Description: An exceptionally large cold-water coral mound province (CMP) was recently discovered extending over 80 km along the Namibian shelf (offshore southwestern Africa) in water depths of 160-270 m. This hitherto unknown CMP comprises 〉2000 mounds with heights of up to 20 m and constitutes the largest CMP known from the southeastern Atlantic Ocean. Uranium-series dating revealed a short but intense pulse in mound formation during the early to mid-Holocene. Coral proliferation during this period was potentially supported by slightly enhanced dissolved oxygen concentrations compared to the present Benguela oxygen minimum zone (OMZ). The subsequent mid-Holocene strengthening of the Benguela Upwelling System and a simultaneous northward migration of the Angola-Benguela Front resulted in an intensification of the OMZ that caused the sudden local extinction of the Namibian corals and prevented their reoccurrence until today.

Description: Highlights • First granulometric record of Holocene Atlantic water inflow into Labrador Sea. • Good temporal correlation with North Atlantic current speeds and AMOC. • Highest current speeds in the early Holocene, lowest during the Neoglaciation. • Distinction between Atlantic water inflow and influence crucial for paleo-studies. • Local ice sheet advances in SW Greenland during “8.2”-event and Neoglaciation. Abstract The hydrodynamics of the Labrador Sea, controlled by the complex interplay of oceanographic, atmospheric and ice-sheet processes, play a crucial role for the Atlantic Meridional Overturning Circulation (AMOC). An improved understanding of the hydrodynamics and its forcing in the past could therefore hold a key to understanding its future behaviour. At present, there is a remarkable temporal mismatch, in that the largely microfossil-based reconstructions of Holocene Atlantic-water inflow/influence in the Labrador Sea and Baffin Bay appear to lag grain size-based current strength reconstructions from the adjacent North Atlantic by 〉 2ka. Here, we present the first current strength record from the West Greenland shelf off Nuuk to reconstruct Atlantic Water (AW)-inflow to the Labrador Sea via the West Greenland Current. Our data show that the Holocene AW-inflow into Labrador Sea is well aligned with the Holocene Speed Maximum documented in the North Atlantic (McCave and Andrews, 2019; Quat. Sci. Rev. 223), suggesting a close coupling with the AMOC. The observed lag between the microfossil-based records and the Holocene Speed Maximum can be explained when considering the presence of an extended meltwater lens that prevented the shoaling of the inflowing Atlantic waters. Once the meltwater discharge waned after the cessation of large-scale melting of the surrounding ice sheets, the AW could influence the surface waters, independently of the strength of its inflow. Only then was an effective ocean-atmosphere heat transfer enabled, triggering the comparably late onset of the regional Holocene Thermal Maximum. Furthermore, sediment geochemical analyses show that short term cooling events, such as the 8.2 ka event related to the final drainage of glacial Lake Agassiz, lead to glacier advances of the Greenland Ice Sheet. Since the grain size data show that these events had no influence on the AW-inflow to the north eastern Labrador Sea, these advances must have been caused by atmospheric cooling. Consequently, we argue that (i) in this region, surface water-based proxies register AW influence rather than inflow (ii) the AW inflow into the Labrador Sea is controlled by the AMOC, but (iii) its impact on an effective ocean-atmosphere heat transfer was hindered by a prevailing meltwater lens in the early Holocene, i.e. until the cessation of large-scale melting of the surrounding ice sheets.