Eastern Mediterranean Deep Water Formation During Sapropel S1: A Reconstruction Using Geochemical Records Along a Bathymetric Transect in the Adriatic Outflow Region

Abstract

Eastern Mediterranean thermohaline circulation is directly influenced by middle‐ and low‐latitude climate systems. The dramatic paleoclimate changes during the last African Humid Period (~10–6 ka BP) were captured in Mediterranean sediments as the distinctly organic‐rich unit sapropel S1. Here, deepwater formation variability during S1 deposition is reconstructed. We use geochemical records of three cores along a bathymetric transect (775‐, 1,359‐, and 1,908‐m water depths), at the transition between the Adriatic DW‐ formation area and the Eastern Mediterranean. In all three cores, sedimentation rates are distinctly higher during S1, corresponding with enhanced runoff emanating from the Adriatic hinterland. Hence, major runoff did not only come from southern but also from northern borderlands in this period. During sapropel formation, enhanced levels of primary productivity occurred in the surface waters and oxygen‐depleted conditions in the bottom waters for all sites. Conditions for sediment and bottom‐water below ~1.4 km water depth were sulfidic throughout S1, but for intermediate depth (775 m) were anoxic only during the first part (S1a). Bottom‐water oxygenation interrupted S1 formation at water depths down to ~1.4 km, during two brief episodes, at 8.2 and 7.4 cal. ka BP. From the 7.4 cal. ka BP ventilation onward, the transition to more oxygenated bottom‐water conditions was more progressive for the intermediate water depth site (775 m) than for the deeper sites. Conditions remained fully oxic for all water depths following the S1‐MarkerBed ventilation event. Possibly, the onset of continuously oxic conditions started slightly earlier at intermediate depth (775 m; 6.6 ± 0.3 cal. ka BP) than at greater depths (1,359 m, 1,908 m; 6.0 ± 0.3 cal. ka BP).

The Postglacial Isotopic Record of Intermediate Water Connects Mediterranean Sapropels and Organic‐Rich Layers

“…The final alkenone accumulation in the ODP Site 976 occurred between7.3 and 6.8 ka and matches with a prominent peak in the Aegean Sea oxyphilic benthic foraminifera ( Figure 6k) (Abu-Zied et al, 2008;Casford et al, 2003), with geochemical and micropaleontological evidence of deep water formation in the Adriatic Sea (Di Donato et al, 2019;Filippidi & De Lange, 2019;Maiorano et al, 2019) and again with increasing LIW density and improved seafloor ventilation in the Sicily Channel (Figures 6e, 6f, and 6i). The 8.2 and 7.2 ka events, which led to benthic foraminifera repopulation, are compatible with improved deep water oxygenation in the Adriatic and Aegean Sea (Filippidi et al, 2016;Filippidi & De Lange, 2019). Changes in the Sicily Channel seafloor and in the eastern Mediterranean hydrological and micropaleontological records suggest the occurrence of regional (pan-Mediterranean) perturbations, rather than local western Mediterranean phenomena, for explaining the ORL termination.…”

High resolution paleo-environmental changes during the Sapropel 1 in the North Ionian Sea, central Mediterranean

“…section 3.4). The comparison of the recalibrated stable isotope G. ruber records shows that in the study core (ND14Mbis) δ 18 O reaches the lighter interglacial values (–0.22–0.95‰) at 10 kyrs BP (Figure 3a), in Filippidi and De Lange (2019) at 10 kyrs BP (north Ionian Sea, Figure 3c), in Geraga et al (2008) at 9.5 kyrs BP (north eastern Ionian Sea, Figure 3b), in Siani et al (2010) at 9 kyrs BP (South Adriatic Sea, Figure 3d) and in Rohling et al (2002) at 9.5 kyrs BP (South Aegean Sea, Figure 3e). All these records, show relatively constant δ 18 O values around −0.3–1.2‰ throughout the Early Holocene.…”

“…The δ 18 O G.ruber signal shows relatively heavy isotopic ratios (1.25–2.35‰) at the base of the studied interval (11.3–6 kyrs BP) followed by a trend towards lighter δ 18 O values that culminates at the onset of S1a (10 kyrs BP) with relatively light values (−0.22−0.95‰) (Figure 3a). This record has been compared with nearby deep-sea cores (Filippidi and De Lange, 2019; Geraga et al, 2008; Rohling et al, 2002 and Siani et al, 2010; Figure 3). To prevent age model discrepancies between core sites in the data comparison we have recalculated the radiocarbon age models for the published δ 18 O G.ruber records following the same Bayesian age model approach adopted for the study core (cf.…”

“…δ 18 O G.ruber data of core ND14Mbis represented in black, compared to δ 18 O G.ruber data from the north Ionian Sea of core Z1 (in orange, Geraga et al, 2008), core MP50PC (in purple and in black the mean values, Filippidi and De Lange, 2019) and the South Aegean Sea core LC-21 (in pink, Casford et al, 2002), andδ 18 O G .bulloides data from core MD90-917 (in green, Siani et al, 2010), from the Adriatic Sea. Black, orange, purple, green and pink dots correspond to each tie point used in the age model of each core.…”

“…Map of the study area showing the modern surface (white arrows) and deep (blue arrow) circulation patterns based on Rohling et al (2009), and also core ND14Mbis location. Other reference core sites used in the text for comparison (LC-21, Casford et al (2002) and Rohling et al (2002); ODP Site 963, Sprovieri et al (2003); M2 and M3, Roussakis et al (2004); Z1, Geraga et al (2008); MD90-917, Siani et al (2010); MP50PC, Filippidi et al (2016) and Filippidi and De Lange (2019); SK-1 and NS-14, Kontakiotis (2016); AEX-15, Giamali et al (2019) are also shown.…”