Climatic Evolution of the Marmara Region during the past 50,000 years

The Marmara region (Western Turkey) is expected to have been one of the principal key areas in the transcontinental dispersal of modern humans from the Near East to the Balkans. Two prominent time-slices are of particular interest in this area, firstly concerning the primary dispersal of modern humans and the emergence of upper Palaeolithic adaptation (50,000-30,000 a BP) and secondly dealing with the reestablishment of human habitats after the second glacial maximum and the dispersal of Neolithic economy (15,000-7,000 a BP).

This project is targeted on reconstructing the climatic and environmental history of the Marmara region since 50 ka. For this purpose, first long sediment cores will be recovered from Lake Iznik. According to pilot studies, their sedimentological, geochemical, and biological (pollen, ostracods) composition will reflect the climatic and environmental development with a high sensitivity, thus promising to enhance the understanding of environmental controlling factors for human migration from Anatolia to the Balkan Peninsula.

Comparative paleoclimate studies focusing on the circum-Mediterranean clearly have demonstrated that the climate shifts in the Last Glacial-Interglacial transition were dominated by precipitation rather than temperature changes in the Eastern Mediterranean (e.g., Roberts et al. 2008). In order to test the potential for a more detailed understanding of the climate history in western Turkey, elemental pre-site surveys for the CRC project B4 were accomplished on Lake Iznik, the biggest lake in the Marmara region, in 2002 and 2005. Field work included hydroacoustic campaigns and sampling of a coarse grid of short sediment cores as well as two 5.5 meters long sediment cores. The information from more than 200 km of single channel airgun (15 to 40 cu.inch) and sediment-echo sounding (3.5 kHz and chirp) profiles allowed the selection of new coring locations in undisturbed sediment sequences to obtain a continuous, high-resolution, and more long-term record for paleoenvironmental and paleoclimate reconstructions. Furthermore, geochemical data from the 5.5 meters long cores clearly demonstrated that Lake Iznik was sensitive to changes of climate during the last 5 ka. Precipitation and evaporation patterns are possibly linked to AO/NAO variations, which considerably affect climate in the Eastern Mediterranean (i.e., Turkey and the Middle East; Franz unpubl.). Hence, it was proven that Lake Iznik is a key site to reconstruct the environmental and climatic history of the Marmara Region (W Turkey), which is by itself a principal area in the transcontinental dispersal of modern humans from the Near East to the Balkans.

The aim of project B4 on Lake Iznik was to retrieve a continuous sediment record of the past ca. 50 ka. A wide range of laboratory analyses was intended to be employed in order to decipher the environmental changes that have occurred in the region in dependence on climatic and human impact. Paleoenvironmental reconstructions were planned to cover both long-term (orbital time scale) and short-term (sub-millennial time scale) fluctuations. The latter aimed at deciphering the regional expression of atmospheric oscillations (NAO, AO, ENSO), sunspot cycles, and short-term events such as the D/O events, the 8.2 ka event, the Younger Dryas, and the Pleistocene- Holocene transition. Besides, it was intended to compare the reconstructed climatic and environmental changes from the lake sediments with information available on human settlement and migration in the region.

At the beginning of the CRC funding period, we completed sedimentological and faunal analysis of 260 subsamples from the sediment cores recovered in 2005 (LC1/2005 & LC2/2005), which continuously cover the past 4.7 ka cal BP in a time resolution of 47 resp. 25 years. Analyses of ostracods, mineralogy, major and trace elements, carbon, nitrogen, sulfur, total organic carbon, and loss on ignition provided comprehensive paleoenvironmental information, for instance concerning variations in humidity and aridity as well as lake-level changes (Ülgen et al. 2012, Viehberg et al. accepted). An oscillating trend of humid periods, interrupted by rapid climate change to relatively arid periods was identified. Distinct lithological and geochemical changes at 4.2 and 3.3 ka cal BP are traced back to intense droughts, which led to extreme, short-term lowering of the lake level.

New field work on Lake Iznik within the scope of the CRC included limnological investigations to obtain a comprehensive understanding of the lake´s (eco-)system. Respective field campaigns during October 2009 (autumn), July/August 2010 (summer), December 2010 (winter), and April 2011 (spring) covered the seasonal variability of the lake's hydrology and faunal distribution. Lake Iznik is an alkaline lake, which mixes once a year during springtime. Anaerobic bottom water in the northern and southern basins is formed in warm seasons, when the thermocline is established between 15 m and 20 m water depth. A dominant whitening event (massive calcium carbonate precipitation) occurred during the summer campaign in 2010. This enabled us to visually trace surface water currents in the lake and to better understand accumulation patterns found in geochemical proxies of the surface sediments, which were determined at the University of Cologne (Viehberg et al. 2012). The modern diverse ostracod and cladoceran fauna of Lake Iznik is a typical Mediterranean fauna with 41 species (27 ostracods and 14 cladocerans), including halotolerant resp. halophilic (e.g., Tyrrhenocythere amnicola, Loxoconchissa immodulata, Heterocypris salina) (Mesquita-Joanes et al. 2012, in press). In contrast to Nazik et al. (2010), we did not recover any marine species in Lake Iznik. This is probably due to contaminated sample material investigated by these authors (Yaltirak et al. 2012).

The CRC coring campaign on Lake Iznik aiming at older sediments was carried out in September/October 2009. Due to unforeseen disputes about visa and permit responsibilities between Turkish  authorities among regional officers (i.e., Ministry of Environment and Forestry, Ministry of Foreign Affairs, Cultural office and Turkish Gendarmerie), the start of field work became delayed by one week, which led to respective shorting of the campaign. For this reason, we decided to core only consecutive instead of overlapping segments from two sites. Major intention was to gain sufficiently old sediment sequences to meet the proposed time frame. In total, we retrieved 46 core meters using a percussion piston corer from a floating platform (UWITEC, Equipment University of Bonn), encompassing three piston cores with lengths between 10 and 14 meters, and four gravity cores containing the undisturbed water-sediment interfaces. The piston cores recovered were opened and described during a two-week stay in Istanbul by the German investigators in December 2009. Working halves of the cores were subsequently shipped and stored at the University of Bonn, while the archive halves are curated in the cooling facilities of Istanbul Technical University. Prior to core opening and description, we conducted whole-core geophysical logging of the core segments with an MSCL logger (Geotek Corp., UK) at the Istanbul Technical University in 0.5 cm resolution. X-ray fluorescence (XRF) scanning was then used to analyze the elemental composition of the core segments. These measurements were first carried out in resolution of 1 cm on the working halves with an ITRAX XRF scanner at the University of Cologne. Later, the data resolution was increased to 1 mm by scans at the archive halves with an Avaatech XRF scanner at the Istanbul Technical University, which was not available immediately after field work due to long-lasting maintenance.

To accomplish a continuous sedimentological record for Lake Iznik, sediment cores IZN09/LC2 and IZN09/LC3 were correlated using magnetic susceptibility and Ca/Fe ratios (for details see Roeser et al. 2012). In order to establish a robust age model for the core composite, extensive dating efforts have been made. Methods applied include AMS 14C, OSL, and tephra dating. OSL measurements were conducted on three core sections (20 cm long) from a duplicated core in cooperation with Dr. Alexandra Hilgers, CRC project F2. Tephra layers were identified macroscopically during the initial core description. Samples isolated from the tephra layers were investigated for geochemical fingerprints at the German Research Centre for Geosciences – GFZ, Potsdam (co-operation Dr. Sabine Wulf). Radiocarbon dating was conducted at the Leibniz

Laboratory – Kiel. In selected depths, 14C ages were obtained on both terrestrial plant material and aquatic ostracods, thus offering estimations for the reservoir effect in the lake. The age model of the core composite IZN09/LC2&LC3 is based on linear interpolation between the probability distribution functions of the calibrated radiocarbon ages and tephra layers (Roeser et al. 2012). According to this age model, the core composite encompasses the period ca. 3 ka BP to 35 ka BP, completed for the last 3 ka by data from the sediment cores IZN05/LC1 and IZN05/LC2 recovered in 2005. Given the uncertainties related to the basal part of the age model, and to achieve a higher resolution also for the Holocene, additional bulk sediment samples have recently been processed for radiocarbon dating. Hence, the current model will be further developed in the near future (Roeser, PhD Thesis, in preparation).

For sedimentological (mineralogy, water content, loss on ignition), geochemical (carbon, nitrogen, sulfur, total organic carbon, major and trace elements, bulk stable carbon and oxygen isotopes) and paleobiological (cladocerans, ostracods, stable carbon and oxygen isotopes on ostracod valves, and pollen) investigations, discrete samples were taken over the core composite IZN09/LC2&LC3 in 5 cm resolution. In addition, the undisturbed core catcher material offered discrete sample material. According to the present age model, the 390 samples obtained represent temporal resolutions of 60 years for the Holocene and 150 years for the Pleistocene. Since the second PhD student applied for palynology had not been approved, we were not able to provide palynological data in the running CRC phase. However, taking the importance of pollen for paleoenvironmental reconstruction, we intent to complement the existing data by palynological data in the next few years within the framework of master theses at the University Bonn.

 The proxy data as yet available from the core composite particularly suggests that fundamental changes in the hydrological system of Lake Iznik occurred at the Pleistocene/Holocene (P/H) transition (Roeser et al. 2012, Viehberg et al. accepted). This is particularly reflected by repeated shifts from heavier to lighter oxygen isotope ratios in ostracod valves (analyses in cooperation with project F4) and the onset of carbonate accumulation. This suggests a reduction in precipiation, which has causes a significant lake level drop. The results are in synchrony to a reduction in precipitation over the adjacent Sakarya Basin recorded during the H1 and H2 events (Kwiecien et al. 2009). In the course of the Holocene, however, isotope values are constantly becoming lighter. In concordance to reconstructions of Holocene precipitation patterns in the eastern Mediterranean (Roberts et al. 2008), this suggests an increase in precipitation at Lake Iznik and a rather constant rise of the lake level. In addition, the fresh-water ostracod faunal assemblage in the lake record shifts from four species in the Pleistocene to a monospecific (Limnocythere inopinata) fossil record in the Holocene. The exclusive occurrence of freshwater species further supports the assumption that the finding of marine species in the surface sediments by Nazik et al. (2010) is due to contamination of their samples (see above). The low diversity in the ostracod fauna unfortunately prohibited to apply transfer functions as originally intended (Viehberg and Mesquita-Joanes, 2012, in press). On the other hand, the shapes of the valves of L. inopinata from the Pleistocene sequence of Lake Iznik match those of modern reference material from alkaline lakes in Anatolia (Reed et al. 2012), suggesting a high degree of alkalinity. In contrary, the phenological morphospace of Holocene L. inopinata populations expresses a condition unknown from modern sites in the region. This monospecific record suggests further increasing alkalinity values during the Holocene (S. Schwabenitz, Diploma Thesis, University of Cologne).

and 2009, we additionally ran a sub-bottom survey (Innomar, equipment Istanbul Technical University) in April 2011. The data obtained turned out to be essential for the cross-correlation between multiple sites and final positioning of the consecutive core segments within the composite record (Viehberg et al. accepted). Furthermore, they provided evidence of submerged fossil terraces and the formation of a center ridge by a contourite drift system, which is in agreement with the interpretation of the sediment core proxies. Although we did not fully match the proposed time frame of 50 ka, we can conclude that we have recovered the oldest sediment record in the Marmara region, which has provided important new information concerning the regional climatic and environmental history. Lake Iznik potentially holds a continuous climate archive for the last 2.8 Ma (1,400 m sediment; pers. comm. Niessen and Franz), whose upper part fits in the scheme of the CRC. However, the funding requested for retrieving older sediments by drilling technology exceeds the available CRC budget. Therefore, an international science party has to be developed to enable a drilling campaign funded for instance

by the International Continental Scientific Drilling Programm (ICDP) along with national programmes. Besides the work on Lake Iznik, the ostracod identification in projects A2, A3, and B2 led to an outbreak group for taxonomic harmonization of freshwater species flocks in Europe and Africa within the International Research Group on Ostracoda during an international ostracodologists' conference that was organized by the B4 project postdoc Finn Viehberg. The group is led together with D. Danielopol (Graz/Austria), K. Martens (Brussels/Belgium) and T. Namiotko (Gdansk/Poland).

Project completed in 2013

asterisk (*):

  • *Roeser, P., Franz, S., Litt, T. (2016): Aragonite and calcite preservation in sediments from Lake Iznik related to bottom lake oxygenation and water column depth. – In: Sedimentology, Vol. 63(7): p. 2253–2277 , DOI: 10.1111/sed.12306
  • *Miebach, A., Niestrath, P., Roeser, P., Litt, T. (2016): Impacts of climate and humans on the vegetation in northwestern Turkey: palynological insights from Lake Iznik since the Last Glacial. – In: Climate of the Past, Vol. 12: p. 575–593, DOI: doi:10.5194/cp-12-575-2016
  • Roeser, P. (2014): Paleolimnology of Lake Iznik (NW Turkey) during the past ~ 31 ka cal BP. PhD-Thesis, University of Bonn
  • *Viehberg, F., Ülgen, U. B., Daci, E., Franz, S. O., Ön, S. A., Roeser, P., Çagatay, M. N., Litt, T., Melles, M. (2012): Seasonal hydrochemical changes and spatial sedimentological variations in Lake Iznik (NW Turkey). – In: Quaternary International, Vol. 274, DOI: 10.1016/j.quaint.2012.05.038
  • *Ülgen, U. B., Franz, S. O., Biltekin, D., Çagatay, M. N., Roeser, P. A., Doner, L., Thein, J. (2012): Climatic and environmental evolution of Lake Iznik (NW Turkey) over the last ~4700 years. – In: Quaternary International, Vol. 274, DOI: 10.1016/j.quaint.2012.06.016
  • *Roeser, P., Franz, S. O., Litt, T. (2012): Lithostratigraphic and geochronological framework for the paleoenvironmental reconstruction of the last ~36 ka cal BP from a sediment record from Lake Iznik (NW Turkey). – In: Quaternary International, Vol. 274: p. 73-87, DOI: 10.1016/j.quaint.2012.06.006

1st Phase - Research Programme
view Poster


logo koeln schrift 120x120 transparent

     logo uni bonn 40x40 transparent      dfg logo transparent

RWTH Aachen University

  logo geoverbund small 
logo qsga small new