; and this can provide a basis for predicting the likelihood of future cable-breaking flows. Older (1883-1937) cable breaks in the Congo Submarine Canyon occurred in temporal clusters, sometimes after one or more years of high river discharge. Increased hazards to cables may therefore persist for several years after one or more river floods, which cumulatively prime the river mouth for cable-breaking flows. The 14-16th January 2020 flow accelerated from 5 to 8 m/s with distance, such that the closest cable to shore did not break, whilst two cables further from shore were broken. The largest turbidity currents may increase in power with distance from shore, and are more likely to overspill from their channel in distal sites. Thus, for the largest and most infrequent turbidity currents, locations further from shore can face lower-frequency but higher-magnitude hazards, which may need to be factored into cable route planning. Observations off Taiwan in 2006-2015, and the 2020 events in the Congo Submarine Canyon, show that although multiple cables were broken by fast (> 5 m/s) turbidity currents, some intervening cables survived. This indicates that local factors can determine whether a cable breaks or not. Repeat seabed surveys of the canyon-channel floor show that erosion during turbidity currents is patchy and concentrated around steeper areas (knickpoints) in the canyon profile, which may explain why only some cables break. If possible, cables should be routed away from knickpoints, also avoiding locations just upcanyon from knickpoints, as knickpoints move up-slope. This study provides key new insights into long runout cable-breaking turbidity currents, and the hazards they pose to seafloor telecommunication cables.

In October 2020, during the marine expedition FocusX1 onboard the research vessel PourquoiPas? microbathymetric mapping was performed using the ROV Victor6000. The main goal was to map the seafloor expression of the North Alfeo fault and select the best path for deployment of a 6-km long fiber optic strain cable designed to monitor movement along the fault and the deployment sites for 8 geodetic stations.Bathymetric data were collected through a Reson Seabat 7125 multibeam echosounder (400 kHz). ROV navigation data were processed using DelphINS, resulting in an optimal merging of navigation sensors (GPS, USBL, DVL, pressure). The MBES data processing (GLOBE software) mainly consisted in estimating and correcting static angular offsets, applying actual in-situ sound speed profile, and finally performing an automatical and manual soundings filtering.The resulting bathymetric grid spans a region of roughly 3 km x 1.5 km, with a 1m cell size, and allows us to identify a variety of morphological features:1 - a set of narrow, linear, E-W oriented gulleys, all parallel (not merging/branching) on a regional E dipping 5-15° slope2 - a striking, continuous curvi-linear feature, which is interpreted as the primary surface expression of the fault.The fault morphology changes from a smooth less than 10 m depression in the NW to a up to 10-20m high scarp with slopes of 20-30°, and locally sub-vertical cliff faces.3 - a local bathymetric plateau (mesa like feature) with a gently E-dipping summit region, showing signs of eastward sliding / rafting tectonics, indicated by N-S oriented gashes/depressions.The 3-km long segment of the fault covered by our survey includes the mesa-like bathymetric high (at the NW extremity) interpreted as a transpressional pop-up feature and an elongated, fault bounded trough (at the SE extremity) interpreted as a transtensional pull-apart basin. Video-camera images recorded by ROV Victor6000 from the seafloor provide visual documentation of the fault scarp and seafloor morphology. Future surveys with a sub-bottom profiler and/or HR- seismics can help confirm these interpretations. The ongoing monitoring with the fiber-optic strain cable is being calibrated by a 3-4 year deployment of seafloor geodetic instruments (Canopus acoustic beacons manufactured by iXblue) which started in Oct. 2020, and will allow us to quantify relative displacement across the fault.

For the first time, a 6-km long fiber-optic strain cable was deployed across an active fault on the seafloor with the aim to monitor possible tectonic movement using laser reflectometry, 25 km offshore Catania Sicily (an urban area of 1 million people). Brillouin Optical Time Domain Reflectometry (BOTDR) is commonly used for structural health monitoring (bridges, dams, etc.) and under ideal conditions, can measure small strains (10-6) along a fiber-optic cable, across very large distances (10 - 200 km), with a spatial resolution of 10 - 50 m. The FocusX1 expedition, (6-21 October 2020) onboard the R/V Pourquoi Pas? was the first experiment of the European funded FOCUS project (ERC Advanced Grant). We first performed micro-bathymetric mapping and a video camera survey using the ROV Victor6000 to select the best path for the cable track and for deployment sites for eight seafloor geodetic stations. Next we connected a custom designed 6-km long fiber-optic cable (manufactured by Nexans Norway) to the TSS (Test Site South) seafloor observatory in 2100 m water depth operated by INFN-LNS (Italian National Physics Institute) via a new Y-junction frame and cable-end module. Cable deployment was performed by means of a deep-water cable-laying system with an integrated plow (updated Deep Sea Net design Ifremer, Toulon) to bury the cable 20 cm in the soft sediments in order to increase coupling between the cable and the seafloor. The cable track crosses the North Alfeo Fault at four locations. Laser reflectometry measurements began on 18 October 2020 and are being calibrated by a 3 - 4 year deployment of eight seafloor geodetic instruments (Canopus acoustic beacons manufactured by iXblue) deployed on 15 October 2020. During a future marine expedition, tentatively scheduled for early 2022 (FocusX2) a passive seismological experiment is planned to record regional seismicity. This will involve deployment of a temporary network of Ocean Bottom Seismometers (OBS) on the seafloor and seismic stations on land, supplemented by INGV permanent land stations. The simultaneous use of laser reflectometry, seafloor geodetic stations as well as seismological land and sea stations will provide an integrated system for monitoring a wide range of slipping event types along the North Alfeo Fault (e.g. - creep, slow-slip, rupture). A long-term goal of the project is the development of dual-use telecom cables with industry partners.

Volcanic eruptions are foundational events shaping the Earth’s surface and providing a window into deep Earth processes. We document here an ongoing magmatic event offshore Mayotte island (Western Indian Ocean) unprecedented in terms of emitted volume of lava and duration of the seismic crisis.This event gave birth to a deep-sea volcanic edifice 820m tall and ~ 5 km3 in volume, located 50 km from Mayotte. A plume with distinct chemical signatures compared to open-ocean seawater emanated from the edifice, generating an exceptional 1900m-high vertical acoustic anomaly in the water column. Noble gas analyses in the vesicles from a popping rock dredged on the flank of the edifice, indicate rapid magma transfer from the asthenosphere. The edifice is located at the tip of a WNW-ESE–striking volcanic ridge composed of many other edifices, cones and lava flows constructed by past eruptions. Starting in May 2018 thousand of earthquakes were triggered by the magmatic event. The space-time distribution of the seismicity suggests that magma below the center of the ridge was transported to the new edifice over a few weeks in dikes that penetrated the brittle mantle a result of a lithosphere-scale extensional episode accommodating motion along a transfer zone between the East-African rifts and Madagascar. Since the eruption’s onset, the seismicity is mostly concentrated closer to the island, in an exceptionally deep zone (25-50 km) overlain by a zone of enigmatic, very low frequency, tremors.

Turbidity currents form many of the largest sediment accumulations, longest channels, and deepest canyons on our planet. These seabed sediment avalanches can be very (> 10 m/s) fast, runout for hundreds of kilometres, and break seabed cables that now form the backbone of the internet and global data transfer. It was once thought that detailed monitoring of turbidity currents in action was impractical, ensuring these flows were relatively poorly understood. However, a series of recent projects have used new approaches and technology to show how these flows can be measured in shallow water (< 2 km) settings, such as Monterey Canyon and Canadian fjords, where flows ran out for < ~50 km and had speeds of up to 8 m/s. Here we present initial results from an ambitious project to measure active flows that runout for >1,000 km to form a major submarine fan in the deep ocean. The project studies the Congo submarine canyon-channel system that extends for ~1,100 km from the mouth of the Congo River, offshore West Africa. Monitoring in 2010 at a single site in the upper Congo Canyon had previously shown that flows are active for ~30% of the time, and reach speeds of up to 3 m/s. In this new project, direct flow monitoring at 11 sites are being combined with detailed time-lapse mapping and coring of flow deposits, through a series of 4 or 5 major research cruises from 2019 to 2023. Here we present initial results from the first of these cruises (JC187) in August-to-October 2019, which placed 11 moorings with sensors at water depths of 1.6 to 5.5 km. The presentation will initially focus on the geomorphology of the channel system, and how it varies down-slope and through time. For example, it is apparent that a landslide partly blocked one location in the upper canyon in the last 20 years, causing meander bend cut-off and sediment ponding. The talk will then discuss models for how submarine channel bends evolve, and the implications for channel deposits. Recent work in sandy submarine channels suggests that they can be dominated by very fast-moving knickpoints (waterfall like features). However, the much muddier Congo channel displays well-developed meander bend bars for which cores are available. We therefore start to show how muddy deep-sea channels may differ in significant ways from their sandier cousins in shallow water.

The early development and growth of seamounts are poorly known as the birth of a volcano on the sea bottom has been rarely observed. The on-going Mayotte seismo-volcanic crisis is associated with the formation of a new seafloor volcano at a water depth of 3300 m and provides the opportunity to study its early development.Four oceanographic cruises, MAYOBS 1 to 4, were carried out between May and July 2019 aboard the French R/V Marion Dufresne. High resolution bathymetry and backscatter data as well as sub-bottom profiler, gravity and magnetic profiles were collected during each cruise. A dense network of profiles has been achieved over the new volcano at different epochs, allowing to assess its detailed morphology and the evolution through time. During MAYOBS4, a deep-towed underwater camera provided sea bottom videos and photos on the volcano.First results indicate that the new volcano is still growing at the end of July 2019. Repetitive surveys in May, June and July 2019 allow to document the morphological evolution of the volcano, to estimate the volume of material emplaced between each epoch and to discuss the emitted lava rate.The new volcano has a starfish shape and is now 820 m high. Steep slopes are observed close to the summit and several radial ridges developed from its central part, displaying hummocky morphology similar to the ones observed along mid oceanic axial volcanic ridges. At the bottom, flat areas with high backscatter could indicate channelized lava flows emplaced at higher effusion rates. The morphological analysis combined with video imagery brings constraints to the eruptive processes yielding to the formation of a nascent volcano.

International audience; The Lansdowne Bank is a partly drowned, isolated carbonate platform of around 4000 km2 located 300 km west of New Caledonia, in the SW Pacific Ocean, in water depths of 20 to 100 m. New multibeam bathymetric data, high resolution seismic reflection profiles and sediment gravity cores have been acquired on the bank top and adjacent slopes. This dataset reveals an almost continuous 4 km wide outer reef rim located in ca. 50 m water depth, surrounding a gently deepening inner platform, reaching up to 100 m water depth. The bank is bordered by very steep slopes showing numerous erosional morphologies such as canyons, channels and gullies. Along with these bypass features, spectacular bank margin collapses and slope failures are evidenced by up to 20 km-wide bank edge and intraslope failure scars, respectively, resulting in a typical “scalloped” geometry of the bank margin. These failure scars can lead to a complete collapse of the outer reef rim and impact subsequent reef development. Bank margin collapses are evidenced by hectometer to kilometer-scale blocks and debris shed on the slope, likely emplaced by rock fall/avalanching processes originating from the brittle failure of early cemented bank edge and upper slope sediments. In turn, failures triggered on the un-cemented mud-prone middle to lower slopes likely generate more cohesive, submarine debris flows that could be at the origin of erosive morphologies within the debris fields. Estimated individual failure volumes can reach up to 3 km3. Quaternary sea-level lowstands, that would have led to platform exposure, fracturing and karstification, and the development of an erosional sea cliff, as well as subsequent rising sea-level are believed to play a significant role in mass wasting event emplacement, yet “bottom up” submarine processes such as the upslope propagation of bypass morphologies by retrogressive headward erosion cannot be ruled out. In terms of geomorphic and stratigraphic constraints, the documented bank margin collapses affect a terrace located in 70 m water depth around the bank, which, depending on its age and origin, could provide a minimum age for collapse events. Finally, considering the shallow water depth of failure headscarps, the volumes of material involved in the slides as well as their vicinity to the nearby main island of New Caledonia, numerical simulations of the tsunamigenic potential of submarine slides have been performed. They showed that these slides would have been able to produce a meter-scale wave that would reach the northern coast of the island in less than an hour.

International audience; Mixed turbidite-contourite systems can be found in oceans where bottom currents and turbidity currents interact. The Zambezi turbidite system, located in the Mozambique Channel (SW Indian Ocean), is one of the largest sedimentary systems in the world in length and area of the related catchments. The oceanic circulation in the Mozambique Channel is intense and complex, dominated by eddies flowing southwards and deep currents flowing northwards along the Mozambican margin. Current measurements obtained from moorings at 3400–4050 m water depth in the Zambezi and Tsiribihina valleys show periods of intense currents at the seafloor with peaks of 40–50 cm s−1 that last up to one month and are not related to turbidity currents. These strong bottom-current events are correlated with a change in current direction and an increase in temperature. The periods of current intensification may be related to eddies, since they present similar frequencies (around 7 per year). Moreover, modelling results show that during periods of intense deep circulation an anticyclonic eddy is present between the Mozambican slope and the centre of the Mozambique Channel, which may block the northward transport of the deep water mass and thus enhance the southward transport along the western slope of Madagascar. According to our hydrodynamic modelling of the circulation near the seafloor, intense currents are often present along the Zambezi Valley, especially along the valley flanks. Multi-channel seismic reflection data show that the Zambezi turbidite system does not show the typical characteristics of turbidite systems, being dominated by erosional processes, which mainly affect the valley flanks. Levees associated with the valley are absent in the main axis of the system. The effect of bottom currents on sedimentation in the basin is evidenced by the low sedimentation rates that witness winnowing in the basin, the presence of contouritic sand in the Zambezi Valley flanks and the abundance of current-related bedforms observed in multibeam bathymetry and seismic data. The intense oceanic processes observed in the Mozambique Channel may transport a large part of the fine sediment out of the basin and erode the seafloor even at great depths. Therefore, the Zambezi turbidite system could at present be considered as a mixed turbidite-contourite system, with important implications for source-to-sink studies.

Cold-water coral ecosystems have been identified as vulnerable, but quantitative data on their conservation status is very limited. The Marine Strategy Framework Directive (MSFD) is the tool implemented by the European Union’s Integrated Maritime Policy to achieve Good Environmental Status (GES) of marine waters by 2020. In this context, the aim of this study was to evaluate the Ecological Status of benthic habitats in Cassidaigne canyon, focusing in particular on cold-water coral habitats dominated by Madrepora oculata. Data were collected during the Videocor1 cruise (2017). Videos and photos collected during eight dives of the H-ROV Ariane were used to reconstruct, in 3-dimensions, the areas where cnidarians have settled in the canyon. A total of 33 3D models were built, which allowed measuring the spatial and vertical distribution, surface, density and size structure of cnidarian populations at four different sites. When 3D reconstructions were not possible, GIS tools were used. The seven cnidarian species considered were the scleractinian M. oculata; three antipatharians: Leiopathes glaberrima, Antipathella subpinnata, Antipathes dichotoma; and three aclyonaceans: the precious red coral Corallium rubrum and the gorgonians Callogorgia verticllata and Viminella flagellum. Using photogrammetry, we were able to reveal the size structure of the dense population of M. oculata in the canyon, as well as to obtain knowledge on a complex site (Cassis-200) composed of 15 knolls, and to quantify the surface occupied by M. oculata at a separate site (Cassis-500) influenced by industrial discharges. At the southern flank of the canyon we found a highly diverse site (SW Flank) dominated by antipatharians and gorgonians composing large forests, and finally a reservoir of M. oculata was identified under overhangs at a site called the Wall. The diversity of accompanying species is also reported and marine litter quantified. Images collected before 2017 were compared to the 3D models to precisely locate them on the sites, and assess temporal changes in M. oculata colony sizes at Cassis-200 site. We also report on the ground-truthing of predicted habitat maps produced previously, and confirm their good representation of the distribution of cold-water coral habitats. Finally, we quantified the criteria defined by the MSFD, aimed at evaluating the GES of benthic habitats for M. oculata ecosystems, at the scale of the Cassidaigne canyon. Measurements showed that the extent of loss of the observed M. oculata habitat reached 56% according to the MFSD definition.

The study area is close to the Wallis and Futuna Islands in the French EEZ. It exists on the western boundary of the fastest tectonic area in the world at the junction of the Lau and North-Fiji basins. At this place, the unstable back-arc accommodates the plate motion in three ways: (i) the north Fiji transform fault, (ii) numerous unstable spreading ridges, and (iii) large areas of recent volcanic activity. This instability creates bountiful opportunity for hydrothermal discharge to occur. Based on geochemical (CH 4 , TDM, 3 He) and geophysical (nephelometry) tracer surveys: (1) no hydrothermal activity could be found on the Futuna Spreading Centre (FSC) which sets the western limit of hydrothermal activity; (2) four distinct hydrothermal active areas were identified: Kulo Lasi Caldera, Amanaki Volcano, Fatu Kapa and Tasi Tulo areas; (3) extensive and diverse hydrothermal manifestations were observed and especially a 2D distribution of the sources. At Kulo Lasi, our data and especially tracer ratios (CH 4 / 3 He ~50×10 6 and CH 4 /TDM ~4.5) reveal a transient CH 4 input, with elevated levels of CH 4 measured in 2010, that had vanished in 2011, most likely caused by an eruptive magmatic event. By contrast at Amanaki, vertical tracer profiles and tracer ratios point to typical seawater/basalt interactions. Fatu Kapa is characterised by a substantial spatial variability of the hydrothermal water column anomalies, most likely due to widespread focused and diffuse hydrothermal discharge in the area. In the Tasi Tulo zone, the hydrothermal signal is characterised by a total lack of turbidity, although other tracer anomalies are in the same range as in nearby Fatu Kapa. The background data set revealed the presence of a Mn and 3 He chronic plume due to the extensive and cumulative venting over the entire area. To that respect, we believe that the joined domain composed of our active area and the nearby active area discovered in the East by Lupton et al. (2012) highly contribute to the extensive Tonga-Fiji plume and which thus may not originate from a sole source near the Samoa. Our results also emphasize and support the idea that back-arc hydrothermal systems have a significant input to the regional and global ocean and maybe more important than their MOR analogues.

Seismic velocities together with sediment/rock compaction models are of common use in the oil industry to predict reservoir pressures and to detect and identify any potential hazard associated with over-pressured formations. Excess pore pressure is a key parameter for geohazard assessment, nevertheless this approach is very seldom applied to shallow offshore engineering as it requires expensive ground-truth boreholes. The simultaneous availability in the Bourcart-Herault canyon interfluve (Gulf of Lion – NW Mediterranean) of 1) geotechnical and sedimentological data (300 m boreholes) from the PROMESS project; 2) in situ pore pressure measurements; and 3) High Resolution (HR) seismic geophysical data showing the presence of active structures in relation to fluid circulation, allows the importance of a multi-disciplinary approach to shallow geohazard assessment to be illustrated through a case study. The velocity field obtained from HR seismic data is interpreted in light of the neighbouring boreholes and resulting velocity anomalies are translated in terms of fluid overpressure and free gas content. Additional in situ surface geotechnical measurements (piezometers) are then taken into account and the role of these parameters is discussed.

Thick forced regressive units on the wide continental shelf of the Gulf of Lions (western Mediterranean) recorded the composite effect of sea level changes during the Quaternary. They are mostly composed of coastal siliciclastic and bioclastic wedges showing clinoform geometry. These deposits have been intensively explored through high-resolution seismic investigations, but only recently it was possible to ground truth seismic interpretations, based on a long (100 m) borehole that crossed the succession and recovered a large part of the mainly sandy deposits (similar to 84% recovery). A multiproxy analysis of the sedimentary succession shows that (1) the stratal architecture of the shelf margin is defined by major bounding surfaces that are polygenic erosion surfaces associated with coarse-grained material incorporating abundant and diverse shells, including cold-water fauna (presently absent from the Mediterranean Sea). Between each surface, coarsening upward units with steep (up to 5 degrees) foresets are made of massive (more than 20 m thick) sands with possible swaley and hummocky cross-stratification, passing seaward to sands with muddy intervals and, further offshore, alternating highly boiturbated sands and silts. Each prograding wedge corresponds to a forced-regressive shoreface (or delta front/prodelta), deposited during the overall sea level falls occurring at (relatively slow) interglacial/glacial transition and therefore represents the record of 100 ka cyclicity. Higher-frequency Milankovitch cyclicities are also probably represented by distinct shoreface/delta front wedges; (2) detailed examination of the architecture and chronostratigraphy of the most recent sequence shows that minor bounding surfaces, corresponding to abrupt shallowing of sedimentary facies, separate downward stepping parasequences within the last 100 ka sequence...