Abstract
Multi-proxy analyses and lithology of two cores, MRS-CS18 and MRS-CS27, from the İmralı Basin of the Sea of Marmara (SoM) provide novel information on environmental conditions, relative sea level, and sill depths of the straits of Bosporus and Dardanelles during the Marine Isotope Stages (MIS) 5 and 6. The fossil and multi-proxy geochemical records show that lacustrine conditions prevailed in the SoM during most of MIS 6, from 171 to 134 ka BP, and that the transition to marine conditions during Termination II took place at ∼134.06 ± 1.10 ka BP. MIS 5 interstadials a, c, and e witnessed the formation of three sapropels (MSAP-2, MSAP-3 and MSAP-4) under suboxic to anoxic marine conditions, whereas during stadials MIS 5b (∼94–86) and MIS 5d (∼112–105 ka BP), lacustrine and marine conditions with deposition of sediments having relatively low TOC contents (<2%) prevailed, respectively. Consideration of the global sea level, together with the timing of the marine reconnection of the SoM during Termination II and persistence of the marine conditions during MIS 5, except for MIS 5b and later part of MIS 5a, suggests that the Dardanelles sill depth was at ∼ -75 ± 5 m during the reconnection at Termination II and at −55 ± 5 m during most of MIS 5. On similar considerations of the Black Sea marine reconnections and disruptions during the MIS 5, a sill depth of −35 to −40 m (similar to the present day depth) is indicated for the Bosporus Strait. The SoM geochemical proxy records correlate well with the regional terrestrial and marine records and the NGRIP oxygen isotope record with its Stadial and Interstadial phases, showing the common effect of the North Atlantic climatic events triggered by the perturbations in the Atlantic Meridional Overturning Circulation. However, the amplitude of the oscillations recorded in the SoM during MIS 6 (Penultimate Glacial Period) is relatively small compared to the MIS 4 to MIS 2 (Last Glacial Period).
| Original language | English |
|---|---|
| Pages (from-to) | 124-141 |
| Number of pages | 18 |
| Journal | Quaternary Science Reviews |
| Volume | 220 |
| DOIs | |
| Publication status | Published - 15 Sept 2019 |
Bibliographical note
Publisher Copyright:© 2019
Funding
Considering the global sea level curve of Grant et al. (2012) (Fig. 8d), the marine connection of the SoM at ∼134 ka BP and the disconnection during MIS 5b would occur if the sill depth of the Dardanelles Strait was below ∼ −80 m (i.e. lower than the present sill depth of −65 m) during Termination II. A similar sill depth of −84 m for the Dardanelles Strait existed during Termination I (MIS 2/MIS 1 transition), as testified by the widespread existence of paleoshorelines of the Marmara “Lake” in the form of berms and wave-cut notches, which are dated at ∼12.5 cal ka BP (Çağatay et al., 2003, 2009; Polonia et al., 2004; McHugh et al., 2008; Eriş et al., 2011). This implies that the Dardanelles Strait's sill depth during the glacial periods was deeper than that during the interglacial periods, probably due to fluvial down-cutting under a relatively low base level; the isostatic effect of increased water column on the strait's channel after the marine reconnection is estimated to be negligible because of limited width of the strait (Smith et al., 1995). The fluvial-downcutting during the glacial periods is supported by high resolution seismic data of Gökaşan et al. (2008, 2010).Although, the sapropels in the SoM, Black Sea and Mediterranean Sea are related to the same climatic events of global warming and sea level rise, the causal mechanisms and timings of their deposition in these basins were different. The sapropel deposition in the SoM (and the Black Sea) appears to have been triggered by marine flooding events, the timing of which is determined by the rate of sea level rise and the sill depth of the connecting shallow straits of Bosporus and Dardanelles at the time (Çağatay et al., 2009, 2015). The marine flooding of the fresh-brackish basins of the SoM and Black Sea would cause a density stratification of the water column, decreased renewal of deep waters and increased organic matter preservation, all leading to the sapropel formation (Çağatay et al., 2009, 2015). On the other hand, the sapropel formation in the Mediterranean is believed to have been controlled by sea level rise and precession minima, with the former causing preconditioning and prevention of deep water formation, and the latter enhanced monsoon intensity and runoff from North Africa, resulting in density stratification and stagnation (see Rohling et al., 2015 for discussion). The Mediterranean sapropels S3, S4 and S5 were formed during interstadial periods of MIS 5 (a, c, and e), whereas sapropel S6 was deposited during ∼180-171 ka BP within the glacial MIS 6 (Grant et al., 2012). Thus, MSAP2, MSAP-3 and MSAP-4 sapropels of the SoM are broadly the time equivalent of S3, S4 and S5 sapropels in the Mediterranean, but the equivalent of Mediterranean S6 sapropel should be absent in the SoM and the Black Sea, based on the considerations of the global sea level curve (Grant et al., 2012) (Fig. 8d) and the Sofular speleothem oxygen isotope data (Badertscher et al., 2011) (Fig. 8f).Although sapropels were deposited in the same interstadial sub-stages of the MIS 5 and during MIS 1 in the SoM, Black Sea and Mediterranean Sea, the timing of their formation was different in these basins. The onset of MIS 5 sapropels, MSAP-2 and MSAP-4, in the SoM is dated at 86.6, and 131.2 ka BP, while that of MSAP-3 could not only be constrained between 108.2 and 103.3 ka BP, because of its disconformable basal boundary. The sapropels S3, S4 and S5 in the Mediterranean Sea started forming at ∼85 ka BP, ∼107 ka and ∼128 ka BP, respectively (Grant et al., 2012), and the MIS 5e (Eemian) sapropel in the Black Sea at 127.6 ka BP (Wegwerth et al., 2014). Therefore, the onset ages of formation of the two SoM MIS 5 sapropels are in general 1–3 ka older than the corresponding Mediterranean sapropels of MIS 5. Similarly, the Holocene MSAP-1 sapropel started forming at 12.33 ± 0.35 (Çağatay et al., 2015), nearly 1.5 ka earlier than the Mediterranean S1 sapropel (10.8 cal yr BP; De Lange et al., 2008), ∼4 ka earlier than the onset of the formation of the Holocene Black Sea sapropel, dated at ∼8 cal yr BP (Jones and Gagnon, 1994; Bahr et al., 2005, 2006; Lamy et al., 2006). The formation of sapropels in the SoM being earlier than those in the Mediterranean sapropels does not necessarily indicate a causal link between the Black Sea outflow and the sapropel formation in the Eastern Mediterranean, as claimed by some earlier studies (Olausson, 1991; Aksu et al., 1995). They suggest that substantial freshwater outflows from glacial Black “Lake” after each marine reconnection was the trigger for sapropel formation in the Eastern Mediterranean. However, this hypothesis is now mainly discarded especially for the sapropel S1 formation, because it was demonstrated that the sea surface salinity was up to 6 psμ higher-than-present in the SoM during the deposition of S1 and that the salinity gradients indicated a southern Mediterranean freshwater source in the Levantine Basin (Sperling et al., 2003; Vidal et al., 2010). Instead, the differences in timing of the sapropel deposition in the silled SoM and Black Sea marginal basins, relative to those of the quasi-equivalent Mediterranean sapropel, are likely related to timing of establishment of the suboxic/anoxic bottom water conditions, which would in turn be controlled by the sill depth of the connecting shallow straits, rate of primary productivity, the rate and extent of sea level rise, and the size and depth of the basins. The shallow sill depths of straits would delay the marine connections, but would result in establishment of early anoxic conditions, thereby shortening the time lag between the marine connection and the onset of sapropel deposition in the marginal silled basins. Similarly, high sea levels with a rapid rise, as in the case of MIS 5e would result in early connections, but cause a delay in the sapropel formation because of the relatively effective bottom water circulation in small basins such as the SoM. This would explain the difference in the time lapse between the marine connection and the formation of sapropels during MIS 5e (MSAP-4) and Holocene (MSAP-1), which is ∼3 ka (this study) and ∼0.3 ka (Çağatay et al., 2015), respectively.As already explained in section 3.4.2, the μ-XRF Ca data, related to the endogenic carbonate production in the epilimnion of the Marmara “lake”, show good correlation with the NGRIP oxygen isotope data (NGRIP, 2004), with the increase (decrease) in Ca corresponding to Greenland Interstadials (Stadials) ( Fig. 3a–c, 8a, d). There is also a very close correlation between the TOC data of the SoM with the continental records from Lake Ohrid in Macedonia/Albania in the Balkans (i.e. TOC data of Francke et al., 2016 and pollen data of Sinopoli et al., 2018) (Fig. 8f, i, j). The SoM and Lake Ohrid are located ∼500 km apart along similar latitudes and are presently both affected by continental and Mediterranean climatic conditions. The TOC data of SoM and TOC and pollen data of Lake Ohrid show conformable sharp changes during the stadial and interstadial periods of MIS 5, with high organic productivity and preservation during the interstadial periods in both basins. The SoM data also show excellent correlation with the global sea level curve (Fig. 8d) and oxygen isotope ratio of pelagic foraminiferal data from the SE Aegean Sea (Grant et al., 2012) (Fig. 8d), surface temperature (SST) data from core ODP-977 (Martrat et al., 2007) (Fig. 8j), and benthic and planktonic oxygen isotope data from the Iberian Margin (Sanchez-Goni et al., 1999, 2005, 2012; not plotted), as well as with the pollen and TOC data of the Black Sea for MIS 5e that show warm and high precipitation conditions during 126.4–122.9 ka BP in north Anatolia (Shumilovskikh et al., 2013b; Wegwerth et al., 2014) (Fig. 8e).The authors would like to thank the scientific team of Marsite cruise, and in particular the co-chiefs, Louis Geli and Livio Ruffine, and the captains and crews of RV Pourquoi pas? The Marsite cruise was co-funded by the EC FP7 project MARSITE (grant number: 308417) “Longterm monitoring experiment in geologically active regions of Europe prone to natural hazards:the Supersite concept” under the call ENV.2012.6.4-2 and by the “Laboratoire d'Excellence” LabexMER (ANR-10-LABX-19) through the projects called MicroGaMa and MISS Marmara, and by a grant from the French government under the program “Investissements d'Avenir”. The authors also thank the two anonymous JQSR reviewers for their constructive comments that improved the paper. The authors would like to thank the scientific team of Marsite cruise, and in particular the co-chiefs, Louis Geli and Livio Ruffine, and the captains and crews of RV Pourquoi pas? The Marsite cruise was co-funded by the EC FP7 project MARSITE (grant number: 308417 ) “Longterm monitoring experiment in geologically active regions of Europe prone to natural hazards:the Supersite concept” under the call ENV.2012.6.4-2 and by the “Laboratoire d'Excellence” LabexMER (ANR-10-LABX-19) through the projects called MicroGaMa and MISS Marmara, and by a grant from the French government under the program “Investissements d'Avenir”. The authors also thank the two anonymous JQSR reviewers for their constructive comments that improved the paper.
| Funders | Funder number |
|---|---|
| LabexMER | ANR-10-LABX-19 |
| SE Aegean Sea | ODP-977 |
| BP | 103.3, 8a, MSAP-4, 180-171, MSAP-3, MSAP-2 |
| Seventh Framework Programme | 308417 |
| Seventh Framework Programme | |
| scientific team of Marsite cruise |
Keywords
- Eastern Europe
- Geochemical proxies
- Late quaternary
- Marine isotope stages 5–6
- Paleoceanography
- Sapropels
- Sea of Marmara
- Sedimentology-marine cores
