Decoding sea surface and paleoclimate conditions in the eastern Mediterranean over the Tortonian-Messinian Transition

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

Contents lists available at ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology journal homepage: www.elsevier.com/locate/palaeo

Decoding sea surface and paleoclimate conditions in the eastern Mediterranean over the Tortonian-Messinian Transition

T

⁎

G. Kontakiotisa, , E. Besioua, A. Antonarakoua, S.D. Zarkogiannisa, A. Kostisa, P.G. Mortynb,c, P. Moissettea,d, J.-J. Cornéee, C. Schulbertf, H. Driniaa, G. Anastasakisa, V. Karakitsiosa a

Faculty of Geology & Geoenvironment, School of Earth Sciences, Department of Historical Geology-Paleontology, National & Kapodistrian University of Athens, Panepistimiopolis, Zografou 15784, Greece b Institute of Environmental Science and Technology (ICTA), Universitat Autonoma de Barcelona (UAB), Edifici Z - Carrer de les Columnes, Bellaterra 08193, Spain c Department of Geography, Universitat Autonoma de Barcelona (UAB), Spain d Muséum National d'Histoire Naturelle, Département Origines et Evolution, UMR7207 CR2P, 8 rue Buffon, 75005 Paris, France e Géosciences Montpellier, Université des Antilles-Université de Montpellier-CNRS, Pointe à Pitre, FWI, France f GeoZentrum Nordbayern, Friedrich-Alexander Universität Erlangen-Nürnberg, 91054 Erlangen, Germany

A R T I C LE I N FO

A B S T R A C T

Keywords: Siphon cooling event Messinian Salinity Crisis (MSC) Sea surface salinity (SSS) Sr/Ca paleothermometry Late Miocene Hydrological cycle

New sedimentological, micropaleontological and geochemical data from the Upper Miocene pre-evaporitic sedimentary sequence of the Faneromeni section (Crete Island, eastern Mediterranean) revealed a stepwise restriction of the Mediterranean Sea preceding the Messinian Salinity Crisis (MSC), which was modulated by a sedimentary cyclicity responding to orbital parameters. This cyclicity is manifested by lithological alternations from laminated to indurated homogeneous marls and clayey limestones, and covers the Tortonian-Messinian Transition (TMT; 7.6–6.7 Ma). This time window covers the successive closure of the marine MediterraneanAtlantic gateways, which culminated in the onset of the MSC. In the present study, we present the first evidence for changes in the upper water column reflected by sea surface temperature (SST) and salinity (SSS) variations that correlate with pronounced paleoclimatic fluctuations. Planktonic foraminiferal isotopes, in combination with paired mixed layer Sr/Ca-derived SST data, reveal that the very warm late Tortonian interval has been followed by a strong long-term cooling (~10 °C) and desalination (~10‰) trend during the earliest Messinian, attributed to the paroxysmal phase of the so-called “siphon” event. In particular, the climate shift that occurred at the end of a global carbon isotope (δ13C) decrease suggests that changes in the carbon cycle were instrumental in driving late Miocene climate dynamics (cooling and aridity) in the progressively isolated eastern Mediterranean Sea. The observed salinity variability during this time interval also provides further insights about seasonal freshwater inputs and gives new support to the much-debated hydrologic regime (linear salinity increase vs step-function evolution with strong salinity fluctuations) preceding the deposition of evaporites. The novel methodology of foraminiferal Sr/Ca paleothermometry and results of this study could have numerous potential applications to other regions and relevant extreme geological events. Therefore, in the near future we expect this approach to add important new information to our understanding of Neogene climates.

1. Introduction The latest Miocene (8.0–5.33 Ma) is considered as a climatically stable period, characterized by minor long-term (> 1 Myr) cooling and ice growth episodes compared to the entire Neogene (Zachos et al., 2001). However, these long-term trends are punctuated by the Late

Miocene Carbon Isotope Shift (LMCIS, 7.6–6.6 Ma) during the Tortonian and the Messinian Salinity Crisis (MSC, 5.97–5.33 Ma), which have been attributed to changes in the carbon cycle, ocean circulation, and global sea-level variations related to changing ice volume (Bickert et al., 2004; van der Laan et al., 2005; Hüsing et al., 2009; Ohneiser et al., 2015). The LMCIS is the last period of Earth's history that

⁎

Corresponding author. E-mail addresses: [email protected] (G. Kontakiotis), [email protected] (A. Antonarakou), [email protected] (S.D. Zarkogiannis), [email protected] (A. Kostis), [email protected] (P.G. Mortyn), [email protected] (P. Moissette), [email protected] (J.-J. Cornée), [email protected] (C. Schulbert), [email protected] (H. Drinia), [email protected] (G. Anastasakis), [email protected] (V. Karakitsios). https://doi.org/10.1016/j.palaeo.2019.109312 Received 11 June 2019; Received in revised form 26 July 2019; Accepted 1 August 2019 Available online 07 August 2019 0031-0182/ © 2019 Elsevier B.V. All rights reserved.

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

G. Kontakiotis, et al.

while other relevant proxies (e.g., TEX86, GDGTs, LDI) for subsurface production and/or input of terrestrial GDGTs, by freshwater input and oxic degradation (Huguet et al., 2007; Kim et al., 2015; Rodrigo-Gámiz et al., 2016; De Bar et al., 2016). Moreover, the above proxies do not necessarily reflect the conditions (depth, season) in which the key planktonic foraminiferal species calcified (Huguet et al., 2006; Lopes dos Santos et al., 2013; Smith et al., 2013; Jonas et al., 2017) and are thus limited in precision for corrections of the isotopic temperature to resolve δ18OSW. The Sr/Ca paleotemperature proxy is mainly linked to changes in growth rates of foraminifera, as a response to changing environmental conditions (Kisakürek et al., 2008). Temperature dependence of Sr incorporation in planktonic foraminifera appears to be species dependent (Mortyn et al., 2005), and its indirect effect on shell Sr/Ca is related to precipitation rates. However, as culture experiments have shown, shell precipitation in foraminifera is a variable incremental function, with the extreme salinity and temperature conditions indicative of low growth rates outside the natural variability usually found in field studies. We note that although absolute precipitation rates cannot be constrained in late Miocene planktonic foraminiferal species, we examine for the first time the SSTs inferred from Sr/Ca ratios in conditions quite similar (extreme SSTs and SSSs) to those of the laboratory. In this contribution, the upper calibration limit of this approach is extended up to 30 °C, which is the main advantage against the Uk′ 37 proxy. Therefore, in this setting the planktonic foraminiferal Sr/Ca likely represents the best alternative for late Miocene Mediterranean SST reconstructions, since it provides the least amount of bias (Sprovieri et al., 2008) and further produces values within those known from the literature temperature range. We present the first continuous record of late Miocene eastern Mediterranean SST and SSS based on the combined foraminiferal δ18OSr/Ca approach, recovered from sampling the Faneromeni section (Crete Island, eastern Mediterranean). The record was also evaluated for dissolution by X-ray microcomputed tomography. Outcropping marine sediments in Crete provide windows into the Mediterranean that have not been recovered by deep-sea drilling, because the upper Miocene sedimentary record from this area is buried under thick salt deposits. Consequently, this work focuses on and further extends the application of the foraminiferal Sr/Ca paleothermometer to upper Miocene Mediterranean sections, and inserts the climate history of this setting into the framework of global climate through this time period.

experienced a long-term positive carbon excursion coincident with a period of extreme warm climate with tropical-like temperatures and high sea-level before the transition to the Neogene glacial-interglacial climate mode and cryospheric circulation system (Böhme et al., 2008). The MSC represents the most significant event in the history of the Mediterranean, during which the basin experienced a dramatic hydrological and biological crisis induced by a powerful combination of geodynamic and climatic drivers (Salé et al., 2012; van der van der Laan et al., 2012; Roveri et al., 2014; Leroux et al., 2018). The change of the Mediterranean connections with both the Atlantic and Paratethyan basins, caused high-amplitude fluctuations in the hydrology of its basins (Flecker et al., 2015; Karakitsios et al., 2017a; Krijgsman et al., 2018; Vasiliev et al., 2019), which had a great impact on the subsequent geological history of the Mediterranean area, and on the salinity of the global oceans. Although the triggers remain debated as a climate, tectonic or combined cause, a consistent pattern regarding the chronology, stratigraphy and progression of the desiccation has started to emerge (Roveri et al., 2014; Manzi et al., 2013, 2016, 2018; Tulbure et al., 2017; Karakitsios et al., 2017a, 2017b; Krijgsman et al., 2018; Lozar et al., 2018; Gennari et al., 2018; Capella et al., 2018). In the Mediterranean Sea, restricted conditions started well before the MSC. According to foraminiferal Sr data, the Atlantic-Mediterranean water mass exchange started to be increasingly severe about 3 Myr earlier (at ~9 Ma) than the onset of evaporite accumulation, resulting in the salinity increase firstly in the deepest basins at ~7.2 Ma and finally in the surface waters at ~6.8–6.7 Ma (Kouwenhoven et al., 2003). The geological expression of this evolution was the cyclic marl/sapropel succession because freshwater input was not constant but strongly pulsed at precessional time-scales (Hilgen et al., 1995; Lourens et al., 1996; Schenau et al., 1999; van der van der Laan et al., 2012). The precession-controlled cycles persisted up to the MSC and continued after the reflooding of the basin during the basal Pliocene (Langereis and Hilgen, 1991; Hilgen and Krijgsman, 1999; Hüsing et al., 2009; Kontakiotis et al., 2016a). Associated salinity and/or water temperature changes are well documented in various micropaleontological and geochemical datasets from the circum-Mediterranean region (Kouwenhoven et al., 2003; Tzanova et al., 2015; Reghizzi et al., 2017; Vasiliev et al., 2017, 2019; Moissette et al., 2018). The planktonic foraminiferal analysis together with the potential of Mediterranean sequences as paleoclimate archives (e.g., Kontakiotis et al., 2013; Lirer et al., 2014; Kontakiotis, 2016; Drinia et al., 2016; Koskeridou et al., 2017; Antonarakou et al., 2018; Louvari et al., 2019; Giamali et al., 2019) remain to be exploited because of difficulties with current SST proxy reconstructions in the Mediterranean Sea (Kontakiotis et al., 2011, 2016b). Stable oxygen isotope ratios (δ18O) cannot resolve regional Mediterranean SST unambiguously due to the large and underconstrained isotopic component attributable to hydrological variability in the Mediterranean Sea (Kouwenhoven et al., 1999; Di Stefano et al., 2010; Kontakiotis, 2016). The foraminiferal Mg/Ca proxy, which when combined with δ18O allows simultaneous reconstruction of temperature and oxygen isotopic composition of seawater (δ18Osw) using the same signal carrier, is also compromised by burial diagenesis and salinity under high evaporitive conditions, such as those in the eastern Mediterranean basin (Kontakiotis et al., 2011, 2016b, 2017). Similarly, although alkenone unsaturation ratio (Uk′ 37) appears unaltered by diagenesis and lithification (Cleaveland and Herbert, 2009; Beltran et al., 2011), this proxy becomes insensitive to temperature above 27.5 °C (Müller et al., 1998) because alkenoneproducing haptophytes synthesize only trace amounts of triunsaturated alkenones. Mediterranean SSTs are very near (reaching or exceeding) the upper Uk′ 37 sensitivity limit during most of the late Miocene (Tzanova et al., 2015), suggesting inaccuracies of the potential estimated temperatures. For instance, the Uk′ 37 might be affected by nutrient availability, lateral transport, and oxic degradation (Gong and Hollander, 1999; Prahl et al., 2003; Rontani et al., 2013; Lattaud et al., 2018),

2. Geological setting 2.1. Upper Miocene sedimentary environments of Crete Neogene sediments on Crete (e.g., Faneromeni, Kastelli, Potamida, Skouloudhiana, Moni Gorgolaini, Makrilia, and Vrysses sections) have been studied in multidisciplinary fashion by several researchers (Krijgsman et al., 1994; Kouwenhoven et al., 2003; Drinia et al., 2007, 2008; Zachariasse et al., 2011; Brachert et al., 2015; Zidianakis et al., 2015, 2016; Zelilidis et al., 2016; Agiadi et al., 2017; Moissette et al., 2018; Antonarakou et al., 2019), and are found in a mosaic of faulted basins separated by alpine basement (Zachariasse et al., 2008), as a result of intense extensional geodynamic processes in the course of the middle-late Miocene (Reuter et al., 2006). Particularly in eastern Crete, the oldest dated sediments are the alternations of upper Tortonian to lower Messinian marine marls and sapropels of the Faneromeni-Sitia area, reflecting dry-wet climatic conditions paced by the cycle of precession and modified by obliquity and eccentricity on longer time scales (Hilgen et al., 1995; Krijgsman et al., 1995). At several places in eastern Crete (including the study area), these distinct sedimentary cycles pass upward either into an alternation of whitish calcareous marls and sapropels, or into shallower calcarenites with occasional calcareous to sapropelitic marls (Krijgsman et al., 1994, 1995; Moissette et al., 2018). 2

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

G. Kontakiotis, et al.

A

Monte del Casino Monte dei Corvi Italy

Atlantic ocean

Turkey Spain

Greece

nean BETICS terra i d e ern M West Algeria RIF 1000 km

FANEROMENI

Gibliscemi

Metochia Eastern Mediterranean CRETE Lybia B

Chania Heraklion

FANEROMENI Sitia

20 km Preneogene

Neogene

C

MESSINIAN

TORTONIAN

Fig. 1. A) Location of the sections discussed in the text within the Mediterranean Sea: the Monte dei Corvi and Monte del Casino sections in northern Italy, the Gibliscemi section in Sicily, the Metochia section in Gavdos, and Faneromeni section in Crete Island, B) geological sketch map of Crete island and location of the study section, C) general panoramic view of the Faneromeni section. The orange line marks the distinction between the Tortonian and the Messinian. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

3

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

G. Kontakiotis, et al.

Krijgsman et al. (1994 and 1995) Magnetostratigraphy: ATNTS 2012 chrons

Moissette et al. (2018) m

Sedimentary cycles

Slump 2 40 39 38

C3An.2n

50

37 36 35

6.72 Ma

38

LO G. nicolae

6.72 Ma

34

Slump 1

33 32 31 30 29

6.83 Ma

30

28 27 26 25

C3Ar

FO G. nicolae

6.83 Ma 40

MESSINIAN

24 23

22 21 20

30

18

C3Bn

20

7.24 Ma

17

FCO G. miotumida group

16

FO G. conomiozea

7.24 Ma

15

G. menardii 4 PE G. scitula group (d)

14 7.28 Ma

C3Br.1r

13 7.36 Ma 20

7.28 Ma

C3Br.1n

Shear plane

7.36 Ma

FO G. menardii 5

12 11 10 9

10

FO G. menardii 5 8 7

7.45 Ma

C3Br.2r

TORTONIAN

un de t.

19

LO C. parvulus

6

10 5 4

7.51 Ma

LCO G. menardii 4

7.58 Ma

G. scitula group (d to s)

3

7.45 Ma

C3Br.2n

7.51 Ma

2

LCO G. menardii 4

1

C3Br.3r

1 0

t. de un

C4n.1n Sandy limestone/sandstone Halimeda limestone Clayey limestone Marl Laminated marl/sapropel

7.58 Ma

(caption on next page) 4

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

G. Kontakiotis, et al.

Fig. 2. Litho-chrono-stratigraphic correlations between the logs of the Faneromeni section, as previously described by Krijgsman et al. (1994, 1995) and Moissette et al. (2018). To the right of the studied stratigraphic column (Moissette et al., 2018) are the positions of the studied samples. Biostratigraphic events coupled with available magnetostratigraphic data (Krijgsman et al., 1994, 1995; Moissette et al., 2018) are also indicated for the studied time interval. The black lines represent the boundaries of the 40 sedimentary cycles recognized, which are symbolized with squared numbers and are progressively numbered from the base to the top of the studied interval. The red line represents the Tortonian/Messinian (T/M) boundary. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

2.2. Dataset, lithostratigraphy and age model of the Faneromeni section

3. Material and methods

In the eastern Mediterranean, the Faneromeni section (35° 13′ 25.48″N; 26° 03′ 49.63″E) is located on northeastern Crete (Fig. 1A, B), and consists of pre-evaporitic sediments covering the upper Tortonian and lowest Messinian (Fig. 1C), as confirmed by previously applied biomagneto-stratigraphic schemes (Krijgsman et al., 1994, 1995; Negri and Villa, 2000; Moissette et al., 2018). The approximately 57 m thick section is well exposed and undisturbed by faults, as previously described by Krijgsman et al. (1994, 1995), Nijenhuis et al. (1996), Negri and Villa (2000), and Moissette et al. (2018). In Fig. 2 is illustrated the stratigraphic correlation between the studies of Krijgsman et al. (1994, 1995) and Moissette et al. (2018). Overall, the pre-evaporitic deposits of the studied outcrop (Moissette et al., 2018) are characterized by welldeveloped bipartite sedimentary cycles, which consist of sub-horizontally bedded, whitish to greyish indurated- or dark greenish-brown to black laminated- (referred to here as “sapropels”) carbonate-poor marls alternating with homogeneous hemipelagic marls and clayey limestones (Fig. 2). The nature of the sedimentary cycles closely resembles those observed in other astronomically calibrated late Miocene Mediterranean sequences, where 40 precessional cycles were evidenced between 7.6 and 6.6 Ma (Hilgen et al., 1995, 2000). A thin interbed of sandy limestone is also noticeable in the middle of the section (at ~30 m). Sparse thin-shelled pectinids are present (among which Amussium and Chlamys), and several levels with concentrations of other bivalve shells (mostly Neopycnodonte) also occur throughout the section. The color of the homogeneous beds changes from blue-grey in the lower part to dark greenish-brown in the upper half of the section, with the color contrast between homogeneous and laminated sediments gradually lost in the uppermost cycles. Laminated sediments yielded fish scales, sponge spicules and some terrestrial leaves. In the upper half of the section, the Messinian cycles are thicker than those of the Tortonian, with the homogeneous marl layers relatively thicker and the sapropels thinner. Near the top of the section, the carbonate part of the sedimentary cycles is often bioturbated. Firmgrounds and brecciated slump levels occur below a Halimeda rich floatstone. The top of the marine succession is generally truncated by erosion surfaces and overlain by Quaternary continental deposits. The chronology of the studied section is based on the planktonic foraminiferal biostratigraphy published in Moissette et al. (2018). The above astronomically calibrated biostratigraphic events were also used as reference points in order to check the correct sequence of sedimentary cycles, with reference to the works of Krijgsman et al. (1994, 1995). Finally, the resulting stratigraphic framework is established by linear interpolation between the control points (Fig. 3). Sedimentation rates and sapropel occurrences guided our sampling strategy in order to address long term and orbital-scale changes through key late Miocene intervals. We sampled the sequence at a minimum of one sample per cycle, a resolution of approximately 40 kyr. Our sampling starts 190 kyr after the Tortonian Salinity Crisis (TSC at 7.8 Ma; Krijgsman et al., 2000) and terminates ~700 kyr before the onset of the MSC (5.97 Ma; Manzi et al., 2013, 2018). Therefore, the studied sediments correspond to the crucial time period of the Tortonian-Messinian Transition (TMT) for the paleoceanographic history of the Mediterranean basin, with particular emphasis on the interval reflecting the Tortonian/Messinian (T/M) boundary.

3.1. Micropaleontological analyses Thirty-eight samples were collected and processed following standard micropaleontological procedures. The dried bulk samples were weighed, washed over a 63-μm sieve, oven-dried at 50 °C, and residues were sieved into sub-fractions (125, 250 and 300 μm) for both micropaleontological (planktonic foraminiferal assemblages and biostratigraphy) and geochemical (isotopes and trace elements) analyses respectively. Qualitative and quantitative analyses have been performed on planktonic foraminiferal assemblages for the > 125 μm fraction, split into aliquots, each one containing at least 300 specimens. All handpicked shells were identified and counted in each sample, and then converted into percentages, based on the extrapolation of a counted split (Fig. 4). Moreover, the shell masses of 50 G. obliquus specimens were measured with a Sartorius microbalance MP2 in order to assess their calcification variations. 3.2. X-ray micro-CT Two randomly picked specimens of the planktonic foraminifera G. obliquus (one from the Tortonian and the other from the Messinian) were analyzed using X-ray microcomputed tomography (XMCT) in order to test their preservation. The first specimen was from sample 18 for which high shell masses were recorded and the second from sample 23 with low shell mass values. The tests were mounted together on a glass rod 1 mm in diameter with tragacanth gum. The scans were performed at a High-Resolution X-Ray System GE/Phoenix v|tome|x s 240 CT scanner at the Paleontology Department of GeoZentrum Nordbayern in Erlangen, using a nano-focus 180 kV–15 W high power X-ray tube. A high-resolution setting (voltage of 80 kV, current 80 μΑ, detector array size of 1024 × 1024, 1500 projections/360°, 2.5 s/projection) enabled to obtain 3-D images with an isotropic pixel size of about 1.1 μm. 3.3. Stable isotope and trace element analyses Stable oxygen and carbon isotope measurements (δ18O, δ13C) were performed on the entire set of collected samples (38 samples) comprising 20 pre-weighed Globigerinoides obliquus specimens from the 250–300 μm size fraction. This size fraction limitation was used to minimize ontogenetic, growth rate and size effects on shell weight and geochemistry (Elderfield et al., 2002). The choice of the shallowdwelling species G. obliquus for isotopic and trace element analyses is based on the following factors: a) it is abundant and continuous throughout the TMT, b) it represents a reliable recorder of both global climate forcing and local climate-ocean influences in the Miocene Mediterranean basin (Sprovieri et al., 1999; Sierro et al., 2003; Antonarakou et al., 2007; Drinia et al., 2007), c) it has similar ecological characteristics, food requirements and the same mineralization as that of Globigerinoides ruber (Hemleben et al., 1989; Lourens et al., 2004; Williams et al., 2005), and d) in contrast to the late Quaternary (Antonarakou et al., 2015), the results on Neogene time periods are not dependent on the choice of surface-dwelling planktonic foraminiferal species (Medina-Elizalde and Lea, 2005; Kontakiotis et al., 2016b). Therefore, G. obliquus shells appear particularly valuable and consequently constitute a very important archive for studying changes in the Mediterranean surface waters during that time. Once picked, samples 5

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

G. Kontakiotis, et al.

Age (Ma) 7.7 60

7.6

7.5

7.4

7.3

7.2

7.1

7

6.9

Tortonian

6.8

6.7

6.6

6.5

Messinian

38

50

35 34

6.0

9c

m

33 32 31 30

3.4

30

9 4.

cm

y /k

/ cm 4(

ky

28 27 26

r)

25 24 23

8

r

22 21 20

16 15

/k yr

18 17

20

4

14 13

r

6.0

r /ky5-6 cm 5 . 3

cm

19

7

12 11

8 7

7.7

10

3

10 9

7.3

cm /k y r

cm /k y

Stratigrraphic level (m)

29

9

40

6 5

r

2

Schematic log of Faneromeni pre-MSC section

36

/k

yr

37

10

1

6 .7

cm

/ky

4

3 2 1

0

C4n. 1r

C4n.1n

C3Br. C3Br. 3r 2n

C3Br.2r

C3Br. C3Br. C3Bn 1n 1r

C3An

C3An.2n

APTS Fig. 3. Time stratigraphic framework based on planktonic foraminiferal biostratigraphy (Moissette et al., 2018) and previously published magnetogstratigraphy (Krijgsman et al., 1994) for the pre-MSC interval of the Faneromeni section. The age model is established by linear interpolation between the astronomically calibrated planktonic foraminiferal bioevents, which are indicated with numbers 1 to 10 into the blue circles. The biostratigraphic tie-points used for the age model are: (1) G. scitula group coiling change D/S, (2) Last Common Occurrence of G. menardii 4 (sin), (3) Last Occurrence of C. parvulus, (4) First Occurrence of G. menardii 5 (dex), (5) Influx of G. menardii 4 within the range of G. menardii 5, (6) Paracme end of G. scitula group (dex), (7) First Common Occurrence of G. miotumida group, (8) Last Common Occurrence of G. scitula group (sin), (9) First Occurrence of G. nicolae, (10) Last Occurrence of G. nicolae. The average sedimentation rates corresponding to the different time intervals in between the above tie-points are also indicated. The schematic lithological column, modified after Moissette et al. (2018), is on the right-hand side. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

phases. Foraminiferal trace element cleaning procedures and element/ Ca (e.g., Mg/Ca, Sr/Ca) ratio analyses were performed at the Universitat Autònoma de Barcelona (UAB). Given that the use of planktonic foraminiferal Mg/Ca ratios as a paleotemperature proxy is confounded by the salinity and overgrowth effects in the eastern Mediterranean (Kontakiotis et al., 2011, 2016b, 2017), and especially during the TMT where both these parameters play a crucial role, we used the Sr/Ca ratios for T reconstruction (following the findings of Hori et al., 2018 where intratest variations in Sr/Ca are independent of Mg/Ca changes). For all samples used in this study, Sr/Ca ratios were converted to SST by means of the species-specific calibration for Globigerina bulloides of Lea et al. (1999):

were sonicated in methanol for ~10 s to remove fossil and clay particles from inside the tests, then crushed and homogenized prior to analysis. Isotopic analyses were performed using an Elemental Analyzer Isotope Ratio Mass Spectrometer (EA-IRMS) at the GeoLab of the Utrecht University (Netherlands). All values are reported in per mil (‰) relative to V-PDB, with an analytical precision better than 0.08 and 0.05‰ for δ18O and δ13C respectively. Reproducibility and accuracy was monitored by replicate analysis of laboratory standards calibrated by assigning δ13C values of +1.95‰ and δ18O values of −2.20‰ to NBS19. For minor and trace metal analysis, 40–50 G. obliquus specimens were picked from the same size fraction as for isotopic analyses. This approach averages the temperature (T) signal recorded by individual tests comprising the sample population (Barker et al., 2003). Prior to cleaning and under microscopic view, the shells were gently crushed, using methanol-cleaned glass plates, to ensure that all chambers were opened without pulverizing the sample, and the samples were loaded into acid-cleaned microvials. The cleaning procedure followed the “Cd method” (Rosenthal et al., 2004), which involves a number of discrete sequential steps with the objective of removing various contaminant

Sr/Ca (mmmol/mol) = 1.04 + 0.011 ∗ SST

(1)

3.4. Seawater oxygen isotope (δ18Οsw) calculation and conversion to sea surface salinity The foraminiferal δ18O record is a function of temperature and the 6

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

G. Kontakiotis, et al. G. obliquus

G. bulloides G. apertura

60

N. acostaensis (d)

G. siphonifera

T. quinqueloba

N. acostaensis (s)

G. menardii 4

O. universa

G. glutinata

G. trilobus

G. scitula gr. (s)

G. scitula gr.(d)

G. parvulus

G. nepenthes

G. miotumida

G. menardii 5

G. conomiozea

50 10

9

Stratigraphic level (m)

40

8

30 7 5

T/M boundary

6

20

4

3

10 2

1

0 0 3060

0 4080

0 3060

0 2 4

0 3 6

0 1530

0 1020

0 1020

0 8 16

0 8 16

0 3 6

0 3 6

0 4 8

0 1020

0 2 4

0 1 2

0 3 6

0 0.5 1(%)

Fig. 4. Frequency curves of the most indicative for the present study planktonic foraminiferal species in the Faneromeni section. The dashed red line represents the Tortonian/Messinian (T/M) boundary. The study of the planktonic foraminiferal fauna displayed a series of bioevents, which are symbolized with circled numbers in blue color and are progressively numbered from the base to the top of the studied interval. (1) G. scitula group coiling change D/S, (2) Last Common Occurrence of G. menardii 4 (sin), (3) Last Occurrence of C. parvulus, (4) First Occurrence of G. menardii 5 (dex), (5) Influx of G. menardii 4 within the range of G. menardii 5, (6) Paracme end of G. scitula group (dex), (7) First Common Occurrence of G. miotumida group, (8) Last Common Occurrence of G. scitula group (sin), (9) First Occurrence of G. nicolae, (10) Last Occurrence of G. nicolae. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

surface water δ18O (δ18OSW) variation at the time of calcification, which in turn depends on ice volume and salinity. To isolate the δ18OSW, the T-driven component was removed using the Orbulina universa low-light paleotemperature equation of Bemis et al. (1998):

T = 16.5–4.80 ∗ (δ18OC − (δ18OSW − 0.27‰))

Globoturborotalita apertura) and G. bulloides, which exhibit a clear antagonistic distribution pattern (Fig. 4). Peaks characterizing the abundance of the mentioned species are almost never exactly coincident; it appears that G. obliquus replaces G. bulloides and vice versa. The latter reaches significant percentages only before or after the G. obliquus blooms, possibly related to nutrient availability or water mixing. Besides the above species, important components of the assemblage are represented by high productivity indicators, such as Orbulina universa, Turborotalita quinqueloba, Globigerinita glutinata and Neogloboquadriniids, which display mean percentages of about 10–20% (Fig. 4). Globorotaliids are found generally in low frequencies (up to 8%) in the middle and upper part of the section. More explicitly, at 24.6 m (7.24 Ma), G. scitula decreases and the representatives of the G. miotumida group slightly increase (Fig. 4). Some additional planktonic species (e.g., G. parvulus, G. siphonifera) occur at low frequencies and are irregularly distributed over the entire section.

(2)

In order to correct for the global ice volume component of the δ18OSW signal, we used the sea-level curve of Miller et al. (2011). The sea-level data were converted into mean ocean δ18O changes applying a 0.008‰ increase per meter of sea-level lowering (Schrag et al., 2002), and were then subtracted from the δ18OSW profile to obtain the regional ice volume free δ18OSW (δ18OIVF-SW), which is considered to approximate local variations in salinity. The propagation error suggests uncertainties of ~0.31‰, which is in accordance with similar errors reported for the δ18OSW residuals during paleosalinity reconstructions (Schmidt, 1999; Antonarakou et al., 2015; Vasiliev et al., 2019). Absolute errors in δ18OIVF-SW can hardly be assessed for the late Miocene, but are expected to be of minor importance as changes in global ice volume were small during this time interval and do not alter the relative changes between records (Williams et al., 2005). As we had no way to estimate the slope of the δ18OSW-salinity relationship for the late Miocene, we used the modern δ18OSW-salinity relationship for the Mediterranean Sea (Pierre, 1999) to convert the δ18O anomaly to sea surface salinity (SSS) values. The limitation in our approach is the caveat of a constant δ18OSW-salinity relationship and the subsequent normalization-to-modern conditions, which ties all past variations to the isotopic values and S-related reconstructions.

4.2. X-ray micro-CT Based on the appearance of the CT images (Fig. 5), both specimens are found to be well preserved as no particular sign of dissolution was observed in the inner chamber walls and no degradation was evident in the internal ultrastructure of foraminiferal tests at the primary calcite layers/linings (Johnstone et al., 2010). The tests are almost equally and relatively little infilled with miscellaneous material, with thus equally minor effect on their weights. The well-preserved state of the two analyzed specimens suggests that the variable shell wall thicknesses observed during different time periods are due to seawater changes.

4. Results 4.3. Stable isotopes 4.1. Planktonic foraminiferal assemblages The oxygen isotope signal is marked by periodic variations superimposed on long term trends (Fig. 6). From the base of the section to ~7.4 Ma, planktonic foraminiferal δ18Ο values average about −0.5‰.

The planktonic foraminiferal fauna is dominated by the Globigerinoides and Globoturborotalita groups (e.g., G. obliquus, 7

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

G. Kontakiotis, et al.

Fig. 5. High resolution μCT tomographs of two randomly picked G. obliquus specimens indicative of shell mass differences, with no signs of dissolution during the studied time interval.

Heavy δ18Ο values commonly reach 1.1‰, whereas minimum lighter values average about −1‰ (minimum −1.9‰). During the latest Tortonian and progressively through the T/M boundary, there is a brief interval (~7.4–7.2 Ma) where mean δ18Ο values are distinctly higher, averaging about 0.7‰ (range between −0.33 and 1.01‰). Across the T/M boundary, δ18Ο values stabilize around 1.0‰, followed by a significant δ18Ο depletion (−1‰) at the beginning of the Messinian (7.21 Ma). During the Messinian, the isotope values become progressively heavier towards the top of the section, with the heaviest δ18Ο values (1.1‰) occurring at 6.83 Ma. However, this upward shift is

interrupted by two pronounced depletions at ~7.1–7.0 Ma (−1.6‰) and 6.82 Ma (−0.6‰) respectively. In the lower part of the succession corresponding to the late Tortonian, δ13C values range between 1.1‰ and 1.7‰. From 7.27 to 7.21 Ma across the T/M boundary, a decrease in δ13C values of ~0.3‰ (1.22 to 0.95‰) is observed. Upwards from this level during the Messinian, δ13C values typically remain generally below 1.1‰ (with only the exceptions of ~7.15, 7.0 and 6.82 Ma) presenting brief intervals of both higher and lower values. After the temporary shift towards heavier values around 7.1–7.0 Ma, the δ13C record shows a trend

Fig. 6. Comparison between the stable isotopic composition (δ18O, δ13C) of the Faneromeni pre-evaporitic section and time-equivalent records in the Mediterranean Sea (Monte del Casino section; Kouwenhoven et al., 1999). The dashed red line represents the Tortonian/Messinian (T/M) boundary. The grey arrow shows the general tendency to lighter δ13C values during the final stage of LMCIS. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) 8

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

20 16 12 8

32

12 8

28

4

24 Modern SST 21°C

0

20

-4

16 6.7

6.9

7.1

7.3 Age (Ma)

7.5

Eastern Mediterranean SST Anomaly (ºC)

28 26 24 22 20 18

Monte Dei Corvi UK'37-SST (°C)

4

T/M boundary

Faneromeni Section Sr/Ca-SST (°C)

ODP 907 North Atlantic UK'37-SST (°C)

G. Kontakiotis, et al.

7.7

Fig. 7. Sr/Ca-derived SSTs on the pre-MSC interval of Faneromeni section along with eastern Mediterranean SST anomaly (Herbert et al., 2016; differences relative to the modern mean annual sea surface temperature at the site section). For comparison, the isochronous Uk′ 37-based SST records from the Mediterranean Sea (Monte del Casino; Tzanova et al., 2015) and Atlantic Ocean (ODP 907; Tzanova, 2015; Herbert et al., 2016) are also displayed. The dashed vertical red line represents the Tortonian/Messinian (T/M) boundary and the grey horizontal one the modern local sea surface temperature value respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

towards lighter values (Fig. 6). After 6.8 Ma, the δ13C signal shifts back to heavier values up to 1.7‰.

interval between the TSC and MSC events, a succession of three restriction phases of variable duration and magnitude is clearly discerned. The initial phase occurred just after the end of the maximum restriction phase of the TSC (7.8–7.6 Ma) around 7.51–7.48 Ma, producing relatively hypersaline waters to 42.32–44.03‰. The second event centered at 7.41–7.39 Ma was equally brief and of the same amplitude (41.72–42.92‰) as the first one. Both these events seem to be separated from the successive restriction phases by short pauses (indicative of fresher conditions) centered at 7.54, 7.45–7.42 and 7.37 Ma respectively. The rebound to higher S values (39.79–43.92‰) and the sustained interval of more saline conditions centered at 7.36–7.24 Ma represent the largest and more intense and persistent restriction phase. Following this salty interval, we observe in the S record a freshening trend of ~5‰ on average, with significant fluctuations represented by minimum (33.82–35.42‰) and maximum (37.39–37.72‰) values. It is also interesting to note the good agreement between the salinity and the G. obliquus shell mass record.

4.4. SST variability over the Late Miocene The calculated SSTs range between 19.8 and 29.9 °C throughout the early and middle part of the record (7.61–6.98 Ma) and document realistic variations in the surface water temperatures (Fig. 7), in accordance with previously published SSTs (Tzanova et al., 2015), and exclude the contribution of diagenesis to the chemistry of the analyzed carbonate shells. However, in the upper part (6.98–6.68 Ma) of the section, early diagenesis alters the primary geochemical composition of the foraminiferal test (Antonarakou et al., 2019), biasing any resulting SST paleoreconstruction. During this time interval, planktonic foraminifera present a “chalky” taphonomy, characterized by the precipitation of authigenic high-Mg calcite and dolomite crystals in the exterior of the tests (advanced diagenetic stage of Antonarakou et al., 2019). The Sr/Ca-derived SSTs present a range of 0.2–13.7 °C, which is unrealistically low compared to time equivalent SSTs within (19.9–27.7 °C; Tzanova et al., 2015) and beyond the Mediterranean Sea (Pacific Ocean: 8.9–12.9 °C; Drury et al., 2018; Herbert et al., 2016; LaRiviere et al., 2012, Arabian Sea: Huang et al., 2007), and therefore they were excluded from the following discussion. The relative standard deviation of our Sr/Ca measurements is ± 2.65%. When combined with the error on calibration Eq. (1), the analytical error associated with the Sr/Ca measurements contributes an additional 0.4 °C of uncertainty. Overall, the error propagation suggests uncertainties of 1.3 °C.

5. Discussion 5.1. Significance of foraminiferal Sr/Ca-SST proxy and potential biases A geochemical deep-sea sediment proxy record of a time span of over millions of years may have been affected by oceanic processes, such as diagenetic overgrowth or dissolution. Although there are no irrefutable methods demonstrating that samples have not been altered, especially in high-salinity marginal basins (Kontakiotis et al., 2011, 2017), we follow the recent findings of Antonarakou et al. (2019) on the preservation of the studied samples. Their intensive microstructural and geochemical characteristics argue against major diagenetic alteration, and therefore the sequence of preservation states with “glassy” or “frosty” shells could be safely used for the paleoenvironmental interpretations. Moreover, Sr/Ca ratios measured on the same samples are in the range of variability of living cultured (Delaney et al., 1985; Lea et al., 1999; Kisakürek et al., 2008) or core-top/plankton tow

4.5. Paleosalinity record through the TMT Raw salinity data range from 33.82 to 44.03‰ (Fig. 8), which is in the expected range of the Mediterranean SSS estimates (Brachert et al., 2007; Tzanova, 2015; Vasiliev et al., 2019) during the pre-evaporitic phase. Overall, the studied interval shows a hypersaline water column with SSSs approaching or even exceeding 40‰. Within this transitional 9

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

G. Kontakiotis, et al.

Fig. 8. Faneromeni section records of computed Ice Volume Free δ18O (δ18OIVF-SW) as an approximation of surface salinities, sea surface salinities (SSSs), and average G. obliquus shell mass during the Tortonian/Messinian Transition. Periods of saltier conditions in the eastern Mediterranean Sea are characterized by highly enriched δ18OIVF-SW values which correlate with higher G. obliquus shell masses (R2 = 0.7). The horizontal dashed line indicates the modern local SSS value. The grey arrows show the general tendency to fresher conditions during the 7.2–6.9 Ma (grey highlighted) time interval, which is further supported by the significant decrease in the planktonic foraminifer shell mass. The dashed vertical red line represents the Tortonian/Messinian (T/M) boundary. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

the tropical-like Uk′ 37 temperatures (~28 °C) for the oldest time window (12.9–8.0 Ma; Tzanova et al., 2015) in the Mediterranean Sea, confirming the transition character from the “warmhouse” (Middle Miocene) towards the “glacial-interglacial” (Pliocene-Pleistocene) climate (Böhme et al., 2008). On the contrary, the average SST of 20.2 °C for the Messinian shows a good correspondence with the present-day average of 20.9 °C, with the slightly lower values during that time to be possibly attributed to the different paleogeography of the study area. According to paleomagnetic data (Krijgsman and Tauxe, 2004), the paleolatitude of Crete at that time was ~275 km more to the north, and this ~2° latitudinal difference could lead to slightly lower SSTs. In this regard, our results reinforce the previous findings of Zidianakis et al. (2007) and Mertz-Kraus et al. (2009a) about the pre-Pliocene development of the modern Mediterranean-type climate. Overall, the long-term trends can be broken down into four intervals based on T variance and relationship between temperature and lithology. The oldest part of the record (7.61–7.50 Ma) exhibits very warm SSTs (up to ~30 °C; Fig. 9a). That climatic phase corresponds to the end of the climate optimum developed during the Tortonian (Zachos et al., 2001). On the contrary, for the rest of the record our sampling clearly demonstrates a discrete cold episode with high amplitude variance at the time of the TMT. The long-term cold event (7.45–6.91 Ma) falls into the lithological alternations of homogeneous and laminated marls at its base and to a highly calcareous bench of clayey limestones at the top, equivalent to the Rossini and Transitional intervals in Monte dei Corvi section (Hüsing et al., 2009). In the second resolution interval between 7.50 and 7.24 Ma, SST averages ~25 °C (Fig. 9b). However, variability increases over this interval from ~1–2 °C

foraminifera (Elderfield et al., 2000; Mortyn et al., 2005) and/or specimens picked from upper Miocene land sections and core sediments (Lear et al., 2003; Sprovieri et al., 2008). This suggests a negligible influence of calcite overgrowths on the carbonate shells and the efficiency of the adopted cleaning procedures to limit the effects of contaminating trace element incorporation. It is also widely accepted that very little shallow-burial diagenetic influence affects the clay mineral assemblage after its storage in sediments (Chamley, 2001). On the contrary, a stronger diagenetic alteration by pore waters is expected to occur in the more porous limestone beds as compared to the marls. Indeed, the “chalky” shells representing the carbonate-rich topmost part of the section (corresponding to 6.98–6.68 Ma), due to their advanced diagenetic potential (Antonarakou et al., 2019), present Sr-depleted values leading to underestimation of the resulting SSTs, and therefore these samples were fully excluded from our interpretations. Sr/Ca ratios in those samples are < 1.2 mmol/mol, reinforcing the initially proposed lower threshold level by Bralower et al. (1997) as unaltered by diagenesis. 5.2. Insights into the Late Miocene climate and comparison to other records When evaluating the paleo-SST in the study area during the TMT, the average eastern Mediterranean temperatures generally exceeded the modern mean annual SST of ~21 °C at the site by as much as 9 °C. More explicitly, during the late Tortonian the average reconstructed SST is 25.2 °C, indicating that the T over this interval at the site was as warm as “warm pool” regions of the modern ocean. Such warmth exceeds modern temperatures at the site, but is significantly cooler than 10

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

G. Kontakiotis, et al.

A

B

Climatic instabilities Alternating periods of enhanced evaporation and reduced runoff

Increased nutrient input from continental sources

Atlantic

Mediterranean

Reduced river

Atlantic

(e.g. rivers)

7.61 - 7.50 Ma

Enhanced productivity

input

Variable SST (21.1 - 28.6ºC) Variable SSS (36.2 - 44.0‰)

Continent

Normal marine conditions

LSL

Gibraltar Arc

Deep water mixing

Colder

HSL

Warmer

Very high SST (up to 30ºC) High SSS (~39‰)

Mediterranean

Gibraltar Arc

Late phase of theTortonian climatic optimum

7.50 - 7.24 Ma

C

Stratification

Continent

Mesotrophic environment

Transitional interval Evidence of late Messinian cooling

D Enhanced evaporation Intermediate

Atlantic

Mediterranean

Increased

Atlantic

runoff

7.24 - 7.18 Ma

Strong SST and SSS decrase

Continent

Cold

Gibraltar Arc

runoff

SST 16 - 21ºC SSS 34 - 38‰

Main phase of the “siphon” event

HSL

LSL

Coldest

SST ~ 21ºC SSS ~ 37‰

Mediterranean

Gibraltar Arc

7.18 - 6.91 Ma

Oligotrophic conditions Sluggish circulation

Continent

Stagnation of bottom waters

Restricted marine conditions Cool and fresh upper water column

Fig. 9. Four-step paleoenvironmental model of the Mediterranean basin during the pre-evaporitic phase of the Messinian Salinity Crisis. A) Late phase of the Tortonian climatic optimum, B) transitional interval with the first evidence of Late Miocene cooling event. C) Coldest times but under normal salinity conditions indicative of the main phase of the “siphon” event. D) Stable cold and fresh upper water column during period of restricted marine conditions. At 7.18 Ma, the deposition of the first sapropels occurs as a result of increased stratification of the entire water column.

cooling around 7.0 Ma was also documented in the Arabian Sea (Huang et al., 2007) and the north Pacific Ocean (LaRiviere et al., 2012), which implies that the Mediterranean SST decrease should be correlated with a large-scale mid-latitude shift in both temperatures and aridity, probably in response to the weakness of Asian and African monsoons (Holbourn et al., 2018). Finally, at the fourth climatic phase (7.18–6.91 Ma; Fig. 9d) we observe stably low SST values around 21 °C indicating the continuation of the cooling. The subsequent shift to predominantly limestone deposition at the top of the section is indicative of the strong arid regional conditions during this climate phase. The termination of this event and the recovery from the cold episode is not present in our record. There is however evidence (Cunningham and Collins, 2002; van Assen et al., 2006) that it coincides with the onset of the diatomite deposition at 6.72 Ma (Krijgsman et al., 1999; Roger et al., 2000; Münch et al., 2003; Drinia et al., 2004, 2007).

oscillations at the previous phase to 4–5 °C towards the T/M boundary. A three-step cold episode evidenced by three notable cold excursions of ~6 °C, 5 °C and 3 °C positioned at 7.45, 7.37 and 7.28 Ma respectively marks this period, representing the first evidence of the Messinian cooling in the eastern Mediterranean. This finding is confirmed in other Mediterranean sections indicating that the first signs of the upper Miocene cooling start in the latest Tortonian around 7.4 Ma (Hodell et al., 1994; Tzanova et al., 2015). In between these peak minimum values, the average SST rebounds to ~25.5 °C, which is even lower than the average SST recorded in middle-late Tortonian (Tzanova et al., 2015; Herbert et al., 2016; Holbourn et al., 2018). Cooling inferred from the Faneromeni section matches the coeval T decrease in both the Italian and Iberian peninsulas (Jimenez-Moreno et al., 2010; Tzanova et al., 2015), and further correlates to the major shift in the European and north African landscape to cooler and dryer conditions (Pound et al., 2011, 2012). The most remarkable feature in the third climatic phase is the strong SST decrease up to 11 °C during the 7.24–7.18 Ma time interval (Fig. 9c). This significant cooling represents the main phase of the known “siphon” event (7.20–6.58 Ma), which has been attributed to the influx of colder Atlantic waters through the Rifian (Benson et al., 1991; Achalhi et al., 2016; Capella et al., 2018) and/or Gibraltar Corridors (Krijgsman et al., 2018) into the Mediterranean basin. The timing of the minimum SSTs around the T/M boundary is in good agreement with the appearance of the first Sahara dunes (Schuster et al., 2006) at ~7.0 Ma, as a response to aridification and cooling of the circum-Mediterranean region. However, an isochronous similar

5.3. The siphon event: strong SST-SSS decrease or seasonality-deduced signal? A Mediterranean-type climate with pronounced seasonality in precipitation characterized by heavy rain events during the winter months and increasing summer aridity has already been documented for the Late Miocene (Mertz-Kraus et al., 2009a, 2009b), since the atmospheric pressure fields during that time are considered to be similar to today (Brachert et al., 2006). Enhanced summer evaporation started and intensified in the course of the early Messinian, when winters became less humid in comparison to the late Tortonian. The potential seasonality 11

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

G. Kontakiotis, et al.

5.4. The correlation of LMCIS with hydrographic and climate changes

signal could be expressed by both SST and seasonal changes in ambient seawater composition (correlated with SSS) The SST seasonality based on coral Sr/Ca proxy, ranges between 7.1 °C and 11.5 °C (Mertz-Kraus et al., 2009a, 2009b), and agrees with both the observed SST decrease (5.0–9.9 °C) in the Faneromeni section and the relevant SST seasonality for the modern Aegean Sea (9 °C; Poulos et al., 1997). However, although the Late Miocene seasonal salinity change (1.1‰) deduced from the same coral data is within the range of the seasonal salinity changes of present-day Aegean surface waters (0.5–3‰; Poulos et al., 1997), it cannot explain the strong SSS decline (up to 10‰) documented in our record. Therefore, although the seasonal (mixed signal of SST and SSS) component on the siphon event development through the TMT cannot be omitted, at least during its initial stages where seasonality seems to be stronger (Di Stefano et al., 2010), an alternative explanation should be given. The main paroxismal phase of the siphon event occurred at 7.2–7.0 Ma and corresponds to the T/M boundary, defined by the first appearance of G. conomiozea in the Mediterranean (Sierro, 1985; Sierro et al., 1993). This has been interpreted as an influx of colder and less saline deep Atlantic waters into the basin (Hodell et al., 1989; Krijgsman et al., 1997). The sea surface cooling during this interval is also reinforced by the significant presence of N. acostaensis (s) upwards from this level. This species is usually associated with cold, nutrientrich waters and the development of a strong deep chlorophyll maximum (DCM) layer (Sierro et al., 2003). The observations of Ivanovic et al. (2013) about a stronger than present early Messinian Atlantic inflow lead us to speculate on a greater SST and SSS decrease in the upper water column during that time. In this case, the combined continental/fluvial seasonal inputs and Atlantic inflows could have resulted in such strong SST and SSS decreases as recorded in the Faneromeni section (up to 10 units). The occurrence of decreasing mean annual salinity (33–38‰; Fig. 8), together with minor SSS seasonal amplitudes during the beginning of the Messinian (2.3–2.8‰), compared to the latest Tortonian (4.2‰), is also supported by δ18O coral data that imply short events of enhanced salinity followed by the return to normal marine conditions (Brachert et al., 2007) suitable for coral reef growth from the Tortonian to the lower Messinian. This argument is further strengthened by the synchronous reduction observed in the average shell mass of G. obliquus (Fig. 8), which can be explained by the introduction of fresher and of lower density Atlantic waters into the basin (Zarkogiannis et al., 2019). The observed SSS difference (~34–44‰) seems to be realistic, since a) a difference of only 6‰ corresponds to that of SSS values of the modern northern Red Sea (~41‰) compared to open ocean waters (~35‰), and b) the oceanographic setting of the current Red Sea and the late Miocene eastern Mediterranean Sea are comparable (MertzKraus et al., 2009b). Such S estimates also explain the nature of the early Messinian environments documented in the eastern Mediterranean (ongoing development of environmental stress; Kouwenhoven et al., 2003; Moissette et al., 2018) through the relatively abundant and diverse benthic foraminiferal fauna. Within the planktonic foraminiferal assemblages, the co-occurrence of the high-S tolerant T. quinqueloba (Bijma et al., 1990) and G. glutinata can also be considered indicative of stressed conditions possibly due to high salinity (Antonarakou et al., 2007; Drinia et al., 2007; Sierro et al., 2003). Overall, we highlight that salinity was highly variable during the TMT, and most likely the restricted nature of the eastern Mediterranean basin intensified the precipitation, evaporation and therefore salinity distribution patterns. This environmental configuration is opposed to the monotonous upward salination trend towards the MSC, since it was often interrupted by brief salinity reversals indicative of the transitional character of this setting. This finding reinforces the data of Kouwenhoven et al. (2003), showing that the increasing stress in terms of salinity variations was more pronounced in the deepest sections (e.g., Monte del Casino, Monte Gibliscemi, and Metochia).

The shift of the carbon isotopes towards lighter values (~1.6‰) is comparable to those at other Mediterranean sites (Monte del Casino, Gibliscemi, and Metochia sections; Sprovieri et al., 1999; Kouwenhoven et al., 1999, 2003). This negative shift points to a perturbation in the global carbon budget and has been considered as a widespread and globally synchronous chemostratigraphic event (Drury et al., 2017, 2018). The most acceptable forcing mechanism that could shift δ13C values by the observed magnitude (~1‰) would include a global shift in the δ13C of oceanic dissolved inorganic carbon (δ13CDIC; Hodell et al., 1994, 2001; Bickert et al., 2004; Hodell and Venz-Curtis, 2006), rather than changes in the marine productivity (Diester-Haass et al., 2006; Drury et al., 2016). Therefore, the δ13C has been placed into the context of increased continental organic carbon flux to the oceans (Delaney and Boyle, 1987) during eustatic sea-level lowering resulting from Antarctic glaciation (Vincent et al., 1980; Ohneiser et al., 2015), or a change in the fractionation of organic matter in the surface ocean (Δorg; Bickert et al., 2004), and/or changes in the vegetation type (e.g., late Miocene global C3:C4 vegetation shift; Pagani et al., 1999) related to increasing seasonality and aridity (Molnar, 2005). Because C4 type plants represent an adaptation to water stress, the expansion of grassland habitats and C4 vegetation between 7.6 and 6.7 Ma has been related to increased aridity (Hodell et al., 1994; Molnar, 2005). Increased aridity in the Mediterranean region may have contributed to the onset of a strong negative water budget into the basin near the T/ M boundary (Hodell et al., 1989; Benson et al., 1991). The trend to heavier δ18O values around 7.2 Ma was interpreted as reflecting a reversal in deep water flow through the Gibraltar Corridor, whereby warm waters of Mediterranean origin were replaced by cold intermediate Atlantic waters. The significant δ18O increase during that time does not reflect a local cooling event but probably represents an increase in global ice volume that lowered sea-level and contributed to the establishment of a negative water budget in the Mediterranean Sea. This is further supported by both ice sheet expansion on Antarctica (Hodell et al., 1986; Kennett and Barker, 1990; Lear et al., 2015), and upper Miocene regressive phases documented in sedimentary sequences throughout the world (Adams et al., 1977). The reversal in deep water circulation at this time resulted in the inflow of nutrient-rich Atlantic intermediate waters into the Mediterranean, which enhanced primary productivity and finally resulted in the organic-rich sapropel development (~7.17 Ma; Kouwenhoven et al., 1999). Constriction of the Atlantic-Mediterranean connections between 7.2 and 7.1 Ma (Krijgsman et al., 1999) could obstruct outflow of deeper waters (Kouwenhoven and van der Zwaan, 2006), slow down the circulation and further contribute to increasing residence time of water masses and accumulation of light organic carbon. Highlighted here is the shift towards much lighter δ13C values within the eastern Mediterranean than at the Atlantic side of the Rifian Corridor (Hodell et al., 1994). This discrepancy between intra- and extra-Mediterranean carbon isotope records is consistent with previous findings of Kouwenhoven et al. (1999) and van der Zwaan and Gudjonsson (1986), and could be considered indicative of the progressive isolation of the basin from 7.2 Ma. Especially the lightest values recorded at the 6.93–6.75 Ma time interval are equivalent to the extra shift occurring between 7.0 and 6.8 Ma in Monte del Casino, which was attributed by Kouwenhoven et al. (2003) to the continuing restriction of the Mediterranean Sea. 6. Conclusions The integrated study here is based on planktonic foraminiferal faunal and geochemical (δ18O, δ13C, Sr/Ca) data from the pre-evaporitic interval of the Faneromeni section, which provides an excellent illustration of the progressive restriction of the Mediterranean basin during the TMT (7.6–6.7 Ma) in response to climate changes. The proposed paleoenvironmental evolution represents a succession of four12

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

G. Kontakiotis, et al.

step paleoceanographic/paleoclimatic phases, which are biostratigraphically framed, and broadly consistent with those reported in other isochronous sections within the entire Mediterranean Sea. During the first phase (7.61–7.50 Ma), the open Atlantic-Mediterranean exchanges maintained relatively stable marine conditions, expressed by a very warm and salty upper water column. The second phase developed until the T/M boundary (7.50–7.24 Ma) and marks the first evidence of basin restriction with a wide range of temperature and salinity fluctuations. A climatic trend towards increased aridity and continentality during the earliest Messinian (third phase: development of the restriction), as reflected by the LMCIS, may have also predisposed the eastern Mediterranean to desiccation. The increase in δ18O values around 7.2 Ma and onward marks a significant SST (up to 10 °C) decrease that possibly lowered sea-level and finally resulted in the establishment of a negative water budget in the Mediterranean basin. This change is also recorded by both the faunal and geochemical proxies, which document an important cooling and desalination trend in the upper water column just after the T/M boundary, attributed to the paroxysmal phase of the so-called “siphon” event. The 7.2–7.1 Ma time interval could be suggested as an important step to the restriction, and may have resulted from shallowing of the Mediterranean gateways under a tectonic control, although a glacio-eustatic overprint cannot be completely excluded. During the fourth phase (7.18–6.91 Ma), the cooling continues together with a decrease of SSS, that resulted in stressful conditions for the marine microfauna, marked by a notable decrease in the diversity of foraminiferal assemblages. Overall, this study reveals a non-gradual increase in sea surface salinity prior to the onset of the MSC, but substantial variability in response to climatic oscillations, supporting the concept of a stepwise restriction of the Mediterranean Sea.

Antonarakou, A., Drinia, H., Tsaparas, N., Dermitzakis, M.D., 2007. Micropaleontological parameters as proxies of Late Miocene surface water properties and paleoclimate in Gavdos Island, Eastern Mediterranean. Geodiversitas 29, 379–399. Antonarakou, A., Kontakiotis, G., Mortyn, P.G., Drinia, H., Sprovieri, M., Besiou, E., Tripsanas, E., 2015. Biotic and geochemical (δ18Ο, δ13C, Mg/Ca, Ba/Ca) responses of Globigerinoides ruber morphotypes to upper water column variations during the last deglaciation, Gulf of Mexico. Geochim. Cosmochim. Acta 170, 69–93. https://doi. org/10.1016/j.gca.2015.08.003. Antonarakou, A., Kontakiotis, G., Zarkogiannis, S., Mortyn, P.G., Drinia, H., Koskeridou, E., Anastasakis, G., 2018. Planktonic foraminiferal abnormalities in coastal and open marine eastern Mediterranean environments: a natural stress monitoring approach in recent and early Holocene marine systems. J. Mar. Syst. 181, 63–78. https://doi.org/ 10.1016/j.jmarsys.2018.02.004. Antonarakou, A., Kontakiotis, G., Vasilatos, C., Besiou, E., Zarkogiannis, S., Drinia, H., Mortyn, P.G., Tsaparas, N., Makri, P., Karakitsios, V., 2019. Evaluating the effect of marine diagenesis on Late Miocene pre-evaporitic sedimentary successions of eastern Mediterranean Sea. IOP Conference Series: Earth and Environmental Sciences 221, 012051. https://doi.org/10.1088/1755-1315/221/1/012051. Barker, S., Greaves, M., Elderfield, H., 2003. A study of cleaning procedures used for foraminiferal Mg/Ca paleothermometry. Geochem. Geophys. Geosyst. 4 (9), 8407. https://doi.org/10.1029/2003GC000559. Beltran, C., Flores, J.A., Sicre, M.A., Baudin, F., Renard, M., de Rafelis, M., 2011. Long chain alkenones in the Early Pliocene Sicilian sediments (Trubi Formation - Punta di Maiata section): implications for the alkenone paleothermometry. Paleogeogr. Paleoclimatol. Paleoecol. 308, 253–263. Bemis, B.E., Spero, H.J., Bijma, J., Lea, D.W., 1998. Reevaluation of the oxygen isotopic composition of planktonic foraminifera: experimental results and revised paleotemperature equations. Paleoceanography 13 (2), 150–160. https://doi.org/10. 1029/98PA00070. Benson, R.H., Rakic-El Bied, K., Bonaduce, G., 1991. An important current reversal (influx) in the Rifian Corridor (Morocco) at the Tortonian-Messinian boundary: the end of the Tethys Ocean. Paleoceanography 6, 164–192. Bickert, T., Haug, G.H., Tiedemann, R., 2004. Late Neogene benthic stable isotope record of Ocean Drilling Program Site 999: implications for Caribbean paleoceanography, organic carbon burial, and the Messinian Salinity Crisis. Paleoceanography 19 (1). https://doi.org/10.1029/2002PA000799. Bijma, J., Faber Jr., W.W., Hemleben, Ch., 1990. Temperature and salinity limits for growth and survival of some planktonic foraminifers in laboratory cultures. J. Foram. Res. 20, 95–116. Böhme, M., August, I., Winklhofer, M., 2008. Late Miocene "washhouse" climate in Europe. Earth Planet. Sci. Lett. 275, 393–401. Brachert, T.C., Reuter, M., Felis, T., Kroeger, K.F., Lohmann, G., Micheels, A., Fassoulas, C., 2006. Porites Corals from Crete (Greece) open a window into late Miocene (10Ma) seasonal and interannual climate variability. Earth Planet. Sci. Lett. 245, 81–94. Brachert, T.C., Vescogni, A., Bosellini, F.R., Reuter, M., Mertz-Kraus, R., 2007. High salinity variability during the early Messinian revealed by stable isotope signatures from vermetid and Halimeda reefs of the Mediterranean region. Geol. Romana 40, 1–16. Brachert, T., Bornemann, A., Reuter, M., J. Galer, S., Grimm, K., Fassoulas, C., 2015. Upwelling history of the Mediterranean Sea revealed by stunted growth in the planktic foraminifera Orbulina universa (early Messinian, Crete, Greece). Int. J. Earth Sci. (Geol Rundsch), 104, 263–-276. DOI https://doi.org/10.1007/s00531-0141076-8. Bralower, T.J., Thomas, D.J., Zachos, J.C., Hirschmann, M.M., Röhl, U., Sigurdsson, H., Thomas, E., Whitney, D.L., 1997. High-resolution records of the late Paleocene thermal maximum and circum-Caribbean volcanism: is there a causal link? Geology 25 (11), 963–966. Capella, W., Barhoun, N., Flecker, R., Hilgen, F.J., Kouwenhoven, T.E., Matenco, L., Sierro, F.J., Tulbure, M., Yousfi, Z., Krijgsman, W., 2018. Palaeogeographic evolution of the Late Miocene Rifian Corridor (Morocco): reconstructions from surface and subsurface data. Earth-Sci. Rev. 180, 37–59. Chamley, H., 2001. Clay mineralogy. Encyclopedia of ocean. Sciences 563–571. https:// doi.org/10.1016/b978-012374473-9.00186-7. Cleaveland, L.C., Herbert, T.D., 2009. Preservation of the alkenone paleotemperature proxy in uplifted marine sequences: a test from the Vrica outcrop, Crotone, Italy. Geology 37, 179–182. Cunningham, K.J., Collins, L.S., 2002. Controls on facies and sequence stratigraphy of an upper Miocene carbonate ramp and platform, Melilla basin, NE Morocco. Sed. Geol. 146, 285–304. De Bar, M., Dorhout, D.J.C., Hopmans, E.C., Sinninghe Damsté, J.S., Schouten, S., 2016. Constraints on the application of long chain diol proxies in the Iberian Atlantic margin. Org. Geochem. 101, 184–195. https://doi.org/10.1016/j.orggeochem.2016. 09.005. Delaney, M.L., Boyle, E.A., 1987. Cd/Ca in late Miocene benthic foraminifera and changes in the global organic carbon budget. Nature 330, 156–159. Delaney, M.L., Bé, A.W.H., Boyle, E.A., 1985. Li, Sr, Mg, and Na in foraminiferal calcite shells from laboratory culture, sediment traps, and sediment cores. Geochim. Cosmochim. Acta 49, 1327–1341. Di Stefano, A., Verducci, M., Lirer, F., Ferraro, L., Iaccarino, S.M., Hüsing, S.K., Hilgen, F.J., 2010. Paleoenvironmental conditions preceding the Messinian Salinity Crisis in the Central Mediterranean: integrated data from the Upper Miocene Trave section (Italy). Palaeogeogr. Palaeoclimatol. Palaeoecol. 297, 37–53. Diester-Haass, L., Billups, K., Emeis, K.C., 2006. Late Miocene carbon isotope records and marine biological productivity: was there a (dusty) link? Paleoceanography 21, PA4216. https://doi.org/10.1029/2006PA001267. Drinia, H., Antonarakou, A., Tsaparas, N., Dermitzakis, M.D., Kontakiotis, G., 2004.

Acknowledgments The authors are grateful to Ignacio Villarroya for kind assistance during the trace element analyses. Jan Willem Zachariasse and Lucas J. Lourens are warmly thanked for useful discussions that helped to improve the manuscript. Frédéric Quillévéré and an anonymous reviewer are deeply appreciated for their critical revision that helped to improve the manuscript, and Thierry Corrège (Editor) is thanked for his editorial handling. This research has been co-financed by the European Union (European Social Fund–ESF) and Greek national funds through the Operational Program “Education and Lifelong Learning” of the National Strategic Reference Framework (NSRF) -Research Funding Program: THALIS–UOA-“Messinian Salinity Crisis: the greatest Mediterranean environmental perturbation and its repercussions to the biota” (MEDSALC) MIS: 375405. Collaboration was also possible through the COST Action CA15103 “Uncovering the Mediterranean salt giant” (MEDSALT) supported by COST (European Cooperation in Science and Technology). The data associated with this article can be found in Supplementary information. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.palaeo.2019.109312. References Achalhi, M., Münch, P., Cornée, J.J., Azdimousa, A., Melinte-Dobrinescu, M., Quillévéré, F., Drinia, H., Fauquette, S., Jiménez-Moreno, G., Merzeraud, G., Moussa, A.B., 2016. The late Miocene Mediterranean-Atlantic connections through the North Rifian Corridor: new insights from the Boudinar and Arbaa Taourirt basins (northeastern Rif, Morocco). Palaeogeogr. Palaeoclimatol. Palaeoecol. 459, 131–152. Adams, C.G., Benson, R.H., Kidd, R.B., Ryan, W.B.F., Wright, R.C., 1977. The Messinian salinity crisis and evidence of late Miocene eustatic changes in the world ocean. Nature 269, 383–386. Agiadi, K., Antonarakou, A., Kontakiotis, G., Kafousia, N., Moissette, P., Cornée, J.J., Manoutsoglou, E., Karakitsios, V., 2017. Connectivity controls on the late Miocene eastern Mediterranean fish fauna. Int. J. Earth Sci. 106, 1147–1159.

13

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

G. Kontakiotis, et al.

hydrological change drove the late Miocene C-4 plant expansion in the Himalayan foreland and Arabian Peninsula. Geology 35, 531–534. Huguet, C., Kim, J.-H., Sinninghe Damsté, J.S., Schouten, S., 2006. Reconstruction of sea surface temperature variations in the Arabian Sea over the last 23 kyr using organic proxies (TEX86 and Uk′ 37). Paleaoceanography 21, PA3003. https://doi.org/10.1029/ 2005PA001215. Huguet, C., Schimmelmann, A., Thunell, R., Lourens, L.J., Sinninghe Damsté, J.S., Schouten, S., 2007. A study of the TEX86 paleothermometer in the water column and sediments of the Santa Barbara Basin, California. Paleoceanography 22, PA3203. https://doi.org/10.1029/2006PA001310. Hüsing, S.K., Kuiper, K.F., Link, W., Hilgen, F.J., Krijgsman, W., 2009. The upper Tortonian-lower Messinian at Monte dei Corvi (Northern Apennines, Italy): completing a Mediterranean reference section for the Tortonian Stage. Earth Planet. Sci. Lett. 282, 140–157. Ivanovic, R.F., Flecker, R., Gutjahr, M., Valdes, P.J., 2013. First Nd isotope record of Mediterranean–Atlantic water exchange through the Moroccan Rifian Corridor during the Messinian Salinity Crisis. Earth Planet. Sci. Lett. 368, 163–174. Jimenez-Moreno, G., Fauquette, S., Suc, J.P., 2010. Miocene to Pliocene vegetation reconstruction and climate estimates in the Iberian Peninsula from pollen data. Rev. Palaeobot. Palynol. 162, 403–415. Johnstone, H.J.H., Schulz, M., Barker, S., Elderfield, H., 2010. Inside story: an X-ray computed tomography method for assessing dissolution in the tests of planktonic foraminifera. Mar. Micropal. 77 (1), 58–70. Jonas, A.S., Schwark, L., Bauersachs, T., 2017. Late Quaternary water temperature vark′ iations of the Northwest Pacific based on lipid paleothermometers TEXH 86, U37 and LDI. Deep Sea Research, Part I 125, 81–93. https://doi.org/10.1016/j.dsr.2017.04. 018. Karakitsios, V., Roveri, M., Lugli, S., Manzi, V., Gennari, G., Antonarakou, A., Triantaphyllou, M., Agiadi, K., Kontakiotis, G., Kafousia, N., de Rafelis, M., 2017a. A record of the Messinian salinity crisis in the eastern Ionian tectonically active domain (Greece, eastern Mediterranean). Bas. Res. 29, 203–233. Karakitsios, V., Cornée, J.-J., Tsourou, T., Moissette, P., Kontakiotis, G., Agiadi, K., Manoutsoglou, E., Triantaphyllou, M., Koskeridou, E., Drinia, H., Roussos, D., 2017b. Messinian salinity crisis record under strong freshwater input in marginal, intermediate, and deep environments: the case of the North Aegean. Palaeogeogr. Palaeoclimatol. Palaeoecol. 485, 316–335. https://doi.org/10.1016/j.palaeo.2017. 06.023. Kennett, J.P., Barker, P.F., 1990. Latest Cretaceous to Cenozoic climate and oceanographic development in the Weddell Sea, Antarctica: an ocean-drilling perspective. Proc. Ocean Drill. Prog. Sci. Res. 113, 937–960. Kim, J.-H., Schouten, S., Rodrigo-Gámiz, M., Rampen, S., Marino, G., Huguet, C., Helmke, P., Buscail, R., Hopmans, E.C., Pross, J., Sangiorgi, F., Middelburg, J.B.M., Sinninghe Damsté, J.S., 2015. Influence of deep-water derived isoprenoid tetraether lipids on the paleothermometer in the Mediterranean Sea. Geochem. Cosmochim. Acta 150, 125–141. https://doi.org/10.1016/j.gca.2014.11.017. Kisakürek, B., Eisenhauer, A., Böhm, F., Garbe-Schönberg, D., Erez, J., 2008. Controls on shell Mg/Ca and Sr/Ca in cultured planktonic foraminifera Globigerinoides ruber (white). Earth Planet. Sci. Lett. 273 (3–4), 260–269. Kontakiotis, G., 2016. Late Quaternary paleoenvironmental reconstruction and paleoclimatic implications of the Aegean Sea (eastern Mediterranean) based on paleoceanographic indexes and stable isotopes. Quater. Internat. 401, 28–42. https://doi.org/ 10.1016/j.quaint.2015.07.039. Kontakiotis, G., Mortyn, P.G., Antonarakou, A., Martínez-Botí, M.À., Triantaphyllou, M. V., 2011. Field-based validation of a diagenetic effect on G. ruber Mg/Ca paleothermometry: core top results from the Aegean Sea (eastern Mediterranean). Geochem. Geophys. Geosyst., 12(9), Q09004. doi:https://doi.org/10.1029/ 2011GC003692. Kontakiotis, G., Antonarakou, A., Zachariasse, W.J., 2013. Late Quaternary palaeoenvironmental changes in the Aegean Sea: interrelations and interactions between north and south Aegean Sea. Bull.Geol. Soc. Greece 47, 167–177. Kontakiotis, G., Karakitsios, V., Mortyn, P.G., Antonarakou, A., Drinia, H., Anastasakis, G., Agiadi, K., Kafousia, N., De Rafelis, M., 2016a. New insights into the early Pliocene hydrographic dynamics and their relationship to the climatic evolution of the Mediterranean Sea. Palaeogeogr. Palaeoclimatol. Palaeoecol. 459, 348–364. Kontakiotis, G., Mortyn, P.G., Antonarakou, A., Drinia, H., 2016b. Assessing the reliability of foraminiferal Mg/Ca thermometry by comparing field-samples and culture experiments: a review. Geol. Quart. 60 (3), 547–560. https://doi.org/10.7306/gq.1272. Kontakiotis, G., Antonarakou, A., Mortyn, P.G., Drinia, H., Anastasakis, G., Zarkogiannis, S., Möbius, J., 2017. Morphological recognition of Globigerinoides ruber morphotypes and their susceptibility to diagenetic alteration in the eastern Mediterranean Sea. J. Mar. Syst. 174, 12–24. https://doi.org/10.1016/j.jmarsys.2017.05.005. Koskeridou, E., Giamali, C., Antonarakou, A., Kontakiotis, G., Karakitsios, V., 2017. Early Pliocene gastropod assemblages from the eastern Mediterranean (SW Peloponnese, Greece) and their palaeobiogeographic implications. Geobios 50, 267–277. Kouwenhoven, T.J., van der Zwaan, G.J., 2006. A reconstruction of late Miocene Mediterranean circulation patterns using benthic foraminifera. Palaeogeogr. Palaeoclimatol. Palaeoecol. 238, 373–385. Kouwenhoven, T.J., Seidenkrantz, M.S., Van Der Zwaan, G.J., 1999. Deep-water changes: the near-synchronous disappearance of a group of benthic foraminifera from the late Miocene Mediterranean. Paleogeogr. Paleoclimatol. Paleoecol. 152, 259–281. Kouwenhoven, T.J., Hilgen, F.J., Van Der Zwaan, G.J., 2003. Late Tortonian-early Messinian stepwise disruption of the Mediterranean-Atlantic connections: constraints from benthic foraminiferal and geochemical data. Palaeogeogr. Palaeoclimatol. Palaeoecol. 198, 303–319. Krijgsman, W., Tauxe, L., 2004. Shallow bias in Mediterranean paleomagnetic directions caused by inclination error. Earth Planet. Sci. Lett. 222, 685–695.

Foraminiferal record of environmental changes: pre-evaporitic diatomaceous sediments from Gavdos Island. southern Greece. Bull. Geol. Soc. Greece 36, 782–791. Drinia, H., Antonarakou, A., Tsaparas, N., Kontakiotis, G., 2007. Palaeoenvironmental conditions preceding the Messinian Salinity Crisis: a case study from Gavdos Island. Geobios 40, 251–265. Drinia, H., Antonarakou, A., Kontakiotis, G., 2008. On the occurrence of Early Pliocene marine deposits in Ierapetra basin, E. Crete. Bull. Geosci. 83, 63–78. Drinia, H., Antonarakou, A., Tsourou, T., Kontakiotis, G., Psychogiou, M., Anastasakis, G., 2016. Foraminifera eco-biostratigraphy of the southern Evoikos outer shelf, central Aegean Sea, during MIS 5 to present. Cont. Shelf Res. 126, 36–49. Drury, A.J., John, C.M., Shevenell, A.E., 2016. Evaluating climatic response to exter-nal radiative forcing during the late Miocene to early Pliocene: new perspec-tives from eastern equatorial Pacific (IODP U1338) and North Atlantic (ODP 982) locations. Paleoceanography 31 (1), 167–184. https://doi.org/10.1002/2015PA002881. Drury, A.J., Westerhold, T., Frederichs, T., Tian, J., Wilkens, R., Channell, J.E.T., Evans, H., John, C.M., Lyle, M., Röhl, U., 2017. Late Miocene climate and time scale reconciliation: accurate orbital calibration from a deep-sea perspective. Earth Planet. Sci. Lett., 475(C), 254–266, doi:https://doi.org/10.1016/j.epsl.2017.07.038. Drury, A.J., Lee, G.P., Gray, W.R., Lyle, M., Westerhold, T., Shevenell, A.E., John, C.M., 2018. Deciphering the state of the late Miocene to early Pliocene equatorial Pacific. Paleocenagraphy 33 (3), 246–263. Elderfield, H., Cooper, M., Ganssen, G., 2000. Sr/Ca in multiple species of planktonic foraminifera: implications for reconstructions of seawater Sr/Ca. Geochem. Geophys. Geosyst., 1. In: 1999GC00031. Elderfield, H., Vautravers, M., Cooper, M., 2002. The relationship between shell size and Mg/Ca, Sr/Ca, δ18O, and δ13C of species of planktonic foraminifera. Geochem. Geophys. Geosyst. 3 (8). https://doi.org/10.1029/2001GC000194. Flecker, R., Krijgsman, W., Capella, W., de Castro Martíns, C., Dmitrieva, E., Mayser, J.P., Marzocchi, A., Modestu, S., Ochoa, D., Simon, D., Tulbure, M., van den Berg, B., van der Schee, M., de Lange, G., Ellam, R., Govers, R., Gutjahr, M., Hilgen, F., Kouwenhoven, T., Lofi, J., Meijer, P., Sierro, F.J., Bachiri, N., Barhoun, N., Alami, A.C., Chacon, B., Flores, J.A., Gregory, J., Howard, J., Lunt, D., Ochoa, M., Pancost, R., Vincent, S., Yousfi, M.Z., 2015. Evolution of the Late Miocene Mediterranean–Atlantic gateways and their impact on regional and global environmental change. Earth-Sci. Rev. 150, 365–392. Gennari, R., Lozar, F., Turco, E., Dela, Pierre F., Lugli, S., Manzi, V., Natalicchio, M., Roveri, M., Schreiber, B.C., Taviani, M., 2018. Integrated stratigraphy and paleoceanographic evolution of the pre-evaporitic phase of the Messinian salinity crisis in the Eastern Mediterranean as recorded in the Tokhni section (Cyprus Island). Newslett. Stratigr. 51 (1), 33–55. Giamali, C., Koskeridou, E., Antonarakou, A., Kontakiotis, G., Ioakim, Chr, Karageorgis, A., Roussakis, G., Karakitsios, V., 2019. Multiproxy marine ecosystem response of abrupt Holocene climatic changes in the northeastern Mediterranean sedimentary archive. Quat. Res. 1–21. https://doi.org/10.1017/qua.2019.38. Gong, C., Hollander, D.J., 1999. Evidence for differential degradation of alkenones under contrasting bottom water oxygen conditions: implication for paleotemperature reconstruction. Geochim. Cosmochim. Acta 63 (3–4), 405–411. https://doi.org/10. 1016/S0016-7037(98)00283-X. Hemleben, C., Spindler, M., Anderson, O.R., 1989. Modern Planktic Foraminifera. Springer, New York, pp. 1–363. Herbert, T.D., Lawrence, K.T., Tzanova, A., Peterson, L.C., Caballero-Gill, R., Kelly, C.S., 2016. Late Miocene global cooling and the rise of modern ecosystems. Nat. Geosci. https://doi.org/10.1038/NGEO2813. Hilgen, F.J., Krijgsman, W., 1999. Cyclostratigraphy and astrochronology of the Tripoli diatomite Formation (pre-evaporite Messinian, Sicily, Italy). Terra Nova 11, 16–22. Hilgen, F.J., Krijgsman, W., Langereis, C.G., Lourens, L.J., Santarelli, A., Zachariasse, W.J., 1995. Extending the astronomical (polarity) time scale into the Miocene. Earth Planet. Sci. Lett. 136, 495–510. Hilgen, F.J., Bissoli, L., Iaccarino, S., Krijgsman, W., Meijer, R., Negri, A., Villa, G., 2000. Integrated stratigraphy and astrochronology of the Messinian GSSP at Oued Akrech (Atlantic Morocco). Earth Planet. Sci. Lett. 182 (3–4), 237–251. Hodell, D.A., Venz-Curtis, K.A., 2006. Late Neogene history of deep water venti-lation in the Southern Ocean. Geochem. Geophys. Geosyst. 7 (9), Q09001. https://doi.org/10. 1029/2005GC001211. Hodell, D.A., Elmstrom, K.M., Kennett, J.P., 1986. Latest Miocene benthic δ18Ο changes, global ice volume, sea level, and the Messinian salinity crisis. Nature 320, 411–414. Hodell, D.A., Benson, R.H., Kennett, J.P., Rakic-El Bied, K., 1989. Stable isotope stratigraphy of latest Miocene sequences in northwest Morocco: the Bou Regreg section. Paleoceanography 4, 467–482. Hodell, D.A., Benson, R.H., Kent, D.V., Boersma, A., Rakic-El-Bied, K., 1994. Magnetostratigraphic, biostratigraphic, and stable isotope stratigraphy of an Upper Miocene drill core from the Salé Briqueterie (northwest Morocco): a high-resolution chronology for the Messinian stage. Paleoceanography 9, 835–855. Hodell, D., Curtis, J., Sierro, F.J., Raymo, M., 2001. Correlation of Late Miocene-to-Early Pliocene sequences between the Mediterranean and North Atlantic. Paleoceanography 16, 164–179. Holbourn, A., Kuhnt, W., Clemens, S.C., Kochhann, K.G.D., Jöhnck, J., Lübbers, J., Andersen, N., 2018. Late Miocene climate cooling and intensification of southeast Asian winter monsoon. Nat. Commun. 9 (1). https://doi.org/10.1038/s41467-01803950-1. Hori, M., Shirai, K., Kimoto, K., Kurasawa, A., Takagi, H., Ishida, A., Takahata, N., Sano, Y., 2018. Chamber formation and trace element distribution in the calcite walls of laboratory cultured planktonic foraminifera (Globigerina bulloides and Globigerinoides ruber). Mar. Micropaleontol. 140, 46–55. https://doi.org/10.1016/j.marmicro.2017. 12.004. Huang, Y.S., Clemens, S.C., Liu, W.G., Wang, Y., Prell, W.L., 2007. Large-scale

14

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

G. Kontakiotis, et al.

Miller, K.G., Mountain, G.S., Wright, J.D., Browning, J.V., 2011. A 180-million-year record of sea level and ice volume variations from continental margin and deep-sea isotopic records. Oceanography 24, 40–53. Moissette, P., Cornée, J.-J., Antonarakou, A., Kontakiotis, G., Drinia, H., Koskeridou, E., Tsourou, T., Agiadi, K., Karakitsios, V., 2018. Palaeoenvironmental changes at the Tortonian/Messinian boundary: a deep-sea sedimentary record of the eastern Mediterranean Sea. Palaeogeogr. Palaeoclimatol. Palaeoecol. 505, 217–233. Molnar, P., 2005. Mio-Pliocene growth of the Tibetian Plateau and evolution of east Asian climate. Palaeontol. Electron. 8 (1). Mortyn, P.G., Elderfield, H., Anand, P., Greaves, M., 2005. An evaluation of controls on planktonic foraminiferal Sr/Ca: comparison of water column and core-top data from a North Atlantic transect. Geochem. Geophsys. Geosyst. 6. https://doi.org/10.1029/ 2005GC001047. Müller, P.J., Kirst, G., Ruhland, G., von Storch, I., Rosell-Mele, A., 1998. Calibration of the alkenone paleotemperature index U-37(K′) based on core-tops from the eastern South Atlantic and the global ocean (60°N–60°S). Geochim. Cosmochim. Acta 62, 1757–1772. Münch, Ph., Saint-Martin, J.P., Cornée, J.J., Féraud, G., Saint-Martin-Pestrea, S., Roger, S., Conesa, G., 2003. Controls on facies and sequence stratigraphy of an upper Miocene carbonate ramp and platform, Melilla basin, NE Morocco: comment. Sed. Geol. 158, 163–166. Negri, A., Villa, G., 2000. Calcareous nannofossil biostratigraphy, biochronology and paleoecology at the Tortonian/Messinian boundary of the Faneromeni section (Crete). Palaeogeogr. Palaeoclimatol. Palaeoecol. 156 (3), 195–209. Nijenhuis, I.A., Schenau, S.J., Van Der Weijden, C.H., Hilgen, F.J., Lourens, L.J., Zachariasse, W.J., 1996. On the origin of upper Miocene sapropelites: a case study from the Faneromeni Section, Crete (Greece). Paleoceanography 11, 633–645. Ohneiser, C., Florindo, F., Stocchi, P., Roberts, A.P., DeConto, R.M., Pollard, D., 2015. Antarctic glacio-eustatic contributions to late Miocene Mediterranean desiccation and reflooding. Nat. Commun. 6, 8765. Pagani, M., Freeman, K.H., Arthur, M.A., 1999. Late Miocene atmospheric CO2 concentrations and the expansion of C-4 grasses. Science 285, 876–879. Pierre, C., 1999. The oxygen and carbon isotope distribution in the Mediterranean water masses. Mar. Geol. 153, 41–55. Poulos, S.E., Drakopoulos, P.G., Collins, M.B., 1997. Seasonal variability in sea surface oceanographic conditions in the Aegean Sea (Eastern Mediterranean): an overview. J. Mar. Syst. 13, 225–244. https://doi.org/10.1016/S0924-7963(96)00113-3. Pound, M.J., Haywood, A.M., Salzmann, U., Riding, J.B., Lunt, D.J., Hunter, S.J., 2011. A Tortonian (Late Miocene, 11.61-7.25 Ma) global vegetation reconstruction. Paleogeogr. Paleoclimatol. Paleoecol. 300, 29–45. Pound, M.J., Haywood, A.M., Salzmann, U., Riding, J.B., 2012. Global vegetation dynamics and latitudinal temperature gradients during the Mid to Late Miocene (15.975.33 Ma). Earth Sci. Rev. 112, 1–22. Prahl, F.G., Sparrow, M.A., Wolfe, G.V., 2003. Physiological impacts on alkenone paleothermometry. Paleoceanography 18 (2), 1025. https://doi.org/10.1029/ 2002PA000803. Reghizzi, M., Gennari, R., Douville, E., Lugli, S., Manzi, V., Montagna, P., Roveri, M., Sierro, F.J., Taviani, M., 2017. Isotope stratigraphy (87Sr/86Sr, δ18O, δ13C) of the Sorbas basin (Betic Cordillera, Spain): paleoceanographic evolution across the onset of the Messinian salinity crisis. Palaeogeogr. Palaeoclimatol. Palaeoecol. 469, 60–73. https://doi.org/10.1016/j.palaeo.2016.12.039. Reuter, M., Brachert, T.C., Kroeger, K.F., 2006. Shallow-marine carbonates of the tropicaltemperate transition zone: effects of hinterland climate and basin physiography (Late Miocene, Crete/Greece). In: Pedley, H.M., Carannante, S. (Eds.), Cool-water Carbonates. Geol. Soc. London, pp. 157–178. Rodrigo-Gámiz, M., Rampen, S.W., Schouten, S., Sinninghe Damsté, J.S., 2016. The impact of oxic degradation on long chain alkyl diol distributions in Arabian Sea surface sediments. Org. Geochem. 100, 1–9. https://doi.org/10.1016/j.orggeochem.2016.07. 003. Roger, S., Münch, Ph., Cornée, J.J., Saint Martin, J.P., Féraud, G., Pestrea, S., Conesa, G., Ben Moussa, A., 2000. 40Ar/39Ar dating of the pre-evaporitic Messinian marine sequences of the Melilla basin (Morocco): a proposal for some biosedimentary events as isochrons around the Alboran Sea. Earth Planet. Sci. Lett. 179 (1), 101–113. https:// doi.org/10.1016/S0012-821X(00)00094-7. Rontani, J.-F., Volkman, J.K., Prahl, F.G., Wakeham, S.G., 2013. Biotic and abiotic degradation of alkenones and implications for UK′ 37 paleoproxy applications: a review. Org. Geochem. 59, 95–113. https://doi.org/10.1016/j.orggeochem.2013.04.005. Rosenthal, Y., Perron-Cashman, S., Lear, C.H., Bard, E., Barker, S., Billups, K., Bryan, M., Delaney, M.L., deMenocal, P.B., Dwyer, G.S., Eldefield, H., German, C.R., Greaves, M., Lea, D.W., Marchitto, T.M., Pak, D.K., Paradis, G.L., Russell, A.D., Schneider, R.R., Sceiderich, K., Stott, L., Tachikawa, K., Tappa, E., Thunnell, R., Wara, M., Wedeab S., Wilson, P.A., 2004. Interlaboratory comparison study of Mg/Ca and Sr/Ca measurements in planktonic foraminifera for paleoceanographic research. Geochem. Geophys. Geosyst., 5 (4), Q04D09, doi:https://doi.org/10.1029/2003GC000650. Roveri, M., Flecker, R., Krijgsman, W., Lofi, J., Lugli, S., Manzi, V., Sierro, F.J., Bertini, A., Carnerlenghi, A., De Lange, G., Govers, R., Hilgen, F.J., Hubscher, C., Meijer, P.T., Stoica, M., 2014. The Messinian Salinity Crisis: past and future of a great challenge for marine sciences. Mar. Geol. 352, 25–58. Salé, S.O., Gennari, R., Lugli, S., Manzi, V., Roveri, M., 2012. Tectonic and climatic control on the Late Messinian sedimentary evolution of the Nijar Basin (Betic Cordillera, Southern Spain). Bas. Res. 24, 314–337. https://doi.org/10.1111/j.13652117.2011.00527.x. Schenau, S.J., Antonarakou, A., Hilgen, F.J., Lourens, L.J., Nijenhuis, I.A., Van Der Wejden, C.H., Zachariasse, J.W., 1999. Organic rich layers in the Metochia section (Gavdos, Greece): evidence for a single sapropel formation mechanism in the eastern Mediterranean during the last 10 Myr. Paleoceanography 153, 117–135.

Krijgsman, W., Hilgen, F.J., Langereis, C.J., Zachariasse, W.J., 1994. The age of the Tortonian/Messinian Boundary. Earth Planet. Sci. Lett. 121, 533–547. Krijgsman, W., Hilgen, F.J., Langereis, C.G., Santarelli, A., Zachariasse, W.J., 1995. Late Miocene magnetostratigraphy, biostratigraphy and cyclostratigraphy in the Mediterranean. Earth Planet. Sci. Lett. 136, 475–494. https://doi.org/10.1016/0012821X(95)00206-R. Krijgsman, W., Hilgen, F.J., Negri, A., Wijbrans, J.R., Zachariasse, W.J., 1997. The Monte del Casino section (Northern Apennines, Italy): a potential Tortonian/Messinian boundary stratotype? Palaeogeogr. Palaeoclimatol. Palaeoecol. 133 (1), 27–47. https://doi.org/10.1016/S0031-0182(97)00039-4. Krijgsman, W., Hilgen, F.J., Marabini, S., Vai, G.B., 1999. New paleomagnetic and cyclostratigraphic age constraints on the Messinian of the Nothern Apenines (Vera del Gesso Basin, Italy). Mem. Soc. Geol. It. 54, 25–33. Krijgsman, W., Garcés, M., Agustí, J., Raffi, I., Taberner, C., Zachariasse, W.J., 2000. The “Tortonian salinity crisis” of the eastern Betics (Spain). Earth Planet. Sci. Lett. 181 (4), 497–511. https://doi.org/10.1016/S0012-821X(00)00224-7. Krijgsman, W., Capella, W., Simon, D., Hilgen, F.J., Kouwenhoven, T.J., Meijer, P.Th., Sierro, F.J., Tulbure, M.A., van den Berg, B.C.J., van der Schee, M., Flecker, R., 2018. The Gibraltar Corridor: Watergate of the Messinian Salinity Crisis. Mar. Geol. 403, 238–246. https://doi.org/10.1016/j.margeo.2018.06.008. Langereis, C.G., Hilgen, F.J., 1991. The Rossello composite: a Mediterranean and global reference section for the Early to early Late Pliocene. Earth Planet. Sci. Lett. 104, 211–225. LaRiviere, J., Ravelo, A.C., Crimmins, A., Dekens, P.S., Ford, H.L., Lyle, M., Wara, M.W., 2012. Late Miocene decoupling of oceanic warmth and atmospheric carbon dioxide forcing. Nature 486, 97–100. Lattaud, J., Lo, L., Huang, J.-J., Chou, Y.-M., Gorbarenko, S.A., Sinninghe Damsté, J.S., Schouten, S., 2018. A comparison of late Quaternary organic proxy-based paleotemperature records of the central Sea of Okhotsk. Paleoceanogr. Paleocl. 33. https:// doi.org/10.1029/2018PA003388. Lea, D.W., Mashiotta, T., Spero, H., 1999. Controls on magnesium and strontium uptake in planktonic foraminifera determined by live culturing. Geochim. Cosmochim. Acta 63 (16), 2369–2379. Lear, C.H., Rosenthal, Y., Wright, J.D., 2003. The closing of a seaway: ocean water masses and global climate change. Earth Planet. Sci. Lett. 210 (3–4), 425–436. https://doi. org/10.1016/s0012-821x(03)00164-x. Lear, C.H., Coxall, H.K., Fos ter, G.L., Lunt, D.J., Mawbey, E.M., Rosenthal, Y., Sosdian, S.M., Thomas, E., Wil son, P.A., 2015. Neogene ice volume and ocean temperatures: insights from infaunal foraminiferal Mg/Ca paleothermometry. Paleoceanography 30, 1437–1454. Leroux, E., Aslanian, D., Rabineau, M., Pellen, R., Moulin, M., 2018. The late Messinian event: a worldwide tectonic revolution. Terra Nova 00, 1–8. https://doi.org/10. 1111/ter.12327. Lirer, F., Sprovieri, M., Vallefuoco, M., Ferraro, L., Pelosi, N., Giordano, L., Capotondi, L., 2014. Planktonic foraminifera as bio-indicators for monitoring the climatic changes occurred during the last 2000 years in the SE Tyrrhenian Sea. Integr. Zool. 9, 542–554. Lopes dos Santos, R.A., Spooner, M.I., Barrows, T.T., De Deckker, P., Sinninghe Damsté, H J.S., Schouten, S., 2013. Comparison of organic (Uk′ 37, TEX86, LDI) and faunal proxies (foraminiferal assemblages) for reconstruction of late Quaternary sea surface temperature variability from offshore southeastern Australia. Paleoceanography 28, 377–387. https://doi.org/10.1002/palo.20035. Lourens, L.J., Antonarakou, A., Hilgen, F.J., Van Hoof, A.A.M., Vergnaud Grazzini, C., Zachariasse, W.J., 1996. Evaluation of the Plio-Pleistocene astronomical timescale. Paleoceanography 11, 391–413. Lourens, L., Hilgen, F.J., Shackleton, N.J., Laskar, J., Wilson, D., 2004. The Neogene period. In: Gradstein, F.M., Ogg, J.G., Smith, A.G. (Eds.), A Geologic Time Scale 2004. Cambridge University Press, pp. 409–440. Louvari, M.A., Drinia, H., Kontakiotis, G., Di Bella, L., Antonarakou, A., Anastasakis, G., 2019. Impact of latest-glacial to Holocene sea-level oscillations on central Aegean shelf ecosystems: a benthic foraminiferal palaeoenvironmental assessment of South Evoikos Gulf, Greece. J. Mar. Syst. 199, 103181. https://doi.org/10.1016/j.jmarsys. 2019.05.007. Lozar, F., Violanti, D., Bernardi, E., Dela Pierre, F., Natalicchio, M., 2018. Identifying the onset of the Messinian salinity crisis: a reassessment of the biochronostratigraphic tools (Piedmont Basin, NW Italy). Newslett. Stratigr. 51 (1), 11–31. Manzi, V., Gennari, R., Hilgen, F., Krijgsman, W., Lugli, S., Roveri, M., Sierro, F.J., 2013. Age refinement of the Messinian salinity crisis onset in the Mediterranean. Terra Nova 25, 315–322. Manzi, V., Lugli, S., Roveri, M., Dela, Pierre F., Gennari, R., Lozar, F., Natalicchio, M., Schreiber, B.C., Taviani, M., Turco, E., 2016. The Messinian salinity crisis in Cyprus: a further step towards a new stratigraphic framework for Eastern Mediterranean. Bas. Res. 28, 207–236. Manzi, V., Gennari, R., Lugli, S., Persico, D., Reghizzi, M., Roveri, M., Schreiber, B.C., Calvo, R., Gavriel, I., Gvirtzman, Z., 2018. The onset of the Messinian salinity crisis in the deep Eastern Mediterranean basin. Terra Nova 30, 189–198. https://doi.org/10. 1111/ter.12325198. Medina-Elizalde, M., Lea, D.W., 2005. The Mid-Pleistocene transition in the tropical Pacific. Science 310, 1009–1012. Mertz-Kraus, R., Brachert, T.C., Reuter, M., Galer, S.J.G., Fassoulas, C., Iliopoulos, G., 2009a. Late Miocene sea surface salinity variability and paleoclimate conditions in the Eastern Mediterranean inferred from coral aragonite δ18O. Chem. Geol. 262 (3–4), 202–216. Mertz-Kraus, R., Brachert, T.C., Jochum, K.P., Reuter, M., Stoll, B., 2009b. LA-ICP-MS analyses on coral growth increments reveal heavy winter rain in the Eastern Mediterranean at 9 Ma. Palaeogeogr. Palaeoclimatol. Palaeoecol. 273, 25–40.

15

Palaeogeography, Palaeoclimatology, Palaeoecology 534 (2019) 109312

G. Kontakiotis, et al.

domain. Paleoceanography 20 (1). van der Laan, E., Hilgen, F.J., Lourens, L.J., de Kaenel, E., Gaboardi, S., Iaccarino, S., 2012. Astronomical forcing of Northwest African climate and glacial history during the late Messinian (6.5-5.5 Ma). Paleogeogr. Paleoclimatol. Paleoecol. 313, 107–126. van der Zwaan, G.J., Gudjonsson, L., 1986. Middle Miocene-Pliocene stable isotope stratigraphy and paleoceanography of the Mediterranean. Mar. Micropal. 10, 71–90. Vasiliev, I., Mezger, E.M., Lugli, S., Reichart, G.J., Manzi, V., Roveri, M., 2017. How dry was the Mediterranean during the Messinian Salinity Crisis? Palaeogeogr. Palaeoclimatol. Palaeoecol. 471, 120–133. https://doi.org/10.1016/j.palaeo.2017. 01.032. Vasiliev, I., Karakitsios, V., Bouloubassi, I., Agiadi, K., Kontakiotis, G., Antonarakou, A., Triantaphyllou, M., Gogou, A., Kafousia, N., de Rafélis, M., Zarkogiannis, S., Kaczmar, F., Parinos, C., Pasadakis, N., 2019. Large sea surface temperature, salinity, and productivity-preservation changes preceding the onset of the Messinian Salinity Crisis in the eastern Mediterranean Sea. Paleoceanogr. Paleocl. 34 (2), 182–202. https://doi.org/10.1029/2018PA003438. Vincent, E., Killingley, J.S., Berger, W.H., 1980. The magnetic epoch 6 carbon shift: a change in the ocean's 13/C12C ratio 6.2 million years ago. Mar. Micropal. 5, 185–203. Williams, M., Haywood, A.M., Taylor, S.P., Valdes, P.J., Sellwood, B.W., Hillenbrand, C.D., 2005. Evaluating the efficacy of planktonic foraminifer calcite delta O-18 data for sea surface temperature reconstruction for the Late Miocene. Geobios 38, 843–863. Zachariasse, W.J., van Hinsbergen, D.J.J., Fortuin, A.R., 2008. Mass wasting and uplift on Crete and Karpathos during the early Pliocene related to initiation of south Aegean left-lateral, strike-slip tectonics. GSA Bull. 120, 976–993. Zachariasse, W.J., van Hinsbergen, D.J.J., Fortuin, A.R., 2011. Formation and fragmentation of a late Miocene supradetachment basin in central Crete: implications for exhumation mechanisms of high-pressure rocks in the Aegean forearc. Bas. Res. 23, 678–701. Zachos, J., Pagani, M., Sloan, L., Thomas, E., Billups, K., 2001. Trends, rhythms, and aberrations in global climate 65 Ma to present. Science 292 (5517), 686–693. Zarkogiannis, S.D., Antonarakou, A., Tripati, A., Kontakiotis, G., Mortyn, P.G., Drinia, H., Greaves, M., 2019. Influence of surface ocean density on planktonic foraminifera calcification. Sci. Rep. 9 (1), 533. https://doi.org/10.1038/s41598-018-36935-7. Zelilidis, A., Tserolas, P., Chamilaki, Ε., Pasadakis, N., Kostopoulou, S., Maravelis, A.G., 2016. Hydrocarbon prospectivity in the Hellenic trench system: organic geochemistry and source rock potential of upper Miocene-lower Pliocene successions in the eastern Crete Island, Greece. Int. J. Earth Sci. 105, 1859–1878. https://doi.org/10.1007/ s00531-015-1278-8. Zidianakis, G., Mohr, B.A.R., Fassoulas, C., 2007. A late Miocene leaf assemblage from Vrysses, western Crete, Greece, and its paleoenvironmental and paleoclimatic interpretation. Geodiversitas 29 (3), 351–377. Zidianakis, G., Iliopoulos, G., Zelilidis, A., Kovar-Eder, J., 2015. Myrica from the plant assemblage of Pitsidia (Crete, Late Miocene): putting the puzzle together. Palaeontographica Abt B: Palaeobotany-Paleophytology 293 (1–6), 149–171. Zidianakis, G., Iliopoulos, G., Zelilidis, A., Kovar-Eder, J., 2016. Pinus remains from the Pitsidia plant assemblage document coastal pine forests in southern Crete during the late Miocene. Rev. Palaeobοt. Palyno. 235, 11–30.

Schmidt, G.A., 1999. Error analysis of paleosalinity calculation. Paleoceanography 14, 422–429. Schrag, D.P., Adkins, J.F., McIntyre, K., Alexander, J.L., Hodell, D.A., Charles, C.D., McManus, J.F., 2002. The oxygen isotopic composition of seawater during the Last Glacial Maximum. Quat. Sci. Rev. 21, 331–342. https://doi.org/10.1016/S02773791(01)00110-X. Schuster, M., Duringer, P., Ghienne, J.F., Vignaud, P., Mackaye, H.T., Likius, A., Brunet, M., 2006. The age of the Sahara Desert. Science 311, 821. Sierro, F.J., 1985. The replacement of the Globorotalia - menardii group by the Globorotalia - miotumida group - an aid to recognizing the Tortonian - Messinian boundary in the Mediterranean and adjacent Atlantic. Mar. Micropal. 9, 525–535. Sierro, F.J., Flores, J.A., Civis, J., Gonzales Delgado, J.A., Frances, G., 1993. Late Miocene globorotaliid event-stratigraphy and biogeography in the NE-Atlantic and Mediterranean. Mar. Micropal. 21, 143–168. Sierro, F.J., Flores, J.A., Frances, G., Vazquez, A., Utrilla, R., Zamarreno, I., Erlenkeuser, H., Barcena, M.A., 2003. Orbitally-controlled oscillations in planktic communities and cyclic changes in western Mediterranean hydrography during the Messinian. Palaeogeogr. Palaeoclimatol. Palaeoecol. 190, 289–316. Smith, M., De Deckker, P., Rogers, J., Brocks, J., Hope, J., Schmidt, S., Lopes dos Santos, H R., Schouten, S., 2013. Comparison of organic (UK′ 37, TEX86, LDI) and faunal proxies (foraminiferal assemblages) for reconstruction of late Quaternary sea surface temperature variability from offshore southeastern Australia. Org. Geochem. 64, 94–104. https://doi.org/10.1016/j.orggeochem.2013.08.015. Sprovieri, M., Bellanca, A., Neri, R., Mazzola, S., Bonanno, A., Patti, B., Sorgente, R., 1999. Astronomical calibration of late Miocene stratigraphic events and analysis of precessionally driven paleoceanographic changes in the Mediterranean basin. Mem. Soc. Geol. Ital. 54, 7–24. Sprovieri, M., d'Alcala, M.R., Manta, D.S., Bellanca, A., Neri, R., Lirer, F., Taberner, C., Pueyo, J.J., Sammartino, S., 2008. Ba/Ca evolution in water masses of the Mediterranean late Neogene. Paleoceanography 23, PA3205. https://doi.org/10. 1029/2007PA001469. Tulbure, M.A., Capella, W., Barhoun, N., Flores, J.A., Hilgen, F.J., Krijgsman, W., Kouwenhoven, T., Sierro, F.J., Yousfi, M.Z., 2017. Age refinement and basin evolution of the North Rifian Corridor (Morocco): no evidence for a marine connection during the Messinian Salinity Crisis. Palaeogeogr. Palaeoclimatol. Palaeoecol. 485, 416–432. Tzanova, A., 2015. Linking Late Miocene-Pliocene Sea Surface Temperatures, Terrestrial Biome Transitions and Intensification of Glaciation Using Alkenone Paleothermometry. In: PhD Thesis. Providence, Rhode Island. Tzanova, A., Herbert, T.D., Peterson, L., 2015. Cooling Mediterranean Sea surface temperatures during the Late Miocene provide a climate context for evolutionary transitions in Africa and Eurasia. Earth Planet. Sci. Lett. 419, 71–80. van Assen, E., Kuiper, K.F., Barhoun, N., Krijgsman, W., Sierro, F.J., 2006. Messinian astrochronology of the Melilla Basin: stepwise restriction of the Mediterranean–Atlantic connection through Morocco. Palaeogeogr. Palaeoclimatol. Palaeoecol. 238, 15–31. van der Laan, E., Gaboardi, S., Hilgen, F.J., Lourens, L.J., 2005. Regional climate and glacial control on high-resolution oxygen isotope records from Ain el Beida (latest Miocene, northwest Morocco): a cyclostratigraphic analysis in the depth and time

16