Middle and Late Pleistocene loess sequences at Batajnica, Vojvodina, Serbia

ARTICLE IN PRESS

Quaternary International 198 (2009) 255–266

Middle and Late Pleistocene loess sequences at Batajnica, Vojvodina, Serbia Slobodan B. Markovic´a,, Ulrich Hambachb, Norm Cattoc, Mladjen Jovanovic´a, Bjo¨rn Buggled, Bjo¨rn Machalettb,e, Ludwig Zo¨llerb, Bruno Glaserd, Manfred Frechene a

Chair of Physical Geography, Faculty of Sciences, University of Novi Sad, Trg Dositeja Obradovic´a 3, 21000 Novi Sad, Serbia b Chair of Geomorphology, University of Bayreuth, 95440 Bayreuth, Germany c Department of Geography, Memorial University of Newfoundland, St. John, Newfoundland, Canada A1B 3X9 d Chair of Soil Physics, University of Bayreuth, 95440 Bayreuth, Germany e Leibniz Institute for Applied Geosciences (GGA-Institut), S3: Geochronology and Isotope Hydrology, Stilleweg 2, 30655 Hannover, Germany Available online 30 December 2008

Abstract Loess sequences in the Vojvodina region (Northern Serbia) reveal a continuous record of paleoclimatic variations during the Middle and Late Pleistocene. The most detailed stratigraphic information comes from remarkable exposures on the cliffs of the right Danube bank from Vukovar to Belgrade. The Batajnica loess section has been recognized as one of the most complete Middle and Late Pleistocene records in this region. A more than 40 m thick loess–paleosol succession represents environmental transition from relative thin loess layers and rubified soils in lower part of profile to thick loess and fossil chernozems characterizing the last three glacial–interglacial cycles. The proposed stratigraphic model is based on a detailed magnetic susceptibility (MS) record which is related to the deep-sea isotope stratigraphy and on correlation with other Eurasian loess records using the distinct MS pattern of selected loess–paleosol couplets. This new stratigraphic model suggests serious revision of previous chronological interpretations. MS as function of depth shows a well-known pattern of low values in loess and high values in paleosols indicating strong enhancement of magnetic minerals during soil formation. With the exception of the recent soil (V-S0) which is strongly contaminated by archaeological artifacts, the third paleosol V-S3 reveals the highest values in MS and a very distinct double peak. The rock magnetic signal at Batajnica resembles the typical pattern of the enviromagnetic records determined from other Eurasian loess sites. The paleopedological interpretations, rubification index values and rock magnetic record at Batajnica yield valuable data for the reconstruction of paleoclimatic fluctuations for the last 5 glacial–interglacial cycles at least. Moreover, the record provides an important link between the classical Central European loess sites and the Central Asian and Chinese loess provinces. r 2009 Elsevier Ltd and INQUA. All rights reserved.

1. Introduction Aeolian dust was deposited worldwide during the cold/ dry periods of the recent geological past. It underwent pedogenesis when more humid conditions predominated, which is reflected in physical–chemical alteration of the sediment. The product of these alterations is a pedohorizon (paleosol) which shows enhancement of magnetic minerals. In contrast, during dry periods loess was formed with Corresponding author. Tel.: +381 21 485 2837; fax: +381 21 459 696.

E-mail addresses: [email protected], [email protected] (S.B. Markovic´).

magnetic properties similar to the unaltered dust. Therefore, magnetic parameters (e.g. magnetic susceptibility, MS) as function of depth in loess–paleosol sequences can serve as a proxy for palaeoclimatic variations, allowing a close match with all kinds of high-resolution palaeoclimatic archives. Loess is by far the most important terrestrial archive that provides detailed palaeoclimatic information for the whole Quaternary and in China goes back to even the Pliocene. Heller and Liu (1984) first used magnetic susceptibility variations in Chinese loess to correlate the loess deposits to marine records. The MS variations in the loess–palaeosol couplets in the Chinese loess plateau resemble the pattern

1040-6182/$ - see front matter r 2009 Elsevier Ltd and INQUA. All rights reserved. doi:10.1016/j.quaint.2008.12.004

ARTICLE IN PRESS 256

S.B. Markovic´ et al. / Quaternary International 198 (2009) 255–266

of the global ice volume record with higher values in palaeosols (interglacials) and lower values in loess (glacials). In China, magnetic susceptibility was found to reflect the intensity of pedogenesis, which in turn leads to enhancement of magnetic minerals in soils. The thickest and most complete loess–paleosol sequences in Eurasia are preserved in the Chinese loess plateau and Central Asia (e.g. Heller and Liu, 1984; Kukla, 1987; Kukla and An, 1989; Frechen and Dodonov, 1998; Liu and Ding, 1998; Ding et al., 2002; Bronger, 2003; Sun et al., 2006; Machalett et al., 2008). Loess deposition in Europe generally started later and is characterized with smaller total thickness (e.g. Kukla, 1975, 1977; Fink and Kukla, 1977; Tsatskin et al., 1998; Jordanova and Petersen, 1999; Nawrocki et al., 1999; Sartori et al., 1999; Panaiotu et al., 2001; Dodonov et al., 2006; Jordanova et al., 2007; Hambach et al., 2008a; Velichko et al., this issue). Thick sequences of interstratified loess–paleosol sequences from the Vojvodina region (Northern Serbia) intensively investigated in recent years provide one of the most complete and most sensitive European terrestrial records of climatic and environmental changes during the Middle and Late Pleistocene (Markovic´ et al. 2003, 2004a, 2004b, 2005, 2006a, 2006b, 2007a, 2007b, 2008, in preparation; Fuchs et al., 2008; Buggle et al., 2008a, in press; Antoine et al., this issue; Bokhorst et al., this issue). The sections on the southwestern banks of the Danube near the village of Batajnica are among the most complete loess– paleosol successions in the region (Markovic´-Marjanovic´, 1970, 1972; Butrym et al., 1991; Kostic´ and Protic´, 2000; Markovic´, 2001) and provide a great potential for climatic and environmental reconstructions. The present study has three objectives: firstly, to accurately define new chronostratigraphy of the Batajnica loess–paleosol sequence; secondly, to reconstruct the paleoclimatic evolution; and thirdly, to compare these data with corresponding Eurasian loess and pollen records, as well as with marine and ice core paleoclimatic proxy records. 2. Setting, sampling and methods The Batajnica loess section is situated about 15 km northwest of Belgrade (441550 2900 N; 201190 1100 E). The analyzed profiles are exposed in steep loess cliffs of the southeastern Danube bank. The modern soil (V-S0; see chapter 3.1 for explanation of the acronyms), the last glacial loess V-L1 and pedocomplex V-S1 represent profile A. Profile B is located in a deep loess gully uncovering the penultimate glacial/interglacial cycle loess unit, V-L2 and pedocomplex V-S2. Finally, profile C includes the lowermost loess–paleosol sequences from V-L3 to the bottom of the section (Fig. 1). The synthetic Batajnica section was built up on the basis of interprofile correlation (Figs. 3 and 4). Investigations of the loess section at Batajnica began in autumn 2004. Field investigations were focused on detailed

Fig. 1. (a) Map of the loess distribution in the Vojvodina and adjacent regions showing geographical position of the investigated section and other main loess sites (modified from Markovic´ et al., 2004b). Legend: 1. Loess plateau; 2. sandy area; 3. mountain; 4. investigated exposures. (b) Topographic map of the area showing the location of the investigated loess profiles.

cleaning and description of the exposures and sampling. The lowermost part of the section (6.5 m), partly covered by recent Danube sediments, was opened mechanically in May 2005. Samples for initial low-field magnetic susceptibility measurements were taken at 5-cm intervals along 40.5 m profile resulting in 836 individual specimens. The highresolution MS measurements were obtained in the laboratory for paleo- and enviromagnetism at the Chair of Geomorphology, University of Bayreuth using the KLY-3Spinner-Kappa-Bridge (AGICO, Brno, Czech Republic). The bridge is operating with an AC-field of 300 A/m at 920 Hz.

ARTICLE IN PRESS S.B. Markovic´ et al. / Quaternary International 198 (2009) 255–266

As the MS record is the backbone of the present study, some remarks are needed about the fundamentals of enviromagnetism. Enviromagnetism deals with the magnetism of sediments and soils and is rock and mineral magnetism applied to environmental questions. It describes the occurrence, abundance and properties of iron-bearing minerals in the environment: Studying the magnetic properties of sediments and soils as proxies for environmental change involves investigating the physical (magnetic) properties of the recording ‘tape’ itself and not the ‘music’, which is recorded on it. Magnetic grains, dominantly iron oxides and sulphides, occur virtually ubiquitously in Quaternary sediments, soils, dusts and organisms, albeit often in minor or trace concentrations. After sedimentation or reworking they undergo diagenesis and pedogenesis when, e.g. more humid conditions predominate, which is reflected in physical– chemical alteration of the sediment or soil. These alterations result also in the enhancement and transformation of magnetic minerals. Especially the ferromagnetic minerals, which are predominantly formed during pedogenesis in loess environments, react in ambient laboratory magnetic fields (magnetic susceptibility) orders of magnitude stronger than other iron-bearing minerals. Thus, already very small amounts of ferromagnetic minerals control the magnetic properties of a sediment or soil (e.g. Evans and Heller, 2001; Hambach et al., 2008b). The enhancement of magnetic minerals in loess during pedogenesis is by far the most important process that defines the magnetic properties of most loess deposits. The models that deal with the enhancement processes are essential for understanding the direct mineralogical reasons for magnetic susceptibility enhancement. The formation of ferrimagnetic minerals in the course of pedogenesis is the most important mechanism. The most widely accepted model (Thompson and Oldfield, 1986; Evans and Heller, 2001) assumes alternating reducing and oxidizing conditions in the beginning, which leads to a release of Fe2+ from the weathering of iron minerals. Finally, magnetite of extremely fine grain size is formed by dehydration and is still susceptible to dissolution. Only further oxidation to maghemite results in a more stable ferrimagnetic mineral giving the palaeosols in loess their magnetic characteristics (Maher, 1998; Buggle et al., in press). The MS records are used for interprofile and interregional correlation with the marine isotope, Antarctic deuterium and Tenaghi Philippon pollen records. Dry and moist colors were recorded by means of Munsell Soil Color Charts. Determined color indices were used to calculate the rubification index according to Harden (1982) for each loess–paleosol unit of the Batajnica section. 3. Results 3.1. Litho- and pedostratigraphy Previous description and stratigraphic interpretations of the Batajnica exposure have been presented by Markovic´-

257

Marjanovic´ (1972), Butrym et al. (1991) and Kostic´ and Protic´ (2000). The new investigations show that the nearly 40 m high loess cliff is intercalated with 6 major pedocomplexes and 6 weakly developed interstadial soils (Fig. 2). As reported in previous studies (Markovic´Marjanovic´, 1970, 1972; Butrym et al., 1991; Kostic´ and Protic´, 2000; Markovic´, 2001), paleopedological interpretations suggest gradual climatic transition from semi-humid and warm to drier and cooler environments. Fig. 2 presents the comparison between the description of the Batajnica loess–paleosol sequence presented by Butrym et al. (1991) and this new interpretation. According to the profile description, the oldest exposed loess layers and fossil soils are affected by strongly developed carbonate concretions and hydromorphic features, such as manganese nodules, iron coatings and iron oxide patches. Only the uppermost part of the oldest observed paleosol is exposed. The oldest exposed loess layer is only 50 cm thick. Overlying this loess is a 300 cm thick pedocomplex including a lower weakly rubified B horizon (7.5YR 4/4–5YR 4/4), which is gradually transformed to an upper altered A horizon. Above this is fossil soil, with two reddish yellow (7.5YR 6/6–5/6) loess layers intercalated with a yellowish brown (10YR 5/6–4/4) paleosol. The strongly developed pedocomplex V-S5 is 450 cm thick. From bottom to top this composite paleosol unit includes a lower yellowish-red (5YR 5/6–4/4) weakly rubified B horizon disturbed by many hydromorphic features; the middle dark brown (7.5YR 4/4–5YR 4/4) weakly rubified cambic horizon with moderately developed coarse polyhedral structure; and an upper pale brown (10YR 7/4–4/4) horizon. The pale yellow loess V-L5 is 395 cm thick. At 100 cm above the base of this loess unit, the 25 cm thick initial fossil soil V-L5S1 (10YR 6/3–5/4) is observed in this section. Many carbonate concretions (1.5–4 cm diameter) and humus infiltrations in ancient root channels are developed at the contact between rgw V-S4 soil complex and the underlying V-L5 loess. The fossil pedocomplex V-S4 measures 125 cm and includes a lower cambic B horizon (10YR 5/6 4/4) and an upper altered A horizon. Above the paleosol V-S4, the 110 cm thick loess unit V-L4 (2.5Y 7/4 6/4) is developed. The unit comprises many humic infiltrations and carbonate concretions near the base of the overlying pedocomplex V-S3. The polygenetic paleosol V-S3 is 160 cm thick and includes from bottom to top a lower transitional A–B horizon (7.5 YR 6/4 4/4); a thin lighter layer with carbonate nodules; an altered A horizon in the middle (10YR 7/3 4/3) with remarkable crotovinas; and an upper initial horizon (10YR 5/2 3/3) also with many crotovinas. The (10YR 8/3 5/4) loess V-L3 is 520 cm thick, with a weakly developed initial pedological horizon V-L3S1 (10YR 8/2 5/3) with crotovinas included in the center. Overlying this loess, V-S2S2, a 175 cm thick paleosol complex is developed. The lower strongly developed darker (10 YR 5/4 4/3) A2 horizon has

ARTICLE IN PRESS 258

S.B. Markovic´ et al. / Quaternary International 198 (2009) 255–266

Fig. 2. Comparison between Batajnica loess–paleosol sequence description of (A) Butrym et al. (1991) and (B) this interpretation. Legend for Butrym et al. (1991) description: 1. Modern chernozem; 2. Fossil initial cherozem like horizons; 3. Fossil chernozems with signs of degratation; a: weaker; b: stronger; 4. Brown forest and brown soils with slight signs of rubefication; 5. Brown soil with signs of rubefication; a: noticeable; b: very noticeable; 7. Carbonate concretions; 8. Younger loess pale yellow and yellowish; 9. Oldest loess yellowish and yellow-reddish. Legend for this study description: 1. Loess; 2. embryonic pedogenic layer; 3. A horizon; 4. Ah horizons; 5. B horizon; 6. Bw rubified horizon; 7. sand beds; 8. tephra layers; 9. Neolithic artefacts; 10. carbonate concretions; 11. krotovinas.

many carbonate pseudomycelia. Crotovinas are scattered throughout the middle lighter colored (10 YR 5/4–4/3) A1 horizon and the upper weakly developed (10YR 5/2 3/3) A horizon. Subunit V-S2L1 is 85 cm thick and consists of weathered loess separating the subsoils of V-S2. Above this thin loess is the 110 cm thick V-S2S1, a light brown (10 YR 6/3–4/4) weakly developed pedological horizon. In total, the composite pedocomplex V-S2 is 370 cm thick. The following pale yellow (10YR 8/2 2.5Y4/4) loess unit V-L2 measures 340 cm. The middle part of penultimate glacial loess is finely laminated with thin sand beds. Below and above 14 m profile depth are two remarkable horizons, probably related to tephra layers. Very similar lithostratigraphic features of the penultimate loess V-L2 are visible at

the loess profiles at Surduk (Danube bank) and in the Titel loess plateau. The last interglacial–early glacial pedocomplex V-S1 is 210 cm thick and composed of a lower dark A horizon indicating a fossil chernozem formation (10YR 6/3 3/4) overlain by a middle lighter A horizon (10YR 6/3 3/2) and the uppermost weakly developed A horizon (10YR 5/2 3/3) characterized by numerous crotovinas. The composite last glacial loess unit V-L1 is 900 cm thick. The lower subunit V-L1L2 (10YR 8/2 2.5Y 5/4) accumulated above paleosol V-S1 is a porous sandy loess, loosely cemented and in some parts finely laminated with thin sand beds. The middle pleniglacial period is represented by two weakly developed initial pedogenic horizons V-L1S1S2 (10YR 8/2 2.5 4/4)

ARTICLE IN PRESS S.B. Markovic´ et al. / Quaternary International 198 (2009) 255–266

259

and V-L1S1S1 separated by a thin interloess layer V-L1S1L1 (10YR 8/2 5/3). The uppermost late pleniglacial loess subunit V-L1L1 (10YR 8/2 2.5 4/4) is less cemented and very porous. At the top of the investigated section, the modern soil is a 90 m thick carbonate chernozem (Miljkovic´, 2001). The lower Ck horizon contains many CaCO3 nodules of 1–5 cm in diameter, numerous crotovinas and root channels filled with humic material. A transitional AC horizon (10YR 5/1–3/3) is a 25-cm thick, very porous, silty loam with fine blocky structure. The Ah horizon (10YR 6/3 3/3) is a 30-cm thick silty loam with granular structure and some carbonate pseudomycelia. The modern soil is strongly contaminated by archaeological artifacts (including pottery) and animal bones, and in some parts is disturbed by previous human occupation.

3.2. Magnetic susceptibility record The synthetic Batajnica section was constructed on the basis of intercorrelation of the MS records from the 3 investigated sub-profiles (Fig. 3). The complete MS record of the synthetic Batajnica loess–paleosol sequence related to the pedostratigraphic interpretation is presented in Fig. 4. Fig. 4. Synthetic MS record related to pedostratigraphy (compare with Fig. 2).

Fig. 3. Correlation of the MS records from sub-profiles A, B and C located in Fig. 1b.

The variations in the low-field MS well reflect the pedostratigraphy in the Batajnica section. MS values observed in the pedocomplexes related to interglacial periods (35–105  108 m3 kg1) are higher than in the interstadial soils (25–40  108 m3 kg1) and loess units (12–28  108 m3 kg1) (Figs. 4 and 5). Fig. 5 gives a histogram of the MS data obtained at Batajnica. A clear separation between the MS values from almost unaltered loess and the enhanced MS in paleosols as well as in the anthropogenic influenced V-S0 is visible. This type of MS pattern reflects magnetic enhancement via pedogenesis and is similar to that in Chinese and Central Asian loess deposits (e.g. Maher and Thompson, 1992, 1999). The MS signal of the lowermost part of the profile is disturbed by post-depositional hydromorphic processes, especially at sequences between the oldest exposed pedocomplex and polygenic paleosol V-S5, and with decreasing intensity up to V-L4. The MS course with depth in pedocomplex V-S5 shows a tripartite pattern. A broad maximum in the lower part is followed by a broad relative minimum on top of which a surprising sharp peak with highest values for this pedocomplex can be observed. The following decrease of the MS values towards V-L5 is developed as a very characteristic and pronounced long shoulder. The origin of the pronounced peak is not clear yet. However, recent results of the investigations from pedocomplex V-S5 in the Titel loess plateau reveal exactly

ARTICLE IN PRESS 260

S.B. Markovic´ et al. / Quaternary International 198 (2009) 255–266

Fig. 5. Histogram of the MS data obtained at Batajnica. Note the clear separation between the MS values from almost unaltered loess and the enhanced MS in paleosols, as well as in the anthropogenic influenced V-S0.

the same pattern, pointing to a regional rather than local phenomenon. Gently elevated MS values in V-L5 mark weak interstadial pedogenesis. The youngest fossil pedocomplex V-S4, formed under forest environmental conditions, provides lower MS values than all younger steppic pedocomplexes. The thin loess layer V-L4 has higher MS values than in other loess units, probably partly caused by bioturbation. The third pedocomplex V-S3 reveals the highest MS values observed in all fossil soils at Batajnica section with a distinct and unmistakable double peak. This double peak makes the MS pattern of pedocomplex V-S3 to a unique marker. Loess unit V-L3 has slightly higher MS values than the upper loess units V-L2 and V-L1. In the middle part of loess V-L3, MS values gently increase indicating the weakly developed interstadial paleosol V-L3S1. The paleosol V-S2 reveals a typical tripartite MS pattern. A sharp but relatively weak peak in the lower part is followed by the central well-developed MS peak (V-S2S2). The thin loess layer V-S2L1 separates the uppermost pedohorizon V-S2S1 from the main paleosol. V-S2S1 shows relatively gentle increased MS values with very small but distinct peak on top. The penultimate glacial loess unit V-L2 shows relative uniform MS values excluding a sudden peak observed 1 m below paleosol V-S1, indicating a possible tephra layer, as recognized during field and preliminary mineralogical investigations (Dr. Sabine Wulf, University of Texas at Austin, personal communication). The last interglacial–early pleniglacial pedocomplex V-S1 shows a gradual decrease of MS over time. Generally, low MS values characterized the composite loess unit VL-1. Discrete higher MS values are observed in weak developed paleosols V-L1S1S2 and V-L1S1S1 compared to loess layers

V-L1L2, V-L1S1L1 and V-L1L1. In contrast, the highest MS values in the whole profile are related to recent soil V-S0 which is strongly contaminated by archaeological artifacts and was partly disturbed during the Neolithic occupation. Preliminary paleomagnetic results performed for the lowermost 6 m reveal continuous normal polarity, which also supports the proposed chronostratigraphic model. 3.3. Rubification index The quantification of soil redness, as expressed by the rubification index (RI), at the Batajnica loess section provide considerably lower values in the loess layers than in the palaeosols. RI values are quite uniform in the loess (around 0 units), except the determined RI of the loess L4 which can be regarded as the influence of pedogenic overprint, as magnetic susceptibility values are also enhanced (Fig. 5). The differences between the soils show significant age dependence. An increased rubification value founded in the older pedocomplexes V-S5 and V-S4 is the result of different pedogenetical conditions than during the formation of the younger steppic soils V-S3, V-S2, and V-S1, characterized by weaker intensification of soil redness compared to the background value of the loess. The RI values identified for weakly developed interstadial soils are slightly higher than in the loess layers (Fig. 7). 4. Discussion The application of the Chinese loess stratigraphical scheme (e.g. Kukla, 1987; Kukla and An, 1989) and the use of MS as a basis for differentiating loess and paleosol units and correlating them regionally and relating them to the

ARTICLE IN PRESS S.B. Markovic´ et al. / Quaternary International 198 (2009) 255–266

deep-sea isotope stratigraphy, result in a serious revision of the earlier chronostratigraphic interpretations of the Batajnica loess–paleosol sequence (Markovic´-Marjanovic´, 1970; Butrym et al. 1991; Kostic´ and Protic´, 2000) (Fig. 6). Butrym et al. (1991) provided correlation with marine oxygen stratigraphy based on their results of thermoluminescence dating. According to this chronological model, the younger paleosols b (V-L1S1S1 in this stratigraphic subdivision), d (V-L1S1S2), f and g (V-S1) developed during the last glacial period. Soils h2 (V-S2S1) and i (V-S2S2) are time equivalents of the last interglacial period, and fossil soils k, l1 and l2 (V-S3) developed during MIS 7. Paleosols n1 (V-S4), n3 and n5 (probable equivalents of VS5) formed during MIS 9. The oldest double pedocomplex, p1 and p2, developed during MIS 13 or 15. The data presented in Table 1 summarize existing chronostratigraphic data for the Batajnica loess–paleosol sequence. Numerical dates presented by Butrym et al. (1991) were produced using problematic methodology, criticized by members of the luminescence dating community (e.g. Wintle, 1987; Frechen et al., 1997). However, at present luminescence dating in general, including new more lightsensitive techniques such as optically stimulated lumines-

261

cence (OSL) and infrared optically stimulated luminescence (IRSL), is still not adequate for age estimations older than 100 ka (e.g. Singhvi et al., 2001; Wintle and Murray, 2006) and cannot be the base for valid stratigraphic subdivisions of older loess–paleosol units. For example, the most recent results of IRSL dating provide ages of 120.7712.7 ka for the uppermost part of the penultimate loess V-L2 at the Surduk section, located on the right bank of the Danube River between Stari Slankamen and Batajnica (Fuchs et al., 2008). This IRSL date well agrees with the proposed Table 1 Comparison of stratigraphic models for the Batajnica loess–palaeosol sequence. Markovic´-Marjanovic´ (1970)

Butrym et al. (1991)

This model

Unit

Alpine stratigraphy

Unit

MIS

Unit

MIS

PK PK PK PK PK

Wu¨rm

g h2+i k+l1+l2 n1

5 7 9 9

V-S1 V-S2 V-S3 V-S4 V-S5

5 7 9 11 13–15

II III IV IV

Riss–Wu¨rm

Fig. 6. Correlation of MS records Paks (Sartori et al., 1999), Batajnica, Ruma (Markovic´ et al., 2006a), Mostistea (Panaiotu et al., 2001) and Koriten (Jordanova and Petersen, 1999), with the astronomically tuned MS curve of the Chinese loess site Lingtai/Zhaojiachuan (Sun et al., 2006) and with the SPECMAP oxygen isotope record (Bassinot et al., 1994). Note that all SE-European sections are plotted on the depth scale of the loess profiles at Paks.

ARTICLE IN PRESS 262

S.B. Markovic´ et al. / Quaternary International 198 (2009) 255–266

chronostratigraphic model and provides a much older age estimation than the results of Butrym et al. (1991) for the lowermost part of penultimate loess V-L2 at Batajnica section indicating an age around 90 kyr. Moreover, the recent stratigraphical and chronological review of the loess units V-S1, V-L1 and V-S0 in the Vojvodina region by Markovic´ et al. (2008), including up-to-date luminescence dating, clearly assign these loess–paleosol sequences from MIS 5 to 1. The proposed new stratigraphic model is also in agreement with recent results of amino acid racemization (AAR) geochronology of different loess sections in the Vojvodina region (Markovic´ et al., 2004a, 2004b, 2005, 2006a, 2006b, 2007a, 2007b, 2008). These results provide stratigraphic correlations between loess–paleosol units V-L1-S1, V-L2-S2, V-L3-S3 and V-L4-S4 in the Vojvodina region with loess of Glacial cycles B, C, D and E (Kukla, 1975), respectively, at other European loess localities (Zo¨ller et al., 1994; Oches and McCoy, 1995a, 1995b, 2001). The rubification values increase in the older pedocomplexes V-S5 and V-S4. This increase is the result of different pedogenetical conditions compared to those during the formation younger steppic soils (Fig. 7). This is in good agreement with previous studies of rubification, recorded at other Eurasian loess sites (e.g. Vidic´ et al., 2004). However, these results indicate an even stronger trend of paleoclimatic transition from sub-Mediterranean to dry continental climate in this part of Europe during the last five glacial–interglacial cycles. Changes of rubification index values provide additional palaeoclimatic evidence important for the understanding of the Middle and Late Pleistocene palaeoclimatic evolution and development of palaeosols during interstadial and interglacial periods. Similar results of rubification for Mircea Voda in southeastern Romania have been found by Buggle et al. (2008b), indicating a regional climatic trend. Many previous investigations of loess–paleosol sequences around the world have used MS as a basis for differentiating widespread loess and paleosol units, correlating them regionally and relating them to the deep-sea isotope stratigraphy (e.g. Heller and Evans, 1995; Hambach et al., 2008b). The oldest exposed pedocomplex and weaker developed paleosol could be related to MIS 17 and MIS 16.3. However, the lowermost part of profile is strongly affected by intensive post-depositional hydromorphic features disturbing the primary morphological and magnetic properties of loess–paleosol sequences. Therefore, at the present level of investigations, the data from the Batajnica loess–paleosol sequences older than V-S5 are not sufficiently reliable for the correlation with marine isotope stratigraphy and equivalent regional loess records. The strongest developed thick pedocomplex V-S5 was formed during MIS 13–15. This pedocomplex is a characteristic feature of the middle part of all Brunhes loess–paleosol sequences in Eurasia (e.g. Bronger, 2003). MS variations recorded in thick pedocomplex V-S5 show a

Fig. 7. Rubification index values related to pedostratigraphy (compare with Fig. 2).

similar pattern to the Chinese loess section (e.g. Sun et al., 2006). The overlying pedocomplex V-S4 reveals relatively low MS values and is overlain by the thin loess V-L4 with high MS compared to other loess layers. This pattern was observed at the lower Danube in Romania and Bulgaria and seems to be characteristic for the paleoclimatic evolution of the region during MIS 10-11 (Jordanova and Petersen, 1999; Panaiotu et al., 2001). The third pedocomplex V-S3 reveals the highest MS values observed in all fossil soils at Batajnica section with a distinct double peak also visible in equivalent paleosols S3 in China (e.g. Kukla, 1987; Kukla and An, 1989; Liu and Ding, 1998; Sun et al., 2006), PK III in Central Asia (Forster et al., 1994), BA of Paks section in Hungary (Sartori et al., 1999) and S3 of Bulgarian sites Koriten,

ARTICLE IN PRESS S.B. Markovic´ et al. / Quaternary International 198 (2009) 255–266

Viatovo and Ljubenovo (Jordanova et al., 2007). The paleosol V-S2 represents a typical MS pattern which can be observed also at other loess sections in the middle and lower Danube Basin (Sartori et al., 1999; Panaiotu et al., 2001; Markovic´ et al., 2006a; Jordanova et al., 2007; Buggle et al., in press) (Fig. 6). The course of MS with depth in V-S1 and V-L1 shows a characteristic pattern present in China as well as in the Carpathian region. The sudden increase of the values at the base is followed by the main peak which again shows a gradual decrease of MS with time (Jordanova and Petersen, 1999; Panaiotu et al., 2001;Sun et al., 2006; Markovic´ et al., 2006a; Buggle et al., in press). This structure shows the representation of the climatic peak and decline from MIS 5.5–5.1. The MS pattern in V-L1 is marked by two saw-tooth-shaped weak peaks with gentle slopes at the base and a sudden decline at the top revealing the typical climatic trend of the last glacial cycle with interstadial paleosols. At Batajnica, the interstadial paleosols are relatively weakly developed, but can be easily correlated to other well-studied sections in the region (Markovic´ et al., 2008, in preparation). The pattern of MS variation and absence of any erosion sign in the profile suggest correlation of paleosols V-S4, V-S3, V-S2 and V-S1 with MIS 11, 9, 7 and 5, respectively (Fig. 6).

263

The revised chronostratigraphic frame for the Batajnica loess–paleosol sequence provides an opportunity to examine the long-term environmental dynamics within the context of regional, continental and global climate changes. The loess–paleosol sequences at Batajnica deposited during the last five interglacial–glacial cycles are characterized by sharp environmental differences between high dust accumulation rates during the glacials and reduced dust fluxes in the periods of paleosol formation. Low values of MS and relatively high accumulation rates derived from loess units V-L1, V-L2, VL-3 and V-L5 are in good agreement with the relative extent of ice sheets as estimated from marine isotope records (e.g. Bassinot et al., 1994), reconstructed from field evidence from Europe (e.g. Ehlers, 1996), and the duration of minimal presence of arboreal pollen (AP) (Tzedakis et al., 2006). The thin loess layer V-L4 at Batajnica has higher MS values, indicating a shorter duration of pure glacial climate conditions during MIS 10 characterized by lower dust deposition rates. Despite general agreement with the most important global paleoclimatic records, detailed environmental reconstruction determined from loess–paleosol sequences at Batajnica indicates some specific differences. The paleopedological interpretations and rubification index values of

Fig. 8. Correlation between the reconstructed interglacial environments from Batajnica loess sequence plotted on timescale defined SPECMAP dO18 series (Bassinot et al., 1994), EPICA Dome C deuterium variations (EPICA community, 2004) and duration of full interglacial conditions, and Tenaghi Phillipon Quercus and Carpinus pollen percentages (Tzedakis et al., 2006). Legend: dS—dry steppe; S—steppe; SF—steppe/forest; F—forest; St—subtropic environments.

ARTICLE IN PRESS 264

S.B. Markovic´ et al. / Quaternary International 198 (2009) 255–266

the Batajnica loess–paleosol sequence indicate a succession of interglacial environmental changes from semi-humid subtropical environments, to temperate forest, and finally towards landscapes with typical steppe soils. Comparison with equivalent Eurasian loess–paleosol records indicate a similar paleoclimatic trend (e.g. Kukla, 1987; Kukla and An, 1989; Pecsi, 1995; Ding et al., 2002; Bronger, 2003; Velichko et al., this issue), but paleoenvironmental indices of progressive aridization especially during the interglacials are more sharply expressed at Batajnica section. Changes of interglacial environments at Batajnica loess–paleosol sequence are much more dramatically expressed than in marine oxygen isotope (e.g. Bassinot et al., 1994) and Antarctic ice deuterium (EPICA community members, 2004) records, or especially in the NE Mediterranean arboreal pollen successions (Tzedakis et al., 2006), which suggest smaller differences regarding to amplitude of interglacial climates. However, the remarkable changes of interglacial environments preserved in Serbian loess coincide with the presence of warmer and shorter post-Mid-Bruhnes Event (MBE) interglacials recorded in the EPICA ice core deuterium record (EPICA, 2004), and with some quantitative vegetation changes of the Tenaghi Philippon pollen record after MIS 16, such as increasing dominance of Quercus and Carpinus which are drought tolerant taxa (Tzedakis et al., 2006) (Fig. 8). The remarkable coincidence between the Antarctic temperature pattern (EPICA, 2004; Jouzel et al., 2007) and expansion of aridity recorded in Serbian loess–paleosol sequences during the Middle Pleistocene underline the importance of understanding climatic mechanisms responsible for differences between these paleoclimatic records. Recently, Yin and Guo (2008), Yin et al. (2008) and Guo et al. (2008) correlating Chinese loess, deep-sea sediments and Antarctica ice records suggested that hemispheric climate coupling at the glacial–interglacial scale was significantly unstable during the mid-Pleistocene, and that the degree of asymmetry of polar ice-conditions has prominent impacts on the global climate system. 5. Conclusions The investigations of the loess–paleosol sequence at Batajnica demonstrate the importance of this site as a representative record of Middle and Late Pleistocene paleoclimate and paleoenvironment in this part of Europe. The distinct and characteristic MS record at Batajnica provides important and significant similarities to the enviromagnetic records observed at other Eurasian loess sections. This opens possibilities to extend the temporal and spatial correlation across the Eurasian loess belt from China via Central Asia to the middle Danube Basin. The new stratigraphic model based on these correlations suggests serious reinterpretations of previous chronological concepts of the Batajnica site. The primary paleoclimatic signal of the lowermost part of the profile is significantly disturbed by hydromorphic

processes. In contrast, the loess–paleosol sequences formed during the last 5 glacial–interglacial cycles provide one of the most complete European continental paleoclimatic records for the last ca. 620 ka, and to date the only complete record for the middle Danube basin. The inferred trend in dust accumulation rates and the intensity of pedogenesis demonstrate clear evidence for the Middle Pleistocene climate and environmental transition indicating a direct link to the temporally and spatially progressive aridization of interior Eurasia since the lower Pleistocene. These interpretations raise questions about the importance of temporal and spatial reconstructions of the middle Pleistocene climatic shifts and environmental dynamics across the Eurasian continent in the scope of understanding the present and of prediction of future natural processes and their development. Acknowledgments We thank Tivadar Gaudenyi, Nebojsˇa Miljkovic´ and Tin Lukic´ for their help during the field work. This research was supported by Project 146019 of the Serbian Ministry of Science, Project 114-451-00754 of the Secretariat for Science and Technological Development, IV AP Vojvodina and by a grant of the German Federal Ministry for Education and Research (BMBF, MOE 04/R01). Furthermore, our work benefited from a Junior-Senior fellowship of the Alexander von Humboldt foundation and Stability pact for SE Europe awarded to S.B. Markovic´ and M. Jovanovic´. References Antoine, P., Rousseau, D.D., Fuchs, M., Hatte´, C., Gautier, C., Markovic´, S.B., Jovanovic´, M., Gaudeenyi, T., Moine, O., Rossignol, J., this issue. High resolution record of the last climatic cycle in the Southern Carpathian basin (Surduk, Vojvodina, Serbia). Quaternary International, doi:10.1016/j.quaint.2008.12.008. Bokhorst, M.P., Beets, C.J., Markovic´, S.B., Gerasimenko, N.P., Matviishina, Z.N., Frechen, M., this issue. Pedo-chemical climate proxies in Late Pleistocene Serbian–Ukrainian loess sequences. Quaternary International, doi:10.1016/j.quatint.2008.09.003. Bassinot, F.V., Labryie, L.D., Vincent, E., Quidelleur, X., Shackleton, N.J., Lancelot, Y., 1994. The astronomical theory of climate and the age of Brunhes–Matuyama magnetic reversal. Earth Planetary Science Letters 126, 91–108. Bronger, A., 2003. Correlation of loess–paleosol sequences in East and Central Asia with SE Central Europe—towards a continental Quaternary pedostratigraphy and paleoclimatic history. Quaternary International 106–107, 11–31. Butrym, J., Maruszcak, H., Zeremski, M., 1991. Thermoluminescence stratigraphy of Danubian loess in Belgrade environs. Annales, Universite Marie-Curie Sklodowska, B 46, 53–64. Buggle, B., Glaser, B., Zo¨ller, L., Hambach, U., Markovic´, S., Glaser, I., Gerasimenko, N., 2008a. Geochemical characterization and origin of Southeastern and Eastern European loesses (Serbia, Romania, Ukraine). Quaternary Science Reviews 27, 1058–1075. Buggle, B., Hambach, U., Glaser, B., Markovic´, S.B., Glaser, I., Zo¨ller, L., 2008b. Long-Term Paleoclimate Records in SE-Europe—The Loess Paleosol Sequences Batajnica/Stari Slankamen (Serbia) and Mircea Voda (Romania). Hauptversammlung der DEUQUA, Wien, 31

ARTICLE IN PRESS S.B. Markovic´ et al. / Quaternary International 198 (2009) 255–266 August – 6 September 2008, Abhandlungen der Geologischen Bundesanstalt, Band 62, pp. 15–19. Buggle, B., Hambach, U., Glaser, B., Gerasimenko, N., Markovic´, S.B., Glaser, I., Zo¨ller, L., in press. Stratigraphy and spatial and temporal paleoclimatic trends in Southeastern/East European loess paleosol sequences. Quaternary International, doi:10.1016/j.quaint.2008.07.013. Dodonov, A.E., Zhou, L.P., Markova, A.K., Tchepalyga, A.L., Trubikhin, V.M., Akesandrovski, A.L., Simakova, A.N., 2006. Middle– Upper Pleistocene bio-climatic and magnetic records of the Northern Black Sea Coastal Area. Quaternary International 149, 44–54. Ding, Z.L., Ranov, V., Yang, S.L., Finaev, A., Han, J.M., Wang, G.A., 2002. The loess record in southern Tajikistan and correlation with Chinese loess. Earth and Planetary Science Letters 200, 387–400. Ehlers, J., 1996. Quaternary and Glacial Geology. Wiley, Chichester. EPICA community members, 2004. Eight glacial cycles from an Antarctic ice core. Nature 429, 623–628. Evans, M.E., Heller, F., 2001. Magnetism of loess/paleosol sequences: recent developments. Earth Science Reviews 54, 129–144. Fink, J., Kukla, G.J., 1977. Pleistocene climates in Central Europe: at least 17 interglacials after the Olduvai event. Quaternary Research 7, 363–371. Forster, Th., Evans, M.E., Heller, F., 1994. The frequency dependence of low field susceptibility in loess sediments. Geophysical Journal International 118, 636–642. Frechen, M., Dodonov, A.E., 1998. Loess chronology of the Middle and Upper Pleistocene in Tadjikistan. Geologische Rundschau 87, 2–20. Frechen, M., Horva´th, E., Ga´bris, G., 1997. Geochronology of Middle and Upper Pleistocene loess sections in Hungary. Quaternary Research 48, 291–312. Fuchs, M., Rousseau, D.D., Antoine, P., Hatte´, C., Gauthier, C., Markovic´, S., Zo¨ller, L., 2008. Chronology of the last climatic cycle (Upper Pleistocene) of the Surduk loess sequence, Vojvodina, Serbia. Boreas 37, 66–73. Guo, Z.T., Berger, A., Yin, Q.Z., Qin, L., 2008. Strong asymmetry of hemispheric climates during MIS-13 inferred from correlating China loess and Antarctica ice records. Climate of the Past Discussions 4, 1061–1088. Hambach, U., Markovic´, S.B., Rolf, C., Zo¨ller, L., 2008a. Pala¨ound umweltmagnetische Untersuchungen an Sedimenten der Bohrung Frankenbach 1. In: Hansch, W., Rosendahl, R. (Eds.), 600.000 Jahre Zeitgeschichte am Neckar, vol. 24. MUSEO, Heilbronn, pp. S.68–76. Hambach, U., Rolf, C., Schnepp, E., 2008b. Magnetic dating of Quaternary sediments, volcanites and archaeological materials: an overview. Eiszeitalter und Gegenwart/Quaternary Science Journal 57/ 1–2, 25–51. Harden, J.W., 1982. A quantitative index of soil development from field descriptions: examples from a chronosequence in Central California. Geoderma 28, 1–28. Heller, F., Liu, T.S., 1984. Magnetism of Chinese loess deposits. Geophysical Journal of the Royal Astronomical Society 77, 125–141. Heller, F., Evans, M.E., 1995. Loess magnetism. Reviews of Geophysics 33, 211–240. Jordanova, D., Petersen, N., 1999. Paleoclimatic record from a loess-soil profile in northeastern Bulgaria II. Correlation with global climatic events during the Pleistocene. Geophysical Journal International 138, 533–540. Jordanova, D., Hus, J., Geeraerts, R., 2007. Palaeoclimatic implications of the magnetic record from loess/palaeosol sequence Viatovo (NE Bulgaria). Geophysical Journal International 171, 1036–1047. 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, 793–796.

265

Kostic´, N., Protic´, N., 2000. Pedology and mineralogy of loess profiles at Kapela Batajnica and Stalac. Catena 41, 217–227. Kukla, G.J., 1975. Loess stratigraphy of Central Europe. In: Butzer, K.W., Isaac, L.I. (Eds.), After Australopithecines. Mouton Publishers, The Hague, pp. 99–187. Kukla, G.J., 1977. Pleistocene land–sea correlations. Earth Science Review 13, 307–374. Kukla, G.J., 1987. Loess stratigraphy in Central China. Quaternary Science Reviews 6, 191–219. Kukla, G.J., An, Z., 1989. Loess stratigraphy in Central China. Palaeogeography, Palaeoclimatology, Palaeoecology 72, 203–225. Liu, T.S., Ding, Z., 1998. Chinese loess and paleomonsoon. Annual Review of Earth and Planetary Sciences 26, 111–145. Machalett, B., Oches, E.A., Frechen, M., Zo¨ller, L., Hambach, U., Mavlyanova, N.G., Markovic´, S.B., Endlicher, W., 2008. Aeolian dust dynamics in Central Asia during the Pleistocene—driven by the longterm migration, seasonality and permanency of the Asiatic polar front, Kazakhstan. Geophysics, Geochemistry and Geosystems 9, Q09008. Maher, B.A., 1998. Magnetic properties of modern soils and Quaternary loessic paleosols: paleoclimatic implications. Palaeogeography, Palaeoclimatology, Palaeoecology 137, 25–54. Maher, B.A., Thompson, R., 1992. Paleoclimatic significance of the mineral magnetic record of the Chinese loess and paleosols. Quaternary Research 37, 155–170. Maher, B.A., Thompson, R. (Eds.), 1999. Quaternary Climates, Environment and Magnetism. Cambridge University Press, Cambridge. Markovic´-Marjanovic´, J., 1970. Data concerning the stratigraphy and the fauna of the lower and middle Pleistocene of Yugoslavia. Paleogeography, Paleoclimatology, Paleoecology 8, 153–163. Markovic´-Marjanovic´, J., 1972. The extent and stratigraphy of loess in Yugoslavia. Glasnik Prirodnjacˇkog Muzeja, Series A 27, 93–107 (in Serbian, with French summary). Markovic´, S.B., 2001. Paleozemljisˇta Srema (Paleosols of Srem region) (in Serbian with English summary). In: Miljkovic´, N., Markovic´, S.B. (Eds.), Zemljita Serma (Soils of the Srem region,Vojvodina, Yugoslavia). Institut za Geografiju, Novi Sad, pp. 133–155. Markovic´, S.B., Heller, F., Kukla, G., Gaudenyi, T., Jovanovic´, M., Miljkovic´, Lj., 2003. Magnetostratigrafija lesnog profila Cˇot u Starom Slankamenu. Zbornik radova Instituta za geografiju 32, 20–28. Markovic´, S.B., Kostic´, N., Oches, E.A., 2004a. Paleosols in the Ruma loess section. Revista Mexicana de Ciencias Geolo´gicas 21, 79–87. Markovic´, S.B., Oches, E.A., Gaudenyi, T., Jovanovic´, M., Hambach, U., Zo¨ller, L., Su¨megi, P., 2004b. Paleoclimate record in the Late Pleistocene loess–palesol sequence at Miseluk (Vojvodina, Serbia). Quaternaire 15, 361–368. Markovic´, S.B., McCoy, W.D., Oches, E.A., Savic´, S., Gaudenyi, T., Jovanovic´, M., Stevens, T., Walther, R., Ivanisˇevic´, P., Galovic´, Z., 2005. Paleoclimate record in the Late Pleistocene loess–paleosol sequence at Petrovaradin Brickyard (Vojvodina, Serbia). Geologica Carpathica 56, 545–552. Markovic´, S.B., Oches, E., Su¨megi, P., Jovanovic´, M., Gaudenyi, T., 2006a. An introduction to the Upper and Middle Pleistocene loess–paleosol sequences of Ruma section (Vojvodina, Yugoslavia). Quaternary International 149, 80–86. Markovic´, S.B., Hambach, U., Gaudenyi, T., Jovanovic´, M., Savic´, S., Zo¨ller, L., Machalett, B., Savic´, S., Romelic´, J., Mesarosˇ, M., 2006b. An introduction to the Late Pleistocene loess–palaeosol sequences of Susek section (Vojvodina, Serbia). Geographica Pannonica 10, 4–8. Markovic´, S.B., Oches, E.A., McCoy, W.D., Gaudenyi, T., Frechen, M., 2007a. Malacological and sedimentological evidence for ‘‘warm’’ climate from the Irig loess sequence (Vojvodina, Serbia). Geophysics, Geochemistry and Geosystems 8, Q09008. Markovic´, S.B., Hambach, U., Oches, E.A., McCoy, W.D., Zoeler, L., Jovanovic´, M., 2007b. 850 Millennia of paleoclimatic history recorded in the loess sequences of Vojvodina region, Serbia. Quaternary International 167/168 (Suppl.), 269. Markovic´, S.B., Bokhorst, M., Vandenberghe, J., Oches, E.A., Zo¨ller, L., McCoy, W.D., Gaudenyi, T., Jovanovic´, M., Hambach, U., Machalett,

ARTICLE IN PRESS 266

S.B. Markovic´ et al. / Quaternary International 198 (2009) 255–266

B., 2008. Late Pleistocene loess–paleosol sequences in the Vojvodina region, North Serbia. Journal of Quaternary Science 23, 73–84. Markovic´, S.B., Hambach, U., Stevens, T., Kukla, G., Heller, F., Machalett, B., Oches, E.A., McCoy, W.D., Jovanovic´, M., Gaudenyi, T., Buggle, B., Zo¨ller, L., in preparation. Stari Slankamen loess–paleosol sequence: revised chronostratigraphy and long-term environmental trends. Miljkovic´, N., 2001. Recent soil cover. In: Miljkovic´, N., Markovic´, S.B. (Eds.), Soils of Srem Region. Institute of Geography, Novi Sad, pp. 39–132. (in Serbian with English summary). Nawrocki, J., Bakhmutov, V., Bogucki, A., Dolecki, L., 1999. The paleo- and petromagnetic record in the Polish and Ukrainian loess–paleosol sequences. Physics and Chemistry of the Earth (A) 24, 773–777. Oches, E.A., McCoy, W.D., 1995a. Amino acid geochronology applied to the correlation and dating of Central European loess deposits. Quaternary Science Reviews 14, 767–782. Oches, E.A., McCoy, W.D., 1995b. Aminostratigraphic evaluation of conflicting age estimates for the ‘‘Young Loess’’ of Hungary. Quaternary Research 44, 767–782. Oches, E.A., McCoy, W.D., 2001. Historical developments and recent advances in amino acid geochronology applied to loess research: examples from North America, Europe and China. Earth Science Reviews 54, 173–192. Panaiotu, C.G., Panaiotu, C.E., Grama, A., Necula, C., 2001. Paleoclimatic record from a loess–paleosol profile in southeastern Romania. Physics and Chemistry of the Earth (A) 26, 893–898. Pecsi, M., 1995. The role of principles and methods in loess–paleosol investigations. GeoJournal 36, 117–131. Sartori, M., Heller, F., Forster, T., Borkovec, M., Hammann, J., Vincent, E., 1999. Magnetic properties of loess grain size fractions from section Paks (Hungary). Physics Earth and Planetary Interiors 116, 53–64. Singhvi, A.K., Bluszcz, A., Bateman, M.D., Someshwar Rao, M., 2001. Luminescence dating of loess–palaeosol sequences and coversands: methodological aspects and palaeoclimatic implications. Earth Science Reviews 54, 193–212.

Sun, Y., Clemens, S.C., An, Z., Yu, Z., 2006. Astronomical timescale and palaeoclimatic implication of stacked 3.6-Myr monsoon records from the Chinese Loess Plateau. Quaternary Science Reviews 25, 33–48. Thompson, R., Oldfield, F., 1986. Environmental Magnetism. Allen and Unwin Ltd., London. Tsatskin, A., Heller, F., Gendler, T.S., Virina, E.I., Spassov, S., Du Pasquier, J., Hus, J., Hailwood, E.A., Bagin, V.I., Faustov, S.S., 1998. A new scheme of terrestrial paleoclimatic evolution during the last 1.5 Ma in the western Black Sea region: integration of soil studies and loess magnetism. Palaeogeography, Palaeoclimatology, Palaeoecology 143, 111–133. Tzedakis, P.C., Hooghiemstra, H., Pa¨like, H., 2006. The last 1.35 million years at Tenaghi Philippon: revised chronostratigraphy and long term vegetation trends. Quaternary Science Reviews 25, 3416–3430. Velichko, A.A., Catto, N., Kononov, Yu, M., Morozova, T.D., Novenko, E.Y., Panin, P.G., Ryskov, Ya. G., Semenov, V.V., Timireva, S.N., Titov, V.V., Teskov, A.S., this issue. Progressively coller, drier interglacials in southern Russia through the Quaternary: evidence from the Sea of Azov region. Quaternary International, doi:10.1016/ j.quaint.2008.06.005. Vidic´, N., Singer, M.J., Versoub, K.L., 2004. Duration dependance of magnetic susceptibility enhancement in Chinese loess–paleosols of the past 620 ky. Palaeogeography, Palaeoclimatology, Palaeoecology 211, 271–288. Wintle, A.G., 1987. Thermoluminescence dating of loess. Catena 9 (Suppl.), 103–115. Wintle, A.G., Murray, A.S., 2006. A review of quartz optically stimulated luminescence characteristics and their relevance in single-aliquot regeneraton dating protocols. Radiation Measurements 41, 369–391. Yin, Q.Z., Guo, Z.T., 2008. Strong summer monsoon during the cool MIS 13. Climate of the Past 4, 29–34. Yin, Q.Z., Berger, A., Driesschaert, H., Goose, H., Loutre, M.F., Crucifix, M., 2008. The Eurasian ice sheet reinforces the East Asian summer monsoon during the interglacial 5,00,000 years ago. Climate of the Past 4, 79–90. Zo¨ller, L., Oches, E.A., McCoy, W.D., 1994. Towards a revised chronostratigraphy of loess in Austria with respect to key sections in the Czech Republic and in Hungary. Quaternary Science Reviews 13, 465–472.