Dynamics of the last four glacial terminations recorded in Lake Van, Turkey

Quaternary Science Reviews 104 (2014) 42e52

Contents lists available at ScienceDirect

Quaternary Science Reviews journal homepage: www.elsevier.com/locate/quascirev

Dynamics of the last four glacial terminations recorded in Lake Van, Turkey Ola Kwiecien a, b, *, Mona Stockhecke a, b, Nadine Pickarski c, Georg Heumann c, Thomas Litt c, Michael Sturm a, Flavio Anselmetti d, Rolf Kipfer b, Gerald H. Haug a a

ETH Zurich, Climate Geology, Sonneggstrasse 5, 8092 Zurich, Switzerland Eawag Dubendorf, Ueberlandstrasse 130, 8600 Duebendorf, Switzerland Steinmann Institute of Geology, Mineralogy and Paleontology, Bonn University, Nussallee 8, 53115 Bonn, Germany d Institute of Geological Sciences and Oeschger Centre for Climate Change Research, University of Bern, Baltzerstrasse 1-3, 3012 Bern, Switzerland b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 5 August 2013 Received in revised form 22 May 2014 Accepted 1 July 2014 Available online 22 July 2014

A well-dated suite of Lake Van climate-proxy data covering the last 360 ka documents environmental changes over 4 glacial/interglacial cycles in Eastern Anatolia, Turkey. The picture of cold and dry glacials and warm and wet interglacials emerging from pollen, organic carbon, authigenic carbonate content, elemental profiling by XRF and lithological analyses is inconsistent with classical interpretation of oxygen isotopic composition of carbonates pointing to a more complex pattern in Lake Van region. Detailed analysis of glacial terminations allows for the constraining of a depositional model explaining different patterns observed in all the proxies. We hypothesize that variations in relative contribution of rainfall, snowmelt and glacier meltwater recharging the basin have a very important role for all sedimentary processes in Lake Van. Lake level of glacial Lake Van, predominantly fed by snowmelt, was low, the water column was oxic, and carbonates precipitating in the epilimnion recorded the light isotopic signature of inflow. During terminations, increasing rainfall and significant supply of mountain glaciers' meltwater contributed to lake level rise. Increased rainfall enhanced density gradients in the water column, and hindered mixing leading to development of bottom-water anoxia. Carbonates precipitating during terminations show large fluctuations in their isotopic composition. Full interglacial conditions in Lake Van are characterized by high or slowly falling lake level. Rainfall and snowmelt feed the lake but due to re-established mixing, the isotopic composition of authigenic carbonates is heavier and closer to that of evaporation-influenced lake water than that of runoff representing snowmelt and atmospheric precipitation. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Lake Van Glacial/interglacial cycle Termination Multi-proxy approach Oxygen isotopes Eastern Mediterranean

1. Introduction Since their first discovery in the 19th century (for review see Imbrie and Imbrie, 1979) the imprint of Quaternary glacial/interglacial cycles on atmospheric and oceanic circulation has been documented in numerous marine and continental climate archives (Shackleton and Opdyke, 1973; McManus et al., 1999; Jaccard et al., 2005; Prokopenko et al., 2006; Jouzel et al., 2007; Martrat et al., 2007; Meckler et al., 2012; Melles et al., 2012). As proposed originally by Milankovich, the main pacemaker of Ice Ages are the changes in Earth's orbital parameters (Hays et al., 1976) with

* Corresponding author. Present address: Ruhr-University Bochum, Sediment and €tsstrasse 150, D-44801 Bochum, Germany. Isotope Geology, Universita E-mail address: [email protected] (O. Kwiecien). http://dx.doi.org/10.1016/j.quascirev.2014.07.001 0277-3791/© 2014 Elsevier Ltd. All rights reserved.

internal feedbacks playing a significant role (Ruddiman, 2003). Terminations of Ice Ages are the most spectacular examples of nearsynchronous global and abrupt climate change. The structure of termination after Denton et al. (2010) includes following scenario: rising Northern Hemisphere summer insolation leads to ice-sheet melting, which delivers meltwater into the Atlantic. Reduced Atlantic Meridional Overturning Circulation (AMOC) facilitates the spread of winter sea ice, hampers the heat transport to the midlatitudes and tropics, and weakens the Asian monsoon. The Intertropical Convergence Zone (ITCZ) and westerly winds of both hemispheres shift southwards, with southern westerlies warming Antarctica and triggering degassing of the Southern Ocean deep CO2 reservoir. Released CO2 promotes further warming and sustains interglacial conditions. The consistency of singular elements of this comprehensive layout and their temporal arrangement has been reported from the marine realm (McManus et al., 2004; Jaccard

O. Kwiecien et al. / Quaternary Science Reviews 104 (2014) 42e52

et al., 2005; Marchitto et al., 2007; Meckler et al., 2013), however, the detailed structure and the sequence of changes occurring during terminations have been to date rarely studied in continental archives. In particular, it is difficult to follow the response of the continental ecosystem to climate forcing (e.g., deglacial warming) over a course of several cycles; most of the continental sedimentary sequences are too short or of insufficient resolution. While studies of past interglacials shed light on different aspects of climate change in a warmer world (Tzedakis and Bennett, 1996; Frogley et al., 1999; Brauer et al., 2008), very few comparative approaches of several glacial terminations recovered from the same continental archive exist (Prokopenko et al., 2006; Cheng et al., 2009). The Eastern Mediterranean region yielded a couple of continental records covering at least one full glacial/interglacial cycle (Bar-Matthews et al., 2000; Torfstein et al., 2009; Develle et al., 2011; Djamali et al., 2012; Stevens et al., 2012), or capturing the termination or the glacial inception (e.g. Bartov et al., 2003; Jones et al., 2007; Waldmann et al., 2009). Not all of these records have an independent chronology and most of them focus on one proxy only (e.g.: pollen assemblages, isotopic composition of speleothems, isotopic composition of lacustrine carbonates, lake level). Consequently reconstructions of the past Eastern Mediterranean hydrological regime are not always consistent and provoke a debate between two competing hypothesis: wet interglacials/dry glacials versus dry interglacials/wet glacials (Gasse et al., 2011). Here we present the last four glacial/interglacial cycles recorded in sedimentary proxies of Lake Van in Eastern Anatolia, Turkey. In order to gain insight into paleoenvironmental changes, we analyzed elemental intensities, amount of total organic carbon and carbonate, oxygen isotope composition of sedimentary carbonates and pollen assemblages. Detailed lithological analyses of Lake Van sediments by Stockhecke et al. (2014a), including organic carbon and carbonate data, constituted the conceptual framework for the present study. We took advantage of this work and juxtaposed extended geochemical and pollen time-series with interpretation based on the facies analysis. The age model of Lake Van record is robust, but not independent; therefore we focus primarily on the relation between different proxies from the same profile over the course of the last four glacial/interglacial cycles rather than

43

compare our record with other archives. This work emphasizes three issues: (1) documenting the internal consistency of the events in the glacial/interglacial sequence; (2) understanding mechanisms responsible for the regional climate change at the glacial/interglacial transition (termination); (3) testing the hypothesis of wet interglacials/dry glacials versus dry interglacials/wet glacials in the context of entire Eastern Mediterranean region. 2. Regional setting Lake Van is a terminal alkaline lake in Eastern Anatolia, Turkey, located on a high plateau at 1648 m above sea level (m a.s.l). In the south, the lake is surrounded by Bitlis Massif and in the north by several volcanoes reaching the altitude of 4000 m a.s.l. (Schweizer, 1975). The ~400 m deep bowl-shaped basin has relatively wide and shallow slopes (Cukur et al., 2012). Its water chemistry and the regional climate, in winter influenced by the position of the westerly jet stream and in summer by the extension of subtropical high-pressure belt (Wigley and Farmer, 1982), make Lake Van an excellent witness of environmental changes (Fig. 1). As is the case with many terminal lakes in volcanically active regions, Lake Van's water column is characterized by NaeCO3eCl chemistry and is saturated with respect to carbonate (Lemcke, 1996; Reimer et al., 2008). Today, the pronounced seasonality of regional climate is essential for depositional processes in Lake Van. Atmospheric precipitation of Mediterranean origin falls during the cold and long winters as snow. As winter snowfall changes to spring rainfall, the amount of precipitation increases. The summers are hot and dry and the precipitation increases again in autumn. Ca2þ is supplied to the lake along with clastic particles by the increased river discharge during the spring (snowmelt and rainfall peak). Clastic constituents tend to settle directly to the lake bottom, while part of the Ca2þ precipitates immediately in form of whitings. The remaining Ca2þ is retained in the upper water column where it precipitates in late summer and early autumn facilitated by phytoplankton activity and wind-driven mixing of the lake (Stockhecke et al., 2012). Light autochthonous organic matter settles during winter. The physical expression of this annual cycle is a deposition of dark (clastic particles, organic matter) and light (carbonate) laminae (Lemcke, 1996;

Fig. 1. Climatic setting of Lake Van, modified after Wigley and Farmer (1982) and Akcar and Schlüchter (2005), and location of the Ahlat Ridge (AR). Bathymetry modified after Cukur et al. (2012).

44

O. Kwiecien et al. / Quaternary Science Reviews 104 (2014) 42e52

Stockhecke et al., 2012) preserved in the sediment due to the anoxic conditions at the bottom of the lake. The main process responsible for mixing and deep-water renewal bringing oxygen to the bottom of saline lakes, are density currents driven by gradients in salinity (Peeters and Kipfer, 2009). Increased freshwater inflow raises the lake level and reduces the density of surface water hampering formation of density currents and suppressing deep-water renewal. As an example, a recent 2-m rise of Lake Van level was followed by upward migration of oxiceanoxic boundary and formation of about 100-m thick, anoxic deep-water body (Kaden et al., 2010). This observation implies that even relatively small changes in the hydrological regime can have considerable consequences for the lacustrine ecosystem. 3. Materials and methods 3.1. Ahlat Ridge (AR) composite profile In the frame of the ICDP PALEOVAN drilling campaign (Litt et al., 2009) an international scientific team recovered in summer 2010 a 219-m-long sediment succession down to the rock basement of Lake Van (Litt et al., 2012). Investigated site, the Ahlat Ridge (AR) was drilled at a water depth of 375 m. This location was chosen in order to retrieve continuous and undisturbed sedimentary sequence. The AR composite profile covers ~600 ka of Lake Van's history and documents several glacial/interglacial cycles (Stockhecke et al., 2014a,b). In its initial phase, Lake Van was most probably a freshwater basin. However, a change in depositional regime, deterioration of core quality and lower resolution of proxy data hinder paleoclimatic reconstruction for sediments older than 400 ka. Thus, this work focuses on the last four glacial/interglacial cycles and, respectively, the anatomy of the last four terminations. 3.2. XRF elemental profiling After splitting longitudinally, sediment cores were scanned with an AVAATECH XRF Core Scanner III at the MARUM e University of Bremen. The XRF core scanner, developed at the Netherlands Institute for Sea Research Texel, measures the bulk intensities of major elements (e.g. Al, Si, K, Ca, Ti, Fe) on sediment cores (Jansen €hl and Abrams, 2000). XRF data were collected et al., 1998; Ro every 2 cm down-core over a 1 cm2 area with a slit size of 12 mm using generator settings of 10 kV, a current of 0.2 mA and a sampling time of 10 s. The split-core surface was covered with a 4 micron thick SPEXCerti Prep Ultralene1 foil to avoid contamination of the XRF measurement unit and desiccation of the sediment. The data reported here have been acquired by a Canberra X-PIPS Silicon Drift Detector (SDD; Model SXD 15C-150-500) with 150 eV X-ray resolution, the Canberra Digital Spectrum Analyzer DAS 1000 and an Oxford Instruments 100W Neptune X-ray tube with rhodium (Rh) target material. Raw spectra were processed by the Analysis of X-ray spectra by Iterative Least square software (WIN AXIL) package from Canberra Eurisys. This software provides intensity data in total counts (tc). All elemental data in this paper are measured by means of XRF profiling, elemental ratios presented here are unitless.

samples yielded standard deviation of ±2.6% for TC and ±5.4% for TIC. Total organic carbon (TOC) was calculated as TOC ¼ TC e TIC. Carbonate (CaCO3) content was calculated using formula CaCO3 ¼ TIC * 8.33, under the assumption that all inorganic carbon is bound to calcium carbonate. 3.4. Pollen Pollen assemblages were analyzed at 0.2 and 1-m sampling resolution for intervals 0e60 and 60e220 m composite below lakefloor (mcblf), respectively. Here, we present only the arboreal (tree) pollen (AP) contribution to the pollen assemblage. Sample preparation generally followed the standard method described by Berglund and Ralska-Jasiewiczowa (1986) and Faegri and Iversen (1989). Method's details and the overview of vegetational development in Lake Van region over the last 600 ka BP are provided by Litt et al. (2014). 3.5. Isotopic composition of carbonates (d18Obulk) Stable isotope values of bulk carbonate (d18Obulk) were analyzed at 0.2 and 1-m sampling resolution for intervals 0e60 and 60e220 mcblf, respectively. The freeze-dried and ground sediment samples were determined at the University of Kiel using a Finningan GasBenchII with carbonate option coupled to a DELTAplusXL IRMS. The isotope compositions were given relative to the VPDB standard in the conventional delta notation and were calibrated against two international reference standards (NBS19 and NBS18). The standard deviation for reference analyses was 0.06‰ for both, d13C and d18O. 4. Lithology Stockhecke et al. (2014a) provides detailed description and interpretation of the Lake Van lithological units in terms of changing environmental conditions. Within the considered interval (0e160 mcblf), lacustrine clayey silt consists of five recurring lithotypes manifesting different depositional regimes of background sedimentation. The finely and faintly laminated clayey silts stand out prominently against the banded, massive and mottled clayey silts. The finely laminated clayey silts were deposited during periods of strong seasonality and high or rising lake levels. The other lithologies show evidence of bioturbation and an oxiceanoxic boundary close to the water sediment interface during falling lake level or lowstands. All of these occur either as single packages or intercalations of two or three lithotypes. Two other lithotypes, graded beds and volcaniclastics, reflect event deposition of allochthonous and/or reworked lacustrine material. In general, applying the best-fit age model (Stockhecke et al., 2014b), laminated and faintly laminated as well as CaCO3-rich banded clayey silts characterize interglacial deposition, while CaCO3-poor banded, and mottled clays dominate during glacials. Frequent intercalations of graded layers mark major lake level rises during deglaciation periods. 5. Chronology

3.3. Total organic carbon (TOC) and carbonate content Samples for carbon measurements were taken at 20 cm resolution. In some intervals (e.g., terminations), resolution was increased to 2.5 cm. The freeze-dried and ground sediment samples were analyzed for total carbon (TC) using an elemental analyzer (HEKAtech Euro Elemental Analyzer). Total inorganic carbon (TIC) content was determined using a titration coulometer (Coulometric Inc., 5011 CO2-Coulometer). Repeated measurements for 112

The age model for the complete AR site composite profile is presented in a chronology-dedicated paper in this issue (Stockhecke et al., 2014b). In order to prepare a data set consisting only of background sedimentation, the thickness of all event deposits was removed from the composite profile, resulting in mcblf-nE (meter composite below lake-floor e no event) depth scale (Stockhecke et al., 2014b). Similarly, the values measured in the event deposits were eliminated from the data sets. Briefly, the age model is

O. Kwiecien et al. / Quaternary Science Reviews 104 (2014) 42e52

constrained by tuning several Lake Van proxy records, including total organic carbon (TOC), sediment colour (reflectance b*), Ca/K and arboreal pollen percentage to the two reference curves: GICC05-based NGRIP isotopic record (NGRIP project members et al., 2004; Steffensen et al., 2008; Svensson et al., 2008; Wolff et al., 2010) for the interval between 0 and 116 ka BP, and the speleothem-based synthetic Greenland record GLT-syn (Barker et al., 2011) for the interval between 116 and 400 ka BP. The accuracy of the age model is confirmed by good correspondence between tuned ages and radiometric ages (14C dating of twigs in younger sediments, 40Ar/39Ar ages on tephra in older sediments) and the comparison of the relative paleointensity record of the magnetic field with reference curve PISO-1500 (Stockhecke et al., 2014a,b). 6. Results 6.1. XRF elemental profiling XRF scanning provides qualitative data, and intensities of individual elements need to be treated with caution (Wilhelms-Dick et al., 2012). Metals such as Ti, K or Fe are traditionally linked to siliciclastic components of the sediment, additionally Fe and Mn are sensitive to changes in redox conditions. Si is related to quartz and clay minerals but also to biogenic opal (diatoms or spicules of freshwater sponges, however the latter were not observed in Lake Van sediments), Ca is commonly related to carbonates. The use of elemental ratios is recognized as a more robust approach for representing variations in the relative proportion of biogenic silica/

45

siliciclastic (e.g.: Si/K), authigenic carbonate/siliciclastic material (e.g.: Ca/K) (e.g.: Arz et al., 1998; Kwiecien et al., 2009) or redox conditions (Mn/Fe) (Melles et al., 2012). Overall good correspondence between Si/K and TOC (discussed below) advocates biogenic source of silica. Biogenic silica is produced by diatoms, thus higher values of Si/K indicate increased primary productivity, but in alkaline lakes where diatoms’ frustules are not well preserved, higher Si/K might point to decreased alkalinity. Noteworthy both of these processes are ultimately related to an increased freshwater inflow bringing dissolved nutrients and diluting alkalinity. While biogenic silica is a minor component, authigenic carbonate and terrigenous clays are the two major components of the Lake Van sedimentary system. The Ca/K signal is a proxy of relative changes between carbonate precipitation and terrigenous fluxes and it is well reproduced by the CaCO3 content. In the Lake Van composite profile, Si/K and Ca/K ratios are higher during interglacials and lower during glacials (Fig. 2). Fe and Mn are sensitive to redox changes in aquatic environments with oxic forms of both elements being mostly soluble, but having different Eh stability fields. As Fe(II) is more rapidly oxidized the Mn/Fe can be used as an indicator for syn- and postdepositional redox conditions (e.g. Melles et al., 2012), with lower values suggesting low oxygenation of the lake bottom water. Due to the weak Mn signal Lake Van Mn/Fe ratio is generally low. Still, even if not robust; handled with caution this is our only anoxia-sensitive proxy independent of productivity. Our rationale behind using Mn/ Fe ratio lies in its comparison with the TOC (discussed below) and Si/K records to separate different processes resulting in higher content of organic matter in the sediment. Mn/Fe is lowest during

Fig. 2. Comparison of Lake Van sedimentary proxies including XRF elemental ratios, Ca/K, Si/K and Mn/Fe; arboreal pollen (AP) percentage, total organic carbon (TOC) and CaCO3 weight contents and d18Obulk composition of carbonates. Reference curves: difference between June and December insolation at 39 N, reflecting changes in seasonality (Laskar et al., 2004), isotopic composition of NGRIP ice core (NGRIP project members et al., 2004; Steffensen et al., 2008; Svensson et al., 2008; Wolff et al., 2010) and synthetic isotopic composition of Greenland ice (Barker et al., 2011). *Conventionally, higher Mn/Fe points to more oxygenated bottom water. Due to low Mn intensity in Lake Van profile Mn/Fe is not very robust proxy and here we focus only on the abrupt increases of Mn/Fe during terminations and interpret them in terms of re-establishment of deep-water renewal.

46

O. Kwiecien et al. / Quaternary Science Reviews 104 (2014) 42e52

Conventionally, higher arboreal pollen (AP) percentage is an indicator of more humid conditions. Although of low resolution, the AP profile depicts best the degree of change between glacial and interglacial conditions (Fig. 2), with high amount of tree pollen pointing to more moisture availability during interglacials and very low amounts during the moisture-deficient glacial intervals. TOC and CaCO3 content show consistently higher values during interglacials and lower values during glacials (Fig. 2). TOC is a composite measure of lake productivity, input of allochtonous organic matter and preservation. However, sediment trap analysis (Stockhecke et al., 2012) and microscopic facies analysis (Stockhecke et al., 2014a) suggest that in Lake Van contribution of allochtonous organic matter is negligible. Further, the comparison between the TOC, Si/K and Mn/Fe records implies that only during terminations higher TOC values are related to better preservation of organic carbon. The oxygen isotopic composition of bulk lacustrine carbonates (d18Obulk) reflects temperature and oxygen isotopic composition of the lake water. In lakes, the latter is governed by the moisture balance between input and output, in particular the isotopic composition of runoff, and the precipitation to evaporation ratio (Leng and Marshall, 2004). Closed lakes are more readily affected by evaporative concentration due to the long hydraulic residence time of the water. Overall, the resolution of the d18Obulk profile is yet too low for a clear visual correlation to glacial/interglacial cycles. In the higher-resolved interval (0e60 mcblf, corresponding to 0e136 ka BP) d18Obulk reach maximum values during penultimate interglacial. Glacials are characterized by smaller variations in d18Obulk than interglacials. On average Last Interglacial values are heavier than glacial ones (Fig. 2, Table 1). Notably a comparison between Last Glacial and Holocene d18Obulk values seems inconclusive; in agreement with previously published records (Lemcke and Sturm, 1997; Litt et al., 2009) d18Obulk signature of early Holocene is lighter, while that of late Holocene heavier than glacial values.

associated with reduced precipitation and runoff deduced from lower amount of arboreal pollen and higher detrital input. We define ‘wet’ conditions related to increased precipitation and runoff indicated by higher amount of arboreal pollen and lower detrital input. During wetter interglacials, advancing vegetation cover hampers erosion and surface runoff, and is therefore a factor limiting supply of fine-grained detrital material into the lake. Additionally, higher lake level reduces the exposure of shallow subaquatic slopes further diminishing transport of detrital material into the lake. lu et al. (2010) reported lake terraces of Last (However, Kuzucuog Glacial age above the level of modern lake, pointing to higher lake level during glacial.) Increased freshwater runoff brings dissolved nutrients and forms a less dense layer on top of denser saline lake water. While supply of nutrients facilitates primary productivity, the increase in density gradients between the surface water the deep water slows down the mixing, consequently introducing bottom-water anoxia as observed over the last decade (Kaden et al., 2010; Stockhecke et al., 2012, 2014a). This interpretation is, however, seemingly incompatible with d18Obulk values (heavier during parts of interglacials, lighter during parts of glacials) if explained in a classical way. The course of Termination I (Fig. 3a) was studied previously in detail (Landmann et al., 1996; Lemcke and Sturm, 1997). These works documented general similarities between glacial (15e12 ka) and Holocene isotopic signature of authigenic carbonates, and shown that Younger Dyras (YD) e a cold spell during the last termination e is characterized, in comparison to glacial and Holocene, by heavy d18O values. In congruence with pollen (Wick et al., 2003) and carbonate Mg/Ca data (Lemcke and Sturm, 1997) the d18Obulk record was regarded as reflecting changes in the precipitation evaporation ratio (P:E). Periods of negative net water balance lead to removal of the lighter isotope resulting in more positive isotope values. Alternatively, during periods of positive net water balance, isotope values become more negative. Consequently, YD heavy isotopic excursion was interpreted as a cold and very dry interval superimposed on general dry conditions of climate emerging from the glacial period. However, extending this classical interpretation to our glacial/interglacial record, heavier d18O of authigenic carbonate precipitating in the epilimnion of Lake Van would imply drier conditions during parts of interglacials.

7. Discussion

7.2. Possible explanations for d18Obulk record

7.1. Inter-proxy comparison on glacial/interglacial scale

In their critical assessment of carbonate d18O data from three Eastern Mediterranean lakes Jones and Roberts (2008) carefully concluded, that although precipitation and evaporation are important hydrological parameters, carbonate d18O interpretation based on general assumptions for given lake types might be oversimplified. The authors warn of compromising or forcing the interpretation of isotope data to agree with climate reconstruction based on other proxies. Further, Jones et al. (2007) uses an example

terminations and its most prominent feature that we highlight, is a rapid increase, consequently lagging increase in other proxies (Fig. 2). Therefore it appears that the bottom water was not anoxic during entire interglacials but predominantly during terminations. 6.2. Pollen, TOC, carbonate content, d18Obulk

As outlined above, glacials and interglacials in Lake Van record were determined on the base of visual correlation between our proxies and reference curves. With the exception of d18Obulk, combined geochemical and pollen proxies point to a coherent scenario of glacial/interglacial changes with drier glacial and wetter interglacial conditions (Fig. 2). By ‘dry’ we understand conditions

Table 1 Mean values and ranges for d18O (‰ VPDB) of authigenic carbonates for different time slices within the high-resolution interval (0e60 mcblf) and part of the low-resolution interval (60e67.603 mcblf). YD interval has an anomalous lithology and isotopic signature; therefore we also provide values for TI after removing YD values (TI*). Interval

Depth (mcblf)

Depth (mcblf-nE)

Age range (ka)

d18Omin

d18Omax

d18Omean

Late Holocene Holocene TI* (without YD) TI Last Glacial Last Interglacial TII Penultimate glacial

0e2.288 0e4.605 5.945e6.591 4.605e6.591 6.591e32.028 32.028e56.905 56.905e61.275 61.275e67.603

0e2.180 0e4.443 5.623e6.196 4.443e6.196 6.196e26.548 26.548e45.989 45.989e49.151 49.151e55.871

0e4.33 0e11.65 12.85e14.64 11.65e14.64 14.64e70.00 70.00e128.40 128.40e136.00 136.00e147.00

0.29 1.74 1.18 1.18 1.59 1.9 4.25 0.50

1.87 1.87 2.11 4.46 1.64 5.75 1.54 1.57

0.56 0 0.21 1.6 0.2 0.83 0.72 0.56

O. Kwiecien et al. / Quaternary Science Reviews 104 (2014) 42e52 47

Fig. 3. Detailed comparison of the last four terminations, (a) TI, (b) TII, (c) TIII, (d) TIV. Note that all data are plotted along depth (mmcblf) and that the values measured in event layers (tephras and graded layers) are not shown. Lithotypes following Stockhecke et al. (2014a): (1) intercalation of green or brown fine-laminated clayey silts and grey graded layers, (2) brown fine-laminated clayey silts, analog to modern sediments, (3) yellowish to gray CaCO3-rich banded clayey silts. Note that TI (3a) has an exceptional lithological sequence with unique lithology during YD and recurring lithotype 2. Ages provided are serving as the tuning points for the age model (Stockhecke et al., 2014). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

48

O. Kwiecien et al. / Quaternary Science Reviews 104 (2014) 42e52

€l in central Turkey to put forward an important of lake Eski Acigo concept of hydrological variability within both glacials and interglacials. This work highlights the importance of considered time-scales; on the orbital time-scales glacials might be recognized as dryer in contrast to wetter interglacials, still, glaciation does not rule out millennial or multi-millennial variability, with cold and dry periods following cold and wet periods. In the following sections we provide possible explanations for Lake Van d18Obulk record and cautiously try to reconcile them with information provided by other proxy data. Although we favor one option, which at the moment seems to account for all aspects of sedimentation in Lake Van, and to explain the apparent problem with the classical interpretation of isotope data, we admit that this interpretation is tentative. When extended and subjected to scrutiny, isotope data might reveal more complex course of hydrologic changes than one proposed here. The resolution of d18Obulk record is not uniform along our profile and suboptimal for conclusive and detailed investigation of variability within glacials and interglacials. Therefore, in order to discuss shifts in the isotope record we first focus on the last 140 ka and then by testing our hypothesis in the context of other proxies we extrapolate our findings to older cycles. Within the period 140e14 ka Lake Van record displays shift from lighter (glacial) to heavier (interglacial) values of d18Obulk, which is the reverse of the isotopic trend characterizing the Late Pleistocene/Holocene transition in most other lakes from Mediterranean region (for review see Roberts et al., 2008). Several hypotheses can account for an opposite than an expected change in d18Obulk of Lake Van. Presence of detrital carbonate would modify isotopic composition of bulk carbonate (e.g. Brauer and Casanova, 2001; Leng et al., 2010). During the wet intervals, the dilution of background sedimentation (authigenic carbonate) by detrital material is at its lowest. Minimal admixture of detrital carbonate means that d18Obulk would reflect mostly the environmental signal of lake water at a given time, while higher admixture of detrital carbonate during dry intervals would dampen the signal and produce an offset toward d18O of detrital material. There is a potential source of detrital carbonate with a relative atay negative signature in the northern drainage of Lake Van (Çag et al., 2014) yet no significant detrital carbonate grains have been observed in the sediment. Although the detail mineralogical investigation of Lake Van carbonates is not yet concluded, previous results suggest presence of mixture of calcite and aragonite (Lemcke and Sturm, 1997). Precipitation experiments show that d18O of calcite is depleted by 0.5‰ relatively to aragonite (Kim and O'Neil, 1997). In order to explain apparently lighter glacial and apparently heavier interglacial d18O values we can theoretically assume systematic precipitation of lighter calcite during glacials and heavier aragonite during interglacials. Noteworthy, 0.5‰ offset cannot account for the entire amplitude of observed change (Fig. 2). Similarly, such an assumption conflicts information provided by other proxies; in lacustrine settings precipitation of authigenic aragonite is related to increased Mg/Ca ratios (Deocampo, 2010) which in turn are often an expression of increased salinity. Reflecting on the pollen and Ca/K data suggesting wetter interglacial conditions, increased interglacial salinity does not seem plausible. Another possibility to explain lighter glacial and heavier interglacial d18O invokes changes in isotopic composition of atmospheric precipitation due to the deglacial increase in temperature (Von Grafenstein et al., 1999). The isotopic signature of atmospheric precipitation would be heavier during interglacials and lighter during glacials. Such a relation has been observed in Central Europe (Von Grafenstein et al., 1999) and as far east as the Black Sea (Bahr et al., 2006) during the last deglaciation. Notably, the Black Sea

receives most of its freshwater input from European rivers. On the contrary, both marine and continental Eastern Mediterranean carbonate records show consequently lighter values during interglacials and heavier values during glacials (Rohling and Fenton, 1998; Frogley et al., 1999; Bar-Matthews et al., 2003; Stevens et al., 2012). Therefore in the continental Mediterranean domain heavier-to-lighter shift recorded by authigenic carbonates at the glacial/interglacial transition is explained as the source area effect (Roberts et al., 2008). Briefly, most of the moisture transported to Eastern Mediterranean is coming from Mediterranean Sea, which has a heavier isotopic composition during the glacials than during interglacials, due to the withdrawal of water locked in the ice sheets and additionally reduced freshwater input. Further on, Roberts et al. (2008) conclude that a pronounced source area effect suggests a negligible role of temperature changes on the isotopic composition of Eastern Mediterranean carbonates over glacial/ interglacial timescales. In any case, higher Lake Van surface temperature during interglacials cannot account for observed shift, as the temperature effect would drive the interglacial d18Obulk toward lighter values (Kim and O'Neil, 1997). We favor an alternative hypothesis, featuring a change in Lake Van recharge system, which would have an impact on the isotopic composition of Lake Van carbonates. Today Lake Van is fed mostly by river inflow and to some extent by snowmelt. The runoff peak comes in spring when the snowmelt and increased rainfall coincide. Meteorological data (IAEA, online database) clearly show that most depleted atmospheric precipitation in Lake Van region falls during winter months (JFM). Changes in the seasonality of precipitation influence its isotopic composition. The idea was explored by Stevens et al. (2001, 2006) who proposed shift toward winterdominated precipitation to reconcile light d18O of carbonates from two Iranian lakes (Zeribar and Mirabad) during early Holocene, depicted as dry by pollen records. Isotopic composition of snow reflects equilibrium conditions in the cloud and is even lighter than that of rain. Lake Van is located at the altitude of 1649 m a.s.l in the vicinity of several mountain glaciers (Akcar and Schlüchter, 2005). During glacials, when the altitude of the snow elevation line is lowered (Sarıkaya et al., 2008, 2009) not only winter but also portion of the spring precipitation would fall as snow. With most of the water trapped in snow or ice the level of glacial Lake Van would fall. Its water would be saline, well-mixed and isotopically heavy, corresponding to the concurrent Eastern Mediterranean signal. In spring significant input of freshwater (snowmelt) would form a lid on the lake surface, induce a strong but short-lived (seasonal) water column stratification and initiate precipitation of authigenic carbonates (analog to modern whitings). In that case isotopic composition of carbonates precipitating in the epilimnion would record a signature of light snowmelt rather than that of heavy saline lake water. During terminations not only annual snowmelt but a significant amount of glacier meltwater reaching Lake Van would raise the lake level and drive d18Obulk toward even more negative values. During the interglacials diminishing contribution of snow and increasing contribution of rain to regional precipitation would lead to isotopically heavier runoff, and at the same time would reduce the salinity and keep the lake level high. With reduced salinity inflowing freshwater would faster mix with the bulk of the lake water. The isotopic signature of precipitating carbonate would be heavier and closer to that of evaporation-influenced and mixed lake water than that of freshwater inflow. Conceptually, the idea is similar to the one presented by Torfstein et al. (2009) accounting for lighter isotopic composition of carbonates during the times of low stands of Lake Amora, the Dead Sea precursor. Dean et al. (2013) evoked a similar mechanism and a freshwater lid hypothesis explaining seasonal offsets between d18O signature of diatom growing in spring and carbonate

O. Kwiecien et al. / Quaternary Science Reviews 104 (2014) 42e52

forming in summer. Attempting to test the possibility of modulation of depositional conditions in Lake Van by snowmelt and meltwater inflow we performed a detailed analysis of the last four glacial terminations (Fig. 3aed). 7.3. Anatomy of termination Our comprehensive comparison of the last four terminations (Fig. 3aed) displays the same sequence of lithotypes, each representing different depositional environment (Stockhecke et al., 2014a). These include, from bottom to top: (1) intercalation of green or brown fine-laminated clayey silts and gray graded layers, (2) brown finely laminated clayey silts, analog to modern sediments, (3) yellowish to gray CaCO3-rich banded silts (Fig. 3). Geochemical proxies show small variability in pattern and different amplitudes of change during each termination but, within the three mentioned lithotypes they point to systematic changes in depositional environment. Intercalations (1) are related to rising lake level, low productivity and increased remobilization of sediment, and contain low amount of arboreal pollen. This lithotype is associated with deglacial conditions. Brown laminated silts (2) were deposited during intervals of rising or high lake level, high

49

productivity and diminishing detrital input. Geochemistry of yellowish to gray, CaCO3-rich banded silts (3) suggests lowering of the lake level, high productivity and very low detrital supply. Lithotypes 2 and 3 contain high amount of arboreal pollen and are associated with interglacial conditions. The transition between lithotypes 2 and 3 has a very sharp character (Fig. 3aed). Termination I (Fig. 3a) with the already mentioned YD cold spell, stands out in comparison with Terminations IIeIV. YD sediments are CaCO3- and organic-poor and almost annually intercalated with less than mm-thick event deposits (Stockhecke et al., 2014a). Generally YD intercalations are more frequent but also thinner than that of the typical lithotype 1. Also in contrast to lithotype 1, YD sediments reflect lowering rather than rising lake level (Stockhecke et al., 2014a). It is conceivable that, as previously suggested (Landmann et al., 1996; Lemcke and Sturm, 1997; Wick et al., 2003) the YD was an exceptionally arid interval with very low runoff. In that case reduced volume of snowmelt and rainfall would be not sufficient to produce isotopically depleted freshwater lens on the lake surface and precipitating carbonates would record isotopic composition of evaporation-influenced lake water. Throughout the whole Lake Van record the YD seems to be the only such an anomalous interval.

Fig. 4. Proxy-based schematic model of most important processes driving sedimentation in Lake Van at the glacial/interglacial transition. The behavior of d18O of carbonate is based only on the proxy record during Termination I and II.

50

O. Kwiecien et al. / Quaternary Science Reviews 104 (2014) 42e52

7.4. Reconstructed course of termination Fig. 4 presents a synthetized schematic course of glacial termination in Lake Van. Let us consider proposed scenario of glacial precipitation falling mostly in the form of snow. If mountain glaciers build up on the flanks of Lake Van, most of the moisture will be trapped as snow or ice leading to a significant lowering of the lake level and consequently a saline, well-mixed water column. Lower temperatures and reduced nutrient supply would inhibit primary productivity in the lake. During spring/summer snowmelt, the d18O of runoff reaching Lake Van would be significantly depleted. Once the annual snowmelt reaches the lake, precipitation of carbonate will take place upon inflow (in analogy to modern whitings) and the isotopic composition of resulting carbonate will record the signature of depleted freshwater lid. Exposed slopes and sparse vegetation cover allow for increased erosion and finegrained detrital input. Enhanced supply of detrital material (of fluvial and eolian origin) dilutes the background carbonate sedimentation. As the bottom water is oxic, no clear laminations are preserved. During deglacial warming, increasing meltwater supply and a change from snow- to rainfall would raise the lake level promoting slope instability and contributing to more frequent deposition of event layers resulting in the intercalations (1) observed in the sedimentary record (Fig. 3aed). Raising lake level would promote the density stratification of the water column and result in the onset of anoxic conditions and better preservation of organic matter at the lake bottom. Neither TOC nor Mn/Fe clearly point to anoxic conditions, but preservation of laminae and color of intercalations changing toward brown suggests suboxic conditions at the sediment water interface. With precipitation shifting from snow to rain, advancing vegetation cover on land would reduce the amount of detrital input. An enhanced runoff and a further warming would increase primary productivity in the lake. This would in turn support biologically triggered mode of carbonate precipitation. Anoxic conditions at the lake bottom would permit the deposition of carbonate-rich, organic-rich brown finely laminated silty clays (2) (Fig. 3aed). Interestingly, comparison of Mn/Fe, TOC and Si/K suggest that increase in productivity lagged the onset of anoxic conditions, but productivity stayed high after mixing of deep-water was restored. At the peak interglacial conditions, the lake would be fed mostly by rainfall. High summer temperatures could strengthen evaporation and by increasing density of surface water re-establish overturning of the lake. Vegetation on land would keep detrital input low, while primary productivity and carbonate precipitation would stay high. Organic matter/carbonate couplets could not be preserved due to the well-mixed, oxygenated water column and a deposition of carbonate-rich, organic-poor banded silts (3) would follow. 7.5. Glacial versus interglacial conditions and broader implications The comparison of Lake Van pollen, geochemical proxies and lithological analysis suggests cold glacials dominated by snowfall and warm interglacials with prevailing rainfall precipitation. In that context, correlation between the carbonate content and d18Obulkdeduced atmospheric precipitation type suggest that more carbonate was formed in the lake during rainfall-dominated intervals. In turn, this situation implies reduced glacial runoff (weak snowfall or weak snowmelt due to colder summers) and increased interglacial runoff (strong rainfall) advocating overall dry glacial and wet interglacial conditions. Our interpretation agrees with studies of Mediterranean pollen and stalagmites (Bar-Matthews et al., 2000; Tzedakis et al., 2006; Fleitmann et al., 2009) but is less consistent with reconstructions of Turkish glaciers (Sarıkaya et al., 2008, 2009

and references therein) suggesting cold and wet glacials (at least LGM). Notably, although resolution of our d18Obulk data is as yet insufficient for analysis of millennial-variability, all other proxies consequently point to succession of relatively drier (stadials) and relatively wetter (interstadials) intervals within the cold glacial periods. It seems that one of the reasons for contrasting hydroclimate conditions reconstructed for Lake Van and Dead Sea basins (eg. Bartov et al., 2003; Torfstein et al., 2009; Waldmann et al., 2009) over glacial/interglacial cycles, is a different altitude of lakes. One important finding related to the altitude and topographic setting is, that in contrast to most of the Mediterranean lakes, hydrology of Lake Van was susceptible to atmospheric temperature changes. We propose, that a transition from snowmelt-fed to rainwater-fed runoff is one of the drivers controlling lake level of Lake Van as well as oxygen isotopic composition of its carbonates. Details of this concept not only supplement the general framework proposed by Stockhecke et al. (2014a) but also provide a plausible mechanism explaining observed changes in lithology and geochemistry of the sediment and pollen data. This study emphasizes the essential role of the local/regional settings in the response of given site to changing climate. 8. Conclusions Based on a comparison of multi-proxy sediment core records from Lake Van, we compiled a comprehensive overview of the climate evolution in the lake region over the last 360 ka, with a special focus on the glacial terminations. On the orbital time-scales the picture of cold and dry glacials and warm and wet interglacials emerging from pollen, organic carbon, authigenic carbonate content, elemental profiling by XRF and lithological analyses is inconsistent with classical interpretation of oxygen isotopic composition of carbonates pointing to wet glacial and dry interglacial conditions for the period 140-14 ka. In order to solve this conundrum we compared available records of the four last terminations and developed a depositional model explaining different patterns observed in all the proxies. We hypothesize that the relative contribution of rain, snow and meltwater recharging the basin plays an important role for hydrological conditions, producing and sustaining bottom-water anoxia and influencing isotopic composition of carbonates. During the glacial, Lake Van is generally well ventilated although isotopically depleted snowmelt may form a freshwater lid causing short-lived, seasonal stratification. Authigenic carbonates precipitating in the epilimnion will record light isotopic composition of runoff. During terminations, increased rainfall and snowmelt will enhance runoff and as a consequence rise lake level leading to bottom-water anoxia. During interglacials, after the meltwater supply is reduced, lake mixing is restored and precipitating carbonates will show signature of evaporated and mixed lake water, isotopically heavier than runoff representing snowmelt and atmospheric precipitation. Systematic pattern of the four last terminations implies that the YD was an exceptional event with no analog in the older Lake Van record. This study emphasized the need of correct assessment of regional situation and deep understanding of the sites proposed for paleoclimatological studies. Acknowledgments We thank the ICDP PALEOVAN project team for support during collection and sharing of data. The authors acknowledge funding of the PALEOVAN drilling campaign by the International Continental Scientific Drilling Program (ICDP), the Deutsche Forschungsgemeinschaft (DFG), the Swiss National Science Foundation (SNF) (200021_124981 and 200020_143330) and the Scientific and Technological Research Council of Turkey (Tübitak). Vera Lukies

O. Kwiecien et al. / Quaternary Science Reviews 104 (2014) 42e52

€ hl from the IODP Core Repository in Bremen provided and Ulla Ro valuable help during the XFR scanning campaign. We thank editor, Neil Roberts, Matthew Jones and an anonymous reviewer from whose insightful comments our manuscript greatly benefited.

References Akcar, N., Schlüchter, C., 2005. Paleoglaciations in Anatolia: a schematic review and first results. Eiszeitalt. Gegen. 55, 102e121. €tzold, J., Wefer, G., 1998. Correlated millennial-scale changes in surface Arz, H.W., Pa Hydrography and terrigenous sediment yield inferred from Last-Glacial Marine deposits off Northeastern Brazil. Quat. Res. 50, 157e166. Bahr, A., Arz, H.W., Lamy, F., Wefer, G., 2006. Late glacial to Holocene paleoenvironmental evolution of the Black Sea, reconstructed with stable oxygen isotope records obtained on ostracod shells. Earth Planet. Sci. Lett. 241, 863e875. Barker, S., Knorr, G., Edwards, R.L., Parrenin, F., Putnam, A.E., Skinner, L.C., Wolff, E., Ziegler, M., 2011. 800,000 Years of abrupt climate variability. Science 334, 347e351. Bar-Matthews, M., Ayalon, A., Gilmour, M., Matthews, A., Hawkesworth, Chris J., 2003. Sea e land oxygen isotopic relationships from planktonic foraminifera and speleothems in the Eastern Mediterranean region and their implication for paleorainfall during interglacial intervals. Geochim. Cosmochim. Acta 67, 3181e3199. Bar-Matthews, M., Ayalon, A., Kaufman, A., 2000. Timing and hydrological conditions of Sapropel events in the Eastern Mediterranean, as evident from speleothems, Soreq cave. Isr. Chem. Geol. 169, 145e156. Bartov, Y., Goldstein, S.L., Stein, M., Enzel, Y., 2003. Catastrophic arid episodes in the Eastern Mediterranean linked with the North Atlantic Heinrich events. Geology 31, 439e442. Berglund, B.E., Ralska-Jasiewiczowa, M., 1986. Pollen Analysis and Pollen Diagrams. Wiley & Sons. Brauer, A., Casanova, J., 2001. Chronology and depositional processes of the laminated sediment record from Lac d' Annecy, French Alps. J. Paleolimnol. 25, 163e177. Brauer, A., Mangili, C., Moscariello, A., Witt, A., 2008. Palaeoclimatic implications from micro-facies data of a 5900 varve time series from the Pi anico interglacial sediment record, southern Alps. Palaeogeogr. Palaeoclimatol. Palaeoecol. 259, 121e135. Cheng, H., Edwards, R.L., Broecker, W.S., Denton, G.H., Kong, X., Wang, Y., Zhang, R., Wang, X., 2009. Ice age terminations. Science 326, 248e252. atay, N., Ogretmen, N., Damci, E., Stockhecke, M., Sancar, Ü., Eris, K.K., Ozeren, S., Çag 2014. Lake level and climate records of the last 90 ka from the Northern Basin of Lake Van, eastern Turkey. Quat. Sci. Rev. 104, 97e116. , E., Imren, C., Niessen, F., Cukur, D., Krastel, S., Demirel-Schlüter, F., Demirbag Toker, M., 2012. Sedimentary evolution of Lake Van (Eastern Turkey) reconstructed from high-resolution seismic investigations. Int. J. Earth Sci. 102, 571e585. Dean, J.R., Jones, M.D., Leng, M.J., Sloane, H.J., Roberts, C.N., Woodbridge, J., itbas¸ıog lu, H., 2013. PalaeoSwann, G.E.A., Metcalfe, S.E., Eastwood, W.J., Yig seasonality of the last two millennia reconstructed from the oxygen isotope € lü, central Turkey. composition of carbonates and diatom silica from Nar Go Quat. Sci. Rev. 66, 35e44. Denton, G.H., Anderson, R.F., Toggweiler, J.R., Edwards, R.L., Schaefer, J.M., Putnam, A.E., 2010. The last glacial termination. Science 328, 1652e1656. Deocampo, D., 2010. The geochemistry of continental carbonates. Dev. Sedimentol. 62, 1e59. Develle, A.-L., Gasse, F., Vidal, L., Williamson, D., Demory, F., Van Campo, E., Ghaleb, B., Thouveny, N., 2011. A 250ka sedimentary record from a small karstic lake in the Northern Levant (Yammoûneh, Lebanon). Palaeogeogr. Palaeoclimatol. Palaeoecol. 305, 10e27. pez-Vinyallonga, S., Djamali, M., Baumel, A., Brewer, S., Jackson, S.T., Kadereit, J.W., Lo Mehregan, I., Shabanian, E., Simakova, A., 2012. Ecological implications of Cousinia Cass. (Asteraceae) persistence through the last two glacialeinterglacial cycles in the continental Middle East for the Irano-Turanian flora. Rev. Palaeobot. Palynol. 172, 10e20. Faegri, K., Iversen, J., 1989. Texbook of Pollen Analysis. Wiley & Sons. €ktürk, O.M., Fleitmann, D., Cheng, H., Badertscher, S., Edwards, R.L., Mudelsee, M., Go Fankhauser, A., Pickering, R., Raible, C.C., Matter, A., Kramers, J., Tüysüz, O., 2009. Timing and climatic impact of Greenland interstadials recorded in stalagmites from northern Turkey. Geophys. Res. Lett. 36, L19707. Frogley, M.R., Tzedakis, P.C., Heaton, T.H.E., 1999. Climate variability in Northwest Greece during the last interglacial. Science 285, 1886e1889. Gasse, F., Vidal, L., Develle, A.-L., Van Campo, E., 2011. Hydrological variability in the Northern Levant: a 250 ka multiproxy record from the Yammoûneh (Lebanon) sedimentary sequence. Clim. Past 7, 1261e1284. Hays, J.D., Imbrie, J., Shackleton, N.J., 1976. Variation in the Earth's Orbit: pacemaker of the Ice Ages. Science 4, 1121e1132. IAEA, n.d. http://www-naweb.iaea.org/napc/ih/documents/userupdate/Waterloo/ index.html#Asia (WWW Document). Imbrie, J., Imbrie, K.P., 1979. Ice Ages: Solving the Mystery. Enslow Publishers, Short Hills NJ, ISBN 978-0-89490-015-0.

51

€hl, U., 2005. Jaccard, S.L., Haug, G.H., Sigman, D.M., Pedersen, T.F., Thierstein, H.R., Ro Glacial/interglacial changes in subarctic north pacific stratification. Science 308, 1003e1006. Jansen, J.H.F., Van der Gaast, S.J., Koster, B., Vaars, A.J., 1998. Short communication CORTEX, a shipboard XRF-scanner for element analyses in split sediment cores. Mar. Geol. 151, 143e153. Jones, M.D., Roberts, C.N., Leng, M.J., 2007. Quantifying climatic change through the last glacialeinterglacial transition based on lake isotope palaeohydrology from central Turkey. Quat. Res. 67, 463e473. Jones, M.D., Roberts, N.C., 2008. Interpreting lake isotope records of Holocene environmental change in the Eastern Mediterranean. Quat. Int. 181, 32e38. Jouzel, J., Masson-Delmotte, V., Cattani, O., Dreyfus, G., Falourd, S., Hoffmann, G., Minster, B., Nouet, J., Barnola, J.M., Chappellaz, J., Fischer, H., Gallet, J.C., Johnsen, S., Leuenberger, M., Loulergue, L., Luethi, D., Oerter, H., Parrenin, F., Raisbeck, G., Raynaud, D., Schilt, A., Schwander, J., Selmo, E., Souchez, R., Spahni, R., Stauffer, B., Steffensen, J.P., Stenni, B., Stocker, T.F., Tison, J.L., Werner, M., Wolff, E.W., 2007. Orbital and millennial Antarctic climate variability over the past 800,000 years. Science 317, 793e796. Kaden, H., Peeters, F., Lorke, A., Kipfer, R., Tomonaga, Y., Karabiyikoglu, M., 2010. Impact of lake level change on deep-water renewal and oxic conditions in deep saline Lake Van, Turkey. Water Resour. Res. 46. Kim, S.-T., O'Neil, J.R., 1997. Equilibrium and nonequilibrium oxygen isotope effects in synthetic carbonates. Geochim. Cosmochim. Acta 61, 3461e3475. lu, C., Christol, A., Mouralis, D., Dog u, A.-F., Akko €prü, E., Fort, M., Kuzucuog lu, M., Scaillet, S., Reyss, J.-L., Brunstein, D., Zorer, H., Fontugne, M., Karabiyikog Guillou, H., 2010. Formation of the Upper Pleistocene terraces of Lake Van (Turkey). J. Quat. Sci. 25, 1124e1137. Kwiecien, O., Arz, H.W., Lamy, F., Plessen, B., Bahr, A., Haug, G.H., 2009. North Atlantic control on precipitation pattern in the eastern Mediterranean/Black Sea region during the last glacial. Quat. Res. 71, 375e384. Landmann, G., Reimer, A., Lemcke, G., Kempe, S., 1996. Dating Late Glacial abrupt climate changes in the 14,570 yr long continuous varve record of Lake Van, Turkey. Palaeogeogr. Palaeoclimatol. Palaeoecol. 122, 107e118. Laskar, J., Robutel, P., Joutel, F., Gastineau, M., Correia, A., Levrard, B., 2004. A longterm numerical solution for the insolation quantities of the earth. Astron. Astrophys. 428, 261e285. Lemcke, G., Sturm, M., 1997. d18O and Trace element measurements as proxy for the reconstruction of climate change at Lake Van (Turkey): preliminary results. NATO AS1 Ser. 149, 653e678. €oklimarekonstruktion am Van See. (Ostanatolien, Türkei). Lemcke, G., 1996. Pala Swiss Federal Institute of Technology, Zurich. ETH, PhD Thesis no. 11786. Leng, M.J., Jones, M.D., Frogley, M.R., Eastwood, W.J., Kendrick, C.P., Roberts, C.N., 2010. Detrital carbonate influences on bulk oxygen and carbon isotope composition of lacustrine sediments from the Mediterranean. Glob. Planet. Change 71, 175e182. Leng, M.J., Marshall, J.D., 2004. Palaeoclimate interpretation of stable isotope data from lake sediment archives. Quat. Sci. Rev. 23, 811e831. atay, N., Cukur, D., Damci, E., Litt, T., Anselmetti, F.S., Baumgarten, H., Beer, J., Çag Glombitza, C., Haug, G., Heumann, G., Kallmeyer, J., Kipfer, R., Krastel, S., Kwiecien, O., Meydan, A.F., Orcen, S., Pickarski, N., Randlett, M.-E., Schmincke, H.U., Schubert, C., Sturm, M., Sumita, M., Stockhecke, M., Tomonaga, Y., Vigliotti, L., Wonik, T., Team, the P.S., 2012. 500,000 Years of environmental history in Eastern Anatolia: the PALEOVAN drilling project. Sci. Drill., 18e29. € Litt, T., Krastel, S., Sturm, M., Kipfer, R., Orcen, S., Heumann, G., Franz, S.O., Ülgen, U.B., Niessen, F., 2009. “PALEOVAN”, International Continental Scientific Drilling Program (ICDP): site survey results and perspectives. Quat. Sci. Rev. 28, 1555e1567. Litt, T., Pickarski, N., Heumann, G., Stockhecke, M., Tzedakis, P.C., 2014. A 600,000 year long continental pollen record from Lake Van, Eastern Anatolia (Turkey). Quat. Sci. Rev. 104, 30e41. http://dx.doi.org/10.1016/j.quascirev.2014.03.017. Marchitto, T.M., Lehman, S.J., Ortiz, J.D., Flückiger, J., van Geen, A., 2007. Marine radiocarbon evidence for the mechanism of deglacial atmospheric CO2 rise. Science 316, 1456e1459. Martrat, B., Grimalt, J.O., Shackleton, N.J., de Abreu, L., Hutterli, M.A., Stocker, T.F., 2007. Four climate cycles of recurring deep and surface water destabilizations on the Iberian margin. Science 317, 502e507. McManus, J.F., Francois, R., Gherardi, J.-M., Keigwin, L.D., Brown-Leger, S., 2004. Collapse and rapid resumption of Atlantic meridional circulation linked to deglacial climate changes. Nature 428, 834e837. McManus, J.F., Oppo, D.W., Cullen, J.L., 1999. A 0.5-Million-Year record of millennialscale climate variability in the North Atlantic. Science 283, 971e975. Meckler, A.N., Clarkson, M.O., Cobb, K.M., Sodemann, H., Adkins, J.F., 2012. Interglacial hydroclimate in the tropical West Pacific through the Late Pleistocene. Science 336, 1301e1304. Meckler, A.N., Sigman, D.M., Gibson, K.A., François, R., Martínez-García, A., €hl, U., Peterson, L.C., Tiedemann, R., Haug, G.H., 2013. Deglacial Jaccard, S.L., Ro pulses of deep-ocean silicate into the subtropical North Atlantic Ocean. Nature 495, 495e498. Melles, M., Brigham-Grette, J., Minyuk, P.S., Nowaczyk, N.R., Wennrich, V., DeConto, R.M., Anderson, P.M., Andreev, A.A., Coletti, A., Cook, T.L., Haltian, P., Tarasov, P., Voge, H., Wagner, B., Hovi, E., Kukkonen, M., Lozhkin, A.V., Rose 2012. 2.8 million years of Arctic climate change from Lake El'gygytgyn, NE Russia. Science 337, 315e320. NGRIP project members, Andersen, K.K., Azuma, N., Barnola, J.-M., Bigler, M., Biscaye, P., Caillon, N., Chappellaz, J., Clausen, H.B., Dahl-Jensen, D., Fischer, H.,

52

O. Kwiecien et al. / Quaternary Science Reviews 104 (2014) 42e52

Flückiger, J., Fritzsche, D., Fujii, Y., Goto-Azuma, K., Grønvold, K., Gundestrup, N.S., Hansson, M., Huber, C., Hvidberg, C.S., Johnsen, S.J., Jonsell, U., Jouzel, J., Kipfstuhl, S., Landais, A., Leuenberger, M., Lorrain, R., Masson-Delmotte, V., Miller, H., Motoyama, H., Narita, H., Popp, T., Rasmussen, S.O., Raynaud, D., Rothlisberger, R., Ruth, U., Samyn, D., Schwander, J., Shoji, H., Siggard-Andersen, M.-L., Steffensen, J.P., Stocker, T., € rnsdo ttir, A.E., Svensson, A., Takata, M., Tison, J.-L., Thorsteinsson, T., Sveinbjo Watanabe, O., Wilhelms, F., White, J.W.C., 2004. High-resolution record of Northern Hemisphere climate extending into the last interglacial period. Nature 431, 147e151. Peeters, F., Kipfer, R., 2009. Currents in stratified water bodies. Density-driven flows. Encycl. Inl. Waters, 530e538. Prokopenko, A.A., Hinnov, L.A., Williams, D.F., Kuzmin, M.I., 2006. Orbital forcing of continental climate during the Pleistocene: a complete astronomically tuned climatic record from Lake Baikal, SE Siberia. Quat. Sci. Rev. 25, 3431e3457. Reimer, A., Landmann, G., Kempe, S., 2008. Lake Van, Eastern Anatolia, hydrochemistry and history. Aquat. Geochem. 15, 195e222. Roberts, N., Jones, M.D., Benkaddour, a., Eastwood, W.J., Filippi, M.L., Frogley, M.R., s, B., Lamb, H.F., Leng, M.J., Reed, J.M., Stein, M., Stevens, L., Valero-Garce Zanchetta, G., 2008. Stable isotope records of Late Quaternary climate and hydrology from Mediterranean lakes: the ISOMED synthesis. Quat. Sci. Rev. 27, 2426e2441. € hl, U., Abrams, L.J., 2000. High-resolution, downhole, and nondestructive core Ro measurements from sites 999 and 1001 in the Caribbean Sea: application to the Late Paleocene Thermal Maximum. Proc. ODP Sci. Results 165, 191e203. Rohling, E.J., Fenton, M., 1998. Magnitudes of sea-level lowstands of the past 500,000 years. Nature 394, 162e165. Ruddiman, W.F., 2003. Orbital insolation, ice volume, and greenhouse gases. Quat. Sci. Rev. 22, 1597e1629. Sarıkaya, M.A., Zreda, M., Çiner, A., 2009. Glaciations and paleoclimate of Mount Erciyes, central Turkey, since the Last Glacial Maximum, inferred from 36Cl cosmogenic dating and glacier modeling. Quat. Sci. Rev. 28, 2326e2341. Sarıkaya, M.A., Zreda, M., Çiner, A., Zweck, C., 2008. Cold and wet Last Glacial Maximum on Mount Sandıras, SW Turkey, inferred from cosmogenic dating and glacier modeling. Quat. Sci. Rev. 27, 769e780. Schweizer, G., 1975. Untersuchungen zur Physiographie von Ostanatolien und Nordwestiran. Tübinger Geogr. Stud. 60, 9. Shackleton, N.J., Opdyke, N.D., 1973. Oxygen isotope and Palaeomagnetic stratigraphy of equatorial Pacific core V28-238: oxygen isotope temperatures and Ice Volumes on a 105 Year and 106 Year Scale. Quat. Res. 55, 39e55. Steffensen, J.P., Andersen, K.K., Bigler, M., Clausen, H.B., Dahl-Jensen, D., Fischer, H., Goto-Azuma, K., Hansson, M., Johnsen, S.J., Jouzel, J., Masson-Delmotte, V., € thlisberger, R., Ruth, U., Stauffer, B., SiggaardPopp, T., Rasmussen, S.O., Ro €rnsdo  ttir, A.E., Svensson, A., White, J.W.C., 2008. HighAndersen, M.-L., Sveinbjo resolution Greenland ice core data show abrupt climate change happens in few years. Science 321, 680e684. Stevens, L.R., Djamali, M., Andrieu-Ponel, V., Beaulieu, J.-L., 2012. Hydroclimatic variations over the last two glacial/interglacial cycles at Lake Urmia. Iran. J. Paleolimnol. 47, 645e660.

Stevens, L.R., Ito, E., Schwalb, A., Wright, H.E., 2006. Timing of atmospheric precipitation in the Zagros Mountains inferred from a multi-proxy record from Lake Mirabad, Iran. Quat. Res. 66, 494e500. Stevens, L.R., Wright Jr., H.E., Ito, E., 2001. Proposed changes in seasonality of climate during the Lateglacial and Holocene at Lake Zeribar, Iran. Holocene 11, 747e755. Stockhecke, M., Anselmetti, F.S., Meydan, A.F., Odermatt, D., Sturm, M., 2012. The annual particle cycle in Lake Van (Turkey). Palaeogeogr. Palaeoclimatol. Palaeoecol. 333e334, 148e159. Stockhecke, M., Sturm, M., Brunner, I., Schmincke, H.-U., Sumita, M., Kipfer, R., Cukur, D., Kwiecien, O., Anselmetti, F.S., 2014a. Sedimentary evolution and environmental history of Lake Van (Turkey) over the past 600 000 years. Sedimentology. http://dx.doi.org/10.1111/sed.12118 (in press). atay, N., Stockhecke, M., Kwiecien, O., Vigliotti, L., Anselmetti, F.S., Beer, J., Çag Channell, J.E., Kipfer, R., Lachner, J., Litt, T., Pickarski, N., Sturm, M., 2014b. Chronostratigraphy of the 600,000 year old continental record of Lake Van (Turkey). Quat. Sci. Rev. 104, 8e17. Svensson, A., Andersen, K.K., Bigler, M., Clausen, H.B., Dahl-Jensen, D., Davies, S.M., € thlisberger, R., Johnsen, S.J., Muscheler, R., Parrenin, F., Rasmussen, S.O., Ro Seierstad, I., Steffensen, J.P., Vinther, B.M., 2008. A 60 000 year Greenland stratigraphic ice core chronology. Clim. Past 4, 47e57. Torfstein, A., Haase-Schramm, A., Waldmann, N., Kolodny, Y., Stein, M., 2009. Useries and oxygen isotope chronology of the mid-Pleistocene Lake Amora (Dead Sea basin). Geochim. Cosmochim. Acta 73, 2603e2630. Tzedakis, P.C., Bennett, K.D., 1996. Interglacial vegetation succession: a view from Southern Europe. Quat. Sci. Rev. 14, 967e982. €like, H., 2006. The last 1.35 million years at Tzedakis, P.C., Hooghiemstra, H., Pa Tenaghi Philippon: revised chronostratigraphy and long-term vegetation trends. Quat. Sci. Rev. 25, 3416e3430. Von Grafenstein, U., Erlenkeuser, H., Brauer, A., Jouzel, J., Johnsen, S., 1999. A midEuropean decadal isotope-climate record from 15,500 to 5000 years B.P. Science 284, 1654e1657. Waldmann, N., Stein, M., Ariztegui, D., Starinsky, A., 2009. Stratigraphy, depositional environments and level reconstruction of the last interglacial Lake Samra in the Dead Sea basin. Quat. Res. 72, 1e15. Wick, L., Lemcke, G., Sturm, M., 2003. Evidence of Lateglacial and Holocene climatic change and human impact in eastern Anatolia: high-resolution pollen, charcoal, isotopic and geochemical records from the laminated sediments of Lake Van, Turkey. Holocene 13, 665e675. Wigley, T.M.L., Farmer, G., 1982. Climate of the Eastern Mediterranean and the near east. In: Bintliff, J.L., et al. (Eds.), Paleoclimates, Paleoenvironments and Human Communities in the Eastern Mediterranean Region in Later Prehistory, Oxford. Oxford Press, pp. 3e37. € hl, U., Wilhelms, F., Vogt, C., Hanebuth, T.J.J., Wilhelms-Dick, D., Westerhold, T., Ro €mmermann, H., Kriews, M., Kasten, S., 2012. A comparison of mm scale Ro resolution techniques for element analysis in sediment cores. J. Anal. Atom. Spectrom. 27, 1574. Wolff, E.W., Chappellaz, J., Blunier, T., Rasmussen, S.O., Svensson, A., 2010. Millennial-scale variability during the last glacial: the ice core record. Quat. Sci. Rev. 29, 2828e2838.