And Litho-Stratigraphic Records of the Oligocene-Early Miocene Climatic

Global and Planetary Change 158 (2017) 36–46

Contents lists available at ScienceDirect

Global and Planetary Change

journal homepage: www.elsevier.com/locate/gloplacha

Magneto- and litho-stratigraphic records of the Oligocene-Early Miocene T climatic changes from deep drilling in the Linxia Basin, Northeast Tibetan Plateau

⁎ Fuli Wua, Xiaomin Fanga,b, , Qingquan Mengb, Yan Zhaoa, Fenjun Tanga,c, Tao Zhanga, Weilin Zhanga, Jinbo Zana a CAS Center for Excellence in Tibetan Plateau Earth Sciences, Key Laboratory of Continental Collision and Plateau Uplift, Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing 100101, China b School of Earth Sciences, Key Laboratory of Western China's Mineral Resources of Gansu Province, Lanzhou University, Lanzhou 730000, China c University of Chinese Academy of Sciences, Beijing 100049, China

ARTICLE INFO ABSTRACT

Keywords: The East Asian monsoon is generally regarded to have initiated at the transition from the Late Oligocene to the Oligocene/Miocene boundary Early Miocene. However, little is known about this process because of a lack of continuous strata across the Paleomagnetism boundary between the Late Oligocene and the Early Miocene in Asia. Based on previous drilling (core HZ-1) in Hematite and redness records the Miocene sediments in the southern Linxia Basin in NW China, we drilled a new 620 m core (HZ-2) into the Early monsoon history Late Oligocene strata and obtained 206 m of continuous new core. The detailed paleomagnetism of the new core Linxia Basin reveals eleven pairs of normal and reversed polarity zones that can be readily correlated with chrons 6Bn-9n of the geomagnetic polarity time scale (GPTS), define an age interval of 21.6–26.5 Ma and indicate continuity from the Late Oligocene to Early Miocene. The core is characterized by the remarkable occurrence of brownish-red paleosols of luvic cambisols (brown to luvic drab soils) above reddish-brown floodplain siltstones and mudstones, which suggest that the East Asian monsoon likely began by 26.5 Ma. In contrast to the siltstone and mudstone of the Late Oligocene strata, the Miocene strata begin with a thick fine sandstone bed, which marks sudden increases in erosion and loading that most likely reflect a response to tectonic uplift. The hematite content and redness index records of the core further demonstrate that the monsoonal climate in the Late Oligocene to Early Miocene in this area was mainly controlled by global temperature trends and events.

1. Introduction the Early Miocene was a critical interval for the climate transition and change in the Cenozoic era. The Oligocene was a period of climatic deterioration that witnessed In Asia, evidence from both temporal and spatial investigations the initiation of the continent-wide ice sheet on Antarctica and the shows that a climate transformation from a zonal pattern to a monsoon- formation of the Antarctic circumpolar current. Deep-sea δ18O records dominated pattern occurred roughly from the Late Oligocene to the (e.g., Miller et al., 1987; Zachos et al., 2001) and terrestrial palyno- Early Miocene (Liu et al., 1998; Guo et al., 2002, 2008; Wang et al., morphs recovered from Antarctic margin drill holes (e.g., Mildenhall, 2003a, 2003b, 2005; Sun and Wang, 2005; Wang, 2009; Qiao et al., 1989; Raine, 1998; Raine and Askin, 2001) indicate climatic ameli- 2006; Qiang et al., 2011), though several studies suggest that the East oration in the Late Oligocene, which may be attributable to deep-sea Asian monsoon had already appeared in the Eocene (Licht et al., 2014; warming coupled with a collapse of the Antarctic ice sheet (e.g., Miller Quan et al., 2012). Therefore, the transition between the Oligocene and et al., 1987; Zachos et al., 2001). Subsequently, until the middle Mio- Miocene is a critical interval for global and East Asian climate change. cene, the global ice volume remained low, and the bottom water tem- However, a deﬁnite determination of the timing of the onset of the East peratures increased slightly but were punctuated by several short in- Asian monsoon is required to understand the mechanisms and re- tervals of glaciations (called Mi-events) (Flower and Kennett, 1994; lationships (phases) with global climatic change and the uplift of the Zachos et al., 2001). Therefore, the period from the Late Oligocene to Tibetan Plateau. This has long been impeded by a widespread shortage

⁎ Corresponding author at: CAS Center for Excellence in Tibetan Plateau Earth Sciences, Key Laboratory of Continental Collision and Plateau Uplift, Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing 100101, China. E-mail address: [email protected] (X. Fang). http://dx.doi.org/10.1016/j.gloplacha.2017.09.008 Received 29 March 2017; Received in revised form 16 August 2017; Accepted 10 September 2017 Available online 18 September 2017 0921-8181/ © 2017 Elsevier B.V. All rights reserved. F. Wu et al. Global and Planetary Change 158 (2017) 36–46

Fig. 1. (A) Location of the Linxia Basin in the topographic and climatic transitional zone between the Tibetan Plateau and the humid monsoon front (modified from Fang et al., 2015). (B) Geologic map of the Linxia Basin showing the distribution of the Cenozoic stratigraphy. The red circle indicates the location of the Heilinding section (modified from Fang et al., 2016). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) of well-dated and continuous terrestrial sediment sequences that record zone between the East Asian monsoon and non-monsoon Asian inland the climate changes across the boundary between the Oligocene and dry climate (Fig. 1). Detailed stratigraphic, biostratigraphic and mag- Miocene. netostratigraphic studies suggest that the exceptionally long continuous The Linxia intracontinental foreland basin (Fang et al., 2016) pro- sedimentary succession in this basin was deposited from the Oligocene vides an ideal opportunity for this purpose. It is located in a transitional to the Pliocene (Li et al., 1995; Fang et al., 2003, 2016; Deng et al.,

37 F. Wu et al. Global and Planetary Change 158 (2017) 36–46

(caption on next page)

38 F. Wu et al. Global and Planetary Change 158 (2017) 36–46

Fig. 2. (I), Stratigraphic succession of the HLD outcrop section (226.5 m) and its relation to drilling cores HZT (129.2 m), HZ-1 (477.7 m) and HZ-2 (620 m) at its top and bottom, blue and red arrows show the location of the collected samples; (II), google map of the study area; (III), the photos show the main lithologies of the cores HZ-1 and HZ-2. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

2004, 2013; Qiu et al., 2004; Garzione et al., 2005). In particular, the a cambic Bw horizon, an underlying distinct calcic Bk horizon, and Heilinding (HLD) section (Figs. 1, 2) in the foredeep of the southern occasionally an additional calcic Ck horizon (Fang et al., 2016). The Bt Linxia basin includes continuous thick floodplain sediments with three horizon is approximately 0.5–1 m thick, more reddish in color, has levels of fossil mammals but few lithofacies variations, so it is suitable black Fe/Mn mottling stains and contains a well-developed granular for revealing climate change (Fang et al., 2016). Unfortunately, the ped structure (Fig. 2III-b and e). The Bw horizon is similar to the Bt stratigraphy at Heilinding outcrops over a thickness of only 226.5 m. horizon but is lighter red (mostly dull reddish-brown) and has less Therefore, in 2010 and 2011, we drilled a 129.2 m core (HZT) at the top pedality (ped to block structures). The Bk horizon is located im- of the section and a 477.7 m core (HZ-1) (Fig. 2) at its bottom. Detailed mediately below the Bt or Bw horizon, is light brownish gray or yel- paleomagnetic measurements demonstrated an age of 23.3 Ma for the lowish-gray, and contains many white carbonate coatings, small no- bottom of core HZ-1 (Fang et al., 2016). Encouraged by those results, dules and impregnations that cement the sediments (Fig. 2III-b). Down- we performed additional deeper drilling (core HZ-2, drilling depth leaching carbonate occasionally reaches the soil's parent materials 620 m) at the same site as core HZ-1 in December 2015 using the same (mudstone or siltstone) below the Bk horizon, but its content is less than drilling technology (three-in-one tube drilling, in which the core is di- that in the Bk horizon, and it forms a Ck horizon (Fig. 2III-f). The lower rectly recovered into an inner plastic tube for better protection and ZZ Fm. is composed of alternating thick reddish-brownish parallel- recovery rate) and obtained 206 m of new core with an recovery rate of bedding fine sandstones (Fig. 2III-d) and massive siltstones, which are over 95% (Fig. 2). In this paper, we carried out a detailed paleomag- classified as a river channel facies over a floodplain (Fig. 2)(Fang et al., netic analysis of this core and analyzed its lithofacies, hematite contents 2016). and redness proxies. Based on these data, we attempted to reconstruct the climatic changes across the boundary between the Late Oligocene 3. Sampling and methods and the Early Miocene. Core HZ-2 can be linked to core HZ-1 by the sandstone marker bed 2. Geological setting and stratigraphy (Fig. 2I). We collected samples starting at a depth of 414.5 m in core HZ-2, which overlaps with core HZ-1 by 63.7 m. Paleomagnetic samples – The Linxia Basin is located on the northeastern margin of the were collected at intervals of 0.5 1 m within the overlapping section – Tibetan Plateau and has an average elevation of 2000–2600 m above and at 0.25 0.5 m intervals elsewhere, and a total of 554 block samples sea level (asl). It is surrounded by several high mountains, including the were collected (Fig. 2I). Each block sample was cut into two West Qin Ling Mts. to the south, which have a maximum elevation of 2 × 2 × 2 cm specimens for measurements. One set of specimens was fi 3767 m, the Laji Shan (Mts.) (4524 m asl) to the west, and the Maxian subjected to stepwise alternating- eld (AF) demagnetization in 15 steps Shan (Mts.) (ca. 3000–3500 m asl) to the north, and it opens to the east (0, 2, 4, 8, 12, 16, 20, 30, 40, 50, 60, 80, 100, 120 and 140 mT). For to the large Longzhong Basin (Fig. 1). The modern climate in the Linxia crosschecking, 80 samples from the second set were systematically Basin exhibits the features of the continental monsoon with a mean thermally demagnetized in 19 discrete steps between room temperature – annual temperature (MAT) of 6.7 °C, an average monthly maximum and 690 °C (intervals of 50 °C below 550 °C and 10 30 °C above 550 °C) temperature of 18 °C and mean annual precipitation (MAP) of > 500 in a Magnetic Measurement Thermal Demagnetizer (MMTD80). All of mm. Warm temperate zone montane broadleaved forests and subalpine the remaining measurements were then conducted using a 2G En- fi coniferous forests are distributed in the mountainous areas around the terprises Model 760-R cryogenic magnetometer installed in a eld-free basin, and the landscape within the basin is temperate broadleaved space (< 100 nT) at the Institute of Tibetan Plateau Research, Chinese deciduous forest-grassland. Academy of Sciences. – The Cenozoic stratigraphy of the Linxia Basin mainly consists of red 180 samples were taken at 1 2 m intervals below a depth of 600 m beds of fluvio-lacustrine mudstones and siltstones in the center and (samples over 824 m were collected from core HZ-1, the else from core ff fl thick strata of gray to reddish-brown conglomerates and sandstones on HZ-2) for di use re ectance spectroscopy (DRS) analysis for redness the margins (Fang et al., 2003, 2016). Previous studies have divided the and hematite content measurements (Barranco et al., 1989; Balsam and stratigraphy into eight formations based on lithofacies, contacts, pa- Deaton, 1991; Deaton and Balsam, 1991). Most of the samples were leontology and magnetostratigraphy, including the Late Oligocene Tala collected from mudstones or siltstones so the impact of the sediment (TL) Fm., the Early Miocene Zhongzhuang (ZZ) Fm., the Middle Mio- matrix could be eliminated from the results. First, all of the samples fi cene Shangzhuang (SZ) Fm., the Late Miocene Dongxiang (DX) Fm. and were ground in an agate mortar. The nely powdered samples were – Liushu (LS) Fm., the Early and Late Pliocene Hewangjia (HWJ) Fm. and then pressed by hand into 20 40 mm rectangular plastic holders. The fl Jishi (JS) Fm. or Jishi Conglomerate Bed and the Early Pleistocene re ectance spectra of the samples were measured in a Purkinje General ff fl Dongshan (DS) Fm. (Li et al., 1995, 1997; Fang et al., 2003; Fang et al., TU1901 UV-vis spectrophotometer with a di use re ectance attach- fl 2016)(Fig. 2). ment (re ectance sphere) from 360 to 850 nm at 1 nm intervals. The The HLD section (35.38°N, 103.32°E) is located on the southern DRS analysis was conducted at the Institute of Tibetan Plateau Re- fl margin of the Linxia Basin and consists of the HZT core (129.2 m) in the search, Chinese Academy of Sciences. All of the re ectance data were fi upper part (0–129.2 m), the HLD outcrop (226.5 m) and the HZ-1 core analyzed by the rst derivative test to calculate the hematite contents (477.7 m) in the middle and lower parts (120–824.5 m) (Fang et al., (Balsam and Deaton, 1991; Deaton and Balsam, 1991). 2016) and the HZ-2 core (620 m) in the lowermost part (761–967.15 m) (Fig. 2). This paper studied the intervals of the section that include the 4. Magnetostratigraphy of core HZ-2 TL Fm. and ZZ Fm. (Fig. 2). The TL Fm. and the upper ZZ Fm. are composed of an upward-fining cycle of alternating brownish siltstones The natural remanent magnetization (NRM) of the samples from and reddish-brown mudstones (Fig. 2III-a) and strongly to moderately core HZ-2 is within the range of 0.1–10 mA/m, and most of them are developed calcic paleosol complexes (Fig. 2III-b and e), which are within the range of 1–10 mA/m (Fig. 3). The remanent magnetization mostly interpreted as overbank floodplain deposits (Fig. 2). These pa- (RM) of all of the samples decreases as the intensity of the applied leosols are luvic paleosols that are characterized by a luvic Bt horizon or magnetic field increases. When the intensity of the applied magnetic

39 F. Wu et al. Global and Planetary Change 158 (2017) 36–46

Fig. 3. Representative alternating ﬁeld and thermal demagnetization diagrams for samples from core HZ-2. Open (closed) symbols represent vertical (horizontal) projections.

field reaches 140 mT, the RM of most of samples decreases to within c), which indicates the elimination of secondary RM. After the turning 25% of the NRM (Fig. 3a–d), and the RM of some of the samples de- points, the RM direction approaches the origin and can represent the creases to within 25%–50% of the NRM (Fig. 3e, i and k). There are two stable characteristic remanent magnetization (ChRM) directions. In the types of vector graphs for the RM direction. In the first type, there are other type of vector graph, the RM direction does not change sig- two RM directions, and the RM direction shows remarkable changes at nificantly (Fig. 3d, e and i), and the stable ChRM direction can be ac- turning points of 8 mT or 12 mT (Fig. 3a and k) and 30 mT (Fig. 3b and quired. The higher ChRM values after 140 mT AF-demagnetization of

40 F. Wu et al. Global and Planetary Change 158 (2017) 36–46 the samples indicate that some of the ChRM is carried by hard magnetic 5. Variations of lithology, hematite content and redness and their minerals, such as hematite. The RM of the thermally demagnetized indications of Asian monsoon change samples decreases significantly at 550–580 °C (Fig. 3h) and 610 °C (Fig. 3f), and the most significant RM decrease occurs at temperatures 5.1. DRS measurements and variations of hematite content and redness higher than 650 °C (Fig. 3f–h, j and i), all these indicate that the main records RM carriers in the samples are magnetite and hematite. Due to the existence of hematite with a high coercive force, complete demagneti- DRS analysis is highly sensitive to iron oxide minerals in the soil and zation cannot be achieved through alternating field demagnetization. sediments and has been recognized as an important instrument to The RM of a few samples increases before 120 °C (Fig. 3f) but then identify and estimate iron oxide minerals in soil and sediments decreases with increasing temperature, which suggests the elimination (Schwertmann, 1971; Comell and Schwertmann, 1996; Balsam et al., of the viscous remanence carried by goethite. For thermal demagneti- 2004; Fang et al., 2015; Torrent et al., 2006; Nie et al., 2017). Because zation, the RM direction generally approaches the origin at tempera- the characteristic reflectance spectra are different for different iron tures higher than 400 °C (Fig. 3f and g), and the RM direction ap- oxide minerals, several iron oxide minerals in the samples, especially proaches the origin at temperatures higher than 610 °C in some samples hematite and goethite, can be identified by DRS scanning. The DRS first (Fig. 3h), which can represent the ChRM direction. The demagnetiza- order derivative of the cores HZ-1 and HZ-2 has only one significant tion effects of the samples are generally the same as those of the sam- peak, and the center is at 568 nm, which indicates the content of the ples from core HZ-1 (Fang et al., 2016). hematite (Fig. 6, Fang et al., 2015). The redness is the percentage ratio The directions of the ChRM components of all of the paleomagnetic of the reflectivity of the red band at 630–700 nm to the total reflectivity samples from core HZ-2 are calculated through principle component of the visible band (400–700 nm) (Judd and Wyszecki, 1975). analysis (Kirschvink, 1980), and the results are generally the same for Fig. 7a shows the variations of hematite content and redness for the the alternating field demagnetization and thermal demagnetization stratigraphic interval of 600–976 m (18–26.5 Ma) in the section com- (Fig. 3i, j, k, and l), which substantiates the reliability of the test results. posed of cores HZ-1 and HZ-2. They show a generally homogeneous Because the core azimuth was not controlled during sampling, the pattern with high but narrow ranges of 0.26 –0.45 for the hematite magnetic declination has no significance, and the geomagnetic polar content and 31.55–42.59 for the redness, which are punctuated by column can only be created with geomagnetic inclination data (Fig. 4). several small but clear short-lasting valleys (lower values) at depths of During sample selection, the samples with maximum angular deviations 820 m (22.9 Ma), 773 m (21.9 Ma), 726 m (21 Ma), 669 m (19.7 Ma), (MAD) > 15° that were used to determine the ChRM direction with 639 m (18.9 Ma) and 610 m (18.1 Ma), which are marked as hematite fewer than four points were excluded. A total of 52 samples, which and redness events HR 1 to HR 6, respectively (Figs. 4b, c and 7a, b). account for 9.4% of the samples, were excluded. Because the sampling was conducted at high resolution and the excluded samples were not 5.2. Changes in the Asian monsoon from the Late Oligocene to the Early concentrated in contiguous layers, the precision and interpretation of Miocene the paleomagnetic polar columns was not affected. Based on the principle that at least two samples of the same polarity define each polarity The studied stratigraphic intervals of cores HZ-1 and HZ-2 are event, 54 samples (account for 9.7% of the samples) were excluded, dominated by fine sediments of reddish-brown mudstones and silt- then 11 positive polarity events (N1–N11) and 11 negative polarity stones, which are intercalated with many well developed paleosols and events (R1–R11) were obtained (Fig. 4d). a short interval of fine sandstones between them that typically char- Fig. 4 shows that the polarity zones (R1–R3, N1–N2) obtained from acterize a floodplain environment (Reading, 2009)(Figs. 2 and 4a). We the interval of core HZ-2 that overlaps with core HZ-1 are highly con- interpreted the well-developed luvic or cambic paleosols, which only sistent with those (R38–R40, N39–N40) of core HZ-1 (Fang et al., form under subaerial conditions, and the subaerially oxidized flood- 2016), which confirms the connection between the cores. This provides plain mudstones and siltstones as an indication that the East Asian a reference for determining the age of the upper boundary of core HZ-2 monsoon had formed by approximately 26.5 Ma and throughout the because the correlation of magnetic polarity zones of core HZ-1 with studied Late Oligocene to Early Miocene interval. This is because each the GPTS is very robust, with considerable age constraints from three kinds of soils represent a specific climatic condition (Retallack, 2001), levels of fossil mammals that were found in the section (Fang et al., and the modern soils such as these luvic or cambic paleosols form in 2016). Therefore, we can readily correlate normal polarity zone N1 of easternmost China like Shandong Province (Fig. 1a), where the MAT core HZ-2 with chron 6Bn and reversed polarity zone R2 with chron 6Br and MAP are approximately 12–14 °C and 600–850 mm, respectively, of the GPTS (Hilgen et al., 2012)(Fig. 4f, g). Thus, we correlate N3 and the warmest month temperature is 23–27 °C, and the average pre- N4 with chrons 6Cn.2n and 6Cn.3n, respectively, considering that the cipitation in the summer (June to August) accounts for > 60% of the corresponding channel sandstones indicate high erosion and sedi- MAP (Xiong and Li, 1987). In other words, the climatic condition in- mentation rates (Fig. 5) in comparison with the underlying floodplain dicated by the paleosols is better than that at the modern Linxia Basin. mudstones and paleosols (Fig. 4). The frequent occurrence of long Moreover, the water vapor source at that time cannot be the westerly normal zones N6 to N10 can be correlated with chrons 7Cn to 8n, and wind, as many studies show that the Northwest of the China is more N11 is readily correlated with the top part of 9n (Fig. 4f, g). The de- arid after the Oligocene (e.g. Dupont-Nivet et al., 2007; Xiao et al., position rate of core HZ-2 established by the correspondence between 2010; Sun et al., 2014, Sun and Windley, 2015), the water vapor can the depth and the magnetostratigraphic age (Fig. 5) shows it ranges only be brought in by the East Asian monsoon, and the monsoon in the stably between 2.64 and 6.33 cm/ka, and the average deposition rate is Linxia Basin was stronger in the Late Oligocene and Early Miocene than 4.28 cm/ka, which is similar with the one at bottom of the core HZ-1 at present. (3.84 cm/ka) (Fang et al., 2016). Then this correlation demonstrates The abundant mammal fossils that have been discovered in the that the age interval of core HZ-2 is 21.6–26.5 Ma based on extra- Linxia Basin provide further support for this interpretation (Deng et al., polations with the sedimentation rate (Fig. 5). The results indicate 2004, 2013). The giant rhino fauna in the Linxia Basin in the Late continuity in the stratigraphic record from the Late Oligocene to the Oligocene mainly include Dzungariotherium orgosense, Tsaganomys al- Early Miocene and shows that the beginning of the Miocene is marked taicus, Megalopterodon sp., Schizotherium ordosium, Triplopus sp., Ardynia by the appearance of the sandstone bed above the reddish mudstones sp., Ardynia sp. nov., Allacerops sp., Paraceratherium sp., Ronztherium and paleosols at approximately 23 Ma, which is consistent with the sp., Aprotodon lanzhouensis, and Paraentelodon macrognathus, and this standard age for the start of the Miocene within the errors (Cohen et al., mammalian assemblage indicates a warm and humid habitat (Deng 2013). et al., 2004, 2013), which confirms our interpretation of the monsoon.

41 F. Wu et al. Global and Planetary Change 158 (2017) 36–46

Fig. 4. Depth functions of the lithology (a), changes of hematite content and redness (b and c), and inclination directions of core HZ-2 and the determined polarity zones (black/white bars indicate normal/reverse polarity) (d, f). For comparison, the magnetic inclination directions and the determined polarity zones of the lower part of core HZ-1 are also plotted (e, f) (Fang et al., 2016). Both polarity zones are correlated with the GPTS of Hilgen et al. (2012) (g).

42 F. Wu et al. Global and Planetary Change 158 (2017) 36–46

Fig. 5. Age versus stratigraphic depth plot of the core HZ-2 using the correlations of the magneozones shown in Fig. 4f. GPTS is cited from Hilgen et al., 2012.

Fig. 6. First derivative curves for several samples and the highest ﬁrst derivative values at 568 nm from cores HZ-1 and HZ-2. The dark line indicates the ﬁrst derivative for 100% hematite (Balsam and Deaton, 1991).

The lower section of the ZZ Fm. consists of alternating parallel- occurred on and around the Himalaya-Tibetan Plateau (Yin et al., 1998; bedded sandstones and siltstones that belong to a river channel facies Wang et al., 2003a, 2003b, 2015; Fang et al., 2005; Sobel et al., 2006; (Fig. 2). In contrast to the ﬁne siltstone and mudstone of the TL Fm., Lu et al., 2012; Lease et al., 2011; Wang et al., 2016). these sandstones appeared at the beginning of the Miocene and indicate We also consider the hematite content and redness to be sensitive sudden increases in erosion and loading. These were likely a response to climatic proxies. Previous studies show that hematite is a common the Early Miocene tectonic uplift of the NE Tibetan Plateau (Fang et al., mineral in soils and sediments and is a product of silicate minerals that 2003, 2016) because this event has been widely reported to have have weathered into the soil (Schwertmann, 1971; Balsam and Damuth,

43 F. Wu et al. Global and Planetary Change 158 (2017) 36–46

Fig. 7. Time series of hematite content (a) and redness (b) in cores HZ-1 and HZ-2 based on their magnetostratigraphy (see Fig. 4) and comparison with oxygen isotope records from the South China Sea (c) (Tian et al., 2008) and global marine stacks (d) (Zachos et al., 2008; the heavy line is the 20-point moving average).

2000; Li et al., 2012; Nie et al., 2014). In the period of the 26.5–18 Ma, redness is usually closely related to hematite content (Fang et al., 1999, the main provenance of sediments in the HLD section did not change 2015; Giosan et al., 2002, White et al., 2007). Therefore, when the (Fang et al., 2016). Even though we could not rule out the impact of temperature increases, the oxidation of sediments will be enhanced, original hematite in the rocks, especially Permian-Triassic igneous that leads to the high hematite content then high redness. This could be rocks in the West Qinling to the south of the Linxia Basin (Liu et al., why hematite and redness have been widely and successfully used as 2015), we think it would not exert a signiﬁcant role in our hematite sensitive proxies in reconstructing climatic changes (e.g., Ji et al., 2002, record in million and less million scales because its concentration in the 2006, 2007; Song et al., 2005; Nie et al., 2010; Liu et al., 2011; Fang rocks of the source areas is very low, chemically resistant (White and et al., 2015). Buss, 2014) and hard to impact short times of climatic changes. Rather, We also interpreted that the high values of the hematite content and the hematite in the sediments of the basin should mostly be formed in redness throughout the Late Oligocene and Early Miocene strata are the depositional stage with climate changes (e.g. Hu et al., 2005; Fang records of the strong East Asian monsoon in the region because the et al., 2015). Moreover, hematite is mainly distributed in soils with higher monsoonal precipitation and temperature is relatively favorable good oxidizing conditions in tropical and subtropical zones and there- for hematite formation and high redness. Thus, the six short intervals of fore mainly indicates a relatively hot environment in the modern lower hematite contents and redness values (HR1 to HR6) recorded six (Schwertmann, 1971; Comell and Schwertmann, 1996; Fang et al., short weaker monsoon events with a roughly one million year cycle 2015; Ji et al., 2002, 2006, 2007). The relationship between color and (Figs. 4 and 7). These events, together with the long-term trend, can be climate is complicated, but most studies indicate that color is related to correlated within the age errors, which are caused by using diﬀerent the contents of organic matter and coloring minerals. For example, GPTS times and our sampling intervals, with the global and regional

44 F. Wu et al. Global and Planetary Change 158 (2017) 36–46 temperature records indicated by the global and South China Sea visible light spectra. Rev. Aquat. Sci. 4, 411–447. Balsam, W., Ji, J.F., Chen, J., 2004. Climatic interpretation of the Luochuan and Lingtai marine oxygen isotope records (Zachos et al., 2008; Tian et al., 2008); loess section, China, based on changing iron oxide mineralogy and magnetic sus- events HR1, HR2, HR4 and HR6 are correlated with the Miocene short ceptibility. Earth Planet. Sci. Lett. 223 (3–4), 335–348. cooling events Mi-1, Mi-1a, Mi-1aa and Mi-1b (Miller et al., 1987, 1991; Barranco, J.F., Balsam, W.L., Deaton, B.C., 1989. Quantitative reassessment of brick red lutites: evidence from reflectance spectrophotometry. Mar. Geol. 89, 299–314. Wright et al., 1992), respectively (Fig. 7), and HR3 and HR5 can be Berger, A., Imbrie, J., Hays, J.D., 1984. Milankovitch and Climate. Reidel Publishing. correlated with two unrecognized cold events between Mi-1a and Mi- Cohen, K.M., Fineey, S.C., Gibbard, P.L., Fan, J.X., 2013. The ICS international chron- 1aa and between Mi-1aa and Mi-1b, respectively (Fig. 6). These events ostratigraphic chart. Episodes 36, 199–204. occurred on a roughly one million cycle, which appears to have re- Comell, R.M., Schwertmann, U., 1996. The Iron Oxides: Structure, Properties, Reactions, Occurrence and Uses Weinheim. VCH Verlagsgesellschaft, pp. 573. corded signals of the long-term 0.8–1.2 Ma eccentricity cycles Deaton, B.C., Balsam, W.L., 1991. Visible spectroscopy—a rapid method for determining (Milankovitch, 1930; Berger et al., 1984). hematite and goethite concentration in geological materials. J. Sediment. Petrol. 61, – Because the hematite content and redness events HR1 to HR6 occur 628 632. Deng, T., Wang, X., Ni, X., Liu, L., 2004. Sequence of the Cenozoic mammalian faunas of irrespective of lithology (e.g., event HR2 occurs in the sandstone, HR3 the Linxia Basin in Gansu, China. Acta Geol. Sin.-Engl. 78, 8–14. occurs in the paleosol, and HR4 and 5 occur in the mudstone), we be- Deng, T., Qiu, Z., Wang, B., Wang, X., Hou, S., Wang, X., Flynn, L., Fortelius, M., 2013. lieve that these events were not affected by tectonic events but rather Late Cenozoic biostratigraphy of the Linxia Basin, northwestern China. In: Wang, X., Flynn, L.J., Fortelius, M. (Eds.), Fossil Mammals of Asia: Neogene Biostratigraphy and are real records of monsoon climate weaknesses caused by global Chronology. Columbia University Press, New York, pp. 243–273. cooling events. Therefore, we conclude that the East Asian monsoon in Dupont-Nivet, G., Krijgsman, W., Langereis, C.G., Abels, H.A., Sabbadini, R., Dai, S., Fang, the Late Oligocene to Early Miocene was principally controlled by both X., 2007. Tibetan Plateau aridification linked to global cooling at the Eocene–Oligocene transition. Nature 445, 635–638. the long-term trend and short-term events of global temperature var- Fang, X.M., Ono, Y., Fukuksawa, H., Pan, B.T., Li, J.L., Guan, D.H., Oi, K.C., Tsukamoto, iations during that period. S., Torii, M., Mishima, T., 1999. Asian summer monsoon instability during the past 60,000 years: magnetic susceptibility and pedogenic evidence from the western Chinese Loess Plateau. Earth Planet. Sci. Lett. 168, 219–232. 6. Conclusions Fang, X.M., Garzione, C., Van der Voo, R., Li, J.J., Fan, M.J., 2003. Flexural subsidence by 29 Ma on the NE edge of Tibet from the magnetostratigraphy of Linxia Basin, China. 1) We drilled a new core (HZ-2) that extended 620 m from the bottom Earth Planet. Sci. Lett. 210, 545–560. of the HLD outcrop section in the southern Linxia foreland basin on Fang, X.M., Yan, M.D., Van der Voo, R., Rea, D.K., Song, C.H., Parés, J.M., Gao, J.P., Nie, J.S., Dai, S., 2005. Late Cenozoic deformation and uplift of the NE Tibetan Plateau: the northeastern margin of the Tibetan Plateau and obtained a evidence from high-resolution magnetostratigraphy of the Guide Basin, Qinghai continuous high quality stratigraphic record that straddles the Late Province, China. Geol. Soc. Am. Bull. 117, 1208–1225. Oligocene to Early Miocene boundary. Fang, X.M., Zan, J.B., Appel, E., Lu, Y., Song, C.H., Dai, S., Tuo, S.B., 2015. An Eocene–Miocene continuous rock magnetic record from the sediments in the Xining 2) Detailed paleomagnetic analyses of core HZ-2 yielded an age in- Basin, NW China: indication for Cenozoic persistent drying driven by global cooling terval of 21.6–26.5 Ma. and Tibetan Plateau uplift. Geophys. J. Int. 201, 78–89. 3) The overwhelming occurrence of subaerially oxidized fine sedi- Fang, X.M., Wang, J.Y., Zhang, W.L., Zan, J.B., Song, C.H., Yan, M.D., Appel, E., Zhang, T., Wu, F.L., Yang, Y.B., Lu, Y., 2016. Tectonosedimentary evolution model of an ments in reddish-brown siltstones and mudstones accompanied by intracontinental flexural (foreland) basin for paleoclimatic research. Glob. Planet. many layers of well-developed brownish-red paleosols of luvic Chang. 145, 78–97. cambisols from cores HZ-1 and HZ-2 characterizes a typical flood- Flower, B.P., Kennett, J.P., 1994. The middle Miocene climatic transition: East Antarctic ice sheet development, deep ocean circulation and global carbon cycling. plain environment and suggests an already occurrence of the East Palaeogeogr. Palaeoclimatol. Palaeoecol. 108, 537–555. Asian monsoon at latest at about 26.5 Ma and extended throughout Garzione, C.N., Ikari, M.J., Basu, A.R., 2005. Source of Oligocene to Pliocene sedimentary the studied interval of the Late Oligocene to the Early Miocene. rocks in the Linxia basin in northeastern Tibet from Nd isotopes: implications for tectonic forcing of climate. Geol. Soc. Am. Bull. 117, 1156–1166. 4) The high hematite contents and redness records from cores HZ-2 and Giosan, L., Flood, R.D., Aller, R.C., 2002. Paleoceanographic significance of sediment HZ-1 confirm the onset of the Late Oligocene - Early Miocene color on western North Atlantic drifts: I. Origin of color. Mar. Geol. 189 (1–2), 25–41. monsoon that was indicated by the lithofacies. They reveal six short Guo, Z.T., Ruddiman, W.F., Hao, Q.Z., Wu, H.B., Qiao, Y.S., Zhu, R.X., Peng, S.Z., Wei, fi intervals of lower values (HR1 to HR6) with a roughly one million J.J., Yuan, B.Y., Liu, T.S., 2002. Onset of Asian deserti cation by 22 Myr ago inferred from loess deposits in China. Nature 416, 159–163. year cycle at approximately 22.9 Ma, 21.9 Ma, 21 Ma, 19.7 Ma, Guo, Z.T., Sun, B., Zhang, Z.S., Peng, S.Z., Xiao, G.Q., Ge, J.Y., Hao, Q.Z., Qiao, Y.S., 18.9 Ma and 18.1 Ma, respectively, which suggests six short weak Liang, M.Y., Liu, J.F., Yin, Q.Z., Wei, J.J., 2008. A major reorganization of Asian – monsoon events at those times. Both the long-term trend and the climate by the early Miocene. Clim. Past 4, 153 174. Hilgen, F., Lourens, L., Van Dam, J., 2012. The Neogene period. In: Gradstein, F.M., Ogg, short-term events of the monsoon records correlate well with global J.G., Schmitz, M.D., Ogg, G.M. (Eds.), The Geologic Time Scale. Elsevier, Amsterdam, and South China Sea marine oxygen isotope records, which suggests pp. 923–978. that the global temperature was a major forcing for early monsoonal Hu, S.Y., Goddu, S.R., Appel, E., Verosub, K., Yang, X.D., Wang, S.M., 2005. Palaeoclimatic changes over the past 1 million years derived from lacustrine sedi- climate change in the Late Oligocene and the Early Miocene at both ments of Heqing basin (Yunnan, China). Quat. Int. 136, 123–129. long and short time scales. Ji, J.F., Balsam, W., Chen, J., Liu, L.W., 2002. Rapid and quantitative measurement of hematite and goethite in the Chinese loess-paleosol sequence by diffuse reflectance spectroscopy. Clay Clay Miner. 50, 208–216. Acknowledgements Ji, J., Zhao, L., Balsam, W., Chen, J., Wu, T., Liu, L., 2006. Detecting chlorite in the Chinese loess sequence by diffuse reflectance spectroscopy. Clay Clay Miner. 54, We gratefully acknowledge the assistance in the field by Xiaohui 266–273. Ji, J.F., Chen, Y., Balsam, W., Liu, L.W., Chen, Y., Zhao, L., Zhou, W., 2007. Quantitative Fang and in the laboratory by Dr. Bingshuai Li. And we also acknowl- analysis of hematite and goethite in the Chinese loess-paleosol sequences and its edge the Editor Sierd Cloetingh, two anonymous reviewers and Dr. Yibo implication for dry and humid variability. Quat. Sci. 27, 221–229 (in Chinese with Yang for their valuable suggestions. This work is co-supported by the English abstract). (973) National Basic Research Program of China (2013CB956400), the Judd, D.B., Wyszecki, G., 1975. Color in Business, Science, and Industry. 553 John Wiley and Sons, New York. Strategic Priority Research Program of the Chinese Academy of Kirschvink, J.L., 1980. The least-squares line and plane and the analysis of palaeomag- Sciences (XDB03020103, XDB03020400) and the National Natural netic data. Geophys. J. Int. 62, 699–718. Science Foundation of China (41620104002, 41272183). Lease, R.O., Burbank, D.W., Clark, M.K., Farley, K.A., Zheng, D., Zhang, H., 2011. Middle Miocene reorganization of deformation along the northeastern Tibetan Plateau. Geology 39 (4), 359–362. References Li, J.J., et al., 1995. Uplift of Qinghai-Xizang (Tibet) Plateau and Global Change. Lanzhou University Press, Lanzhou. Li, J.J., Fang, X.M., Van der Voo, R., Zhu, J.J., MacNiocaill, C., Cao, J.X., Zhong, W., Balsam, W.L., Damuth, J.E., 2000. Further investigations of shipboard vs. shore-based Chen, H.L., Wang, J., Wang, J.M., 1997. Late Cenozoic magnetostratigraphy spectral data: implications for interpreting Leg 164 sediment composition. Proc. ODP (11–0 Ma) of the Dongshanding and Wangjiashan sections in the Longzhong Basin, – Sci. Results 164, 313 324. western China. Geol. Mijnb. 76, 121–134. Balsam, W.L., Deaton, B.C., 1991. Sediment dispersal in the Atlantic Ocean: evaluation by Li, X.Y., Fang, X.M., Yang, Y.B., Zan, J.B., 2012. Hematite record of 3 Ma Loess-red clay

45 F. Wu et al. Global and Planetary Change 158 (2017) 36–46

sequences in the central Chinese Loess Plateau 3 Ma and its paleoclimatic sig- Schwertmann, U., 1971. Transformation of hematite to goethite in soils. Nature 232, nificance. Quat. Sci. 32, 700–708 (in Chinese with English abstract). 624–625. Licht, A., Cappelle, M., Abels, H.A., Ladant, J.B., Trabucho-Alexandre, J., France-Lanord, Sobel, E.R., Chen, J., Heermance, R.V., 2006. Late Oligocene-Early Miocene initiation of C., Donnadieu, Y., Vanderberghe, J., Rigaudier, T., Lecuyer, G., Terry Jr., D., shortening in the Southwestern Chinese Tian Shan: implications for Neogene short- Adriaens, R., Boura, A., Guo, Z., Soe, A.N., Quade, J., Dupont-Nivet, G., Jaeger, J.J., ening rate variations. Earth Planet. Sci. Lett. 247 (1), 70–81. 2014. Asian monsoons in a late Eocene greenhouse world. Nature 513, 501–506. Song, C.H., Bai, J.F., Zhao, Y.D., Jin, H.B., Meng, Q.Q., 2005. The color of lacustrine Liu, T.S., Zheng, M.P., Guo, Z.T., 1998. Initiation and evolution of the Asian monsoon sediments recorded climatic changes from 13 to 4.5 Myr in Linxia Basin. Acta system timely coupled with the ice-sheet growth and the tectonic movement in Asia. Sedimentol. Sin. 23, 507–513 (in Chinese with English abstract). Quat. Sci. (3), 194–204 (in Chinese). Sun, X.J., Wang, P.X., 2005. How old is the Asian monsoon system?—Palaeobotanical Liu, Q.S., Torrent, J., Barr'on, V., Duan, Z.Q., Bloemendal, J., 2011. Quantification of records from China. Palaeogeogr. Palaeoclimatol. Palaeoecol. 222, 181–222. hematite from the visible diffuse reflectance spectrum: effects of aluminium sub- Sun, J.M., Windley, B.F., 2015. Onset of aridification by 34 Ma across the Eocene- stitution and grain morphology. Clay Miner. 46, 137–149. Oligocene transition in Central Asia. Geology 34. http://dx.doi.org/10.1130/ Liu, S.P., Li, J.J., Stockli, D.F., Song, C.H., Nie, J.S., Peng, T.J., Wang, X.X., He, K., Hui, G37165.1. Z.C., Zhang, J., 2015. Late Tertiary reorganizations of deformation in northeastern Sun, J.M., Ni, X.J., Bi, S.D., Wu, W.Y., Ye, J., Meng, J., Windley, B.F., 2014. Synchronous Tibet constrained by stratigraphy and provenance data from eastern Longzhong turnover of flora, fauna, and climate at the Eocene–Oligocene Boundary in Asia. Sci Basin. J. Geophys. Res. Solid Earth 120, 5804–5821. Rep 4 (7463), 1–6. Lu, H.J., Wang, E.Q., Shi, X.H., Meng, K., 2012. Cenozoic tectonic evolution of the Tian, J., Zhao, Q.H., Wang, P.X., Li, Q.Y., Cheng, X.R., 2008. Astronomically modulated Elashan range and its surroundings, northern Tibetan Plateau as constrained by pa- Neogene sediment records from the 24 South China Sea. Paleoceanography 23. leomagnetism and apatite fission track analyses. Tectonophysics 580, 150–161. http://dx.doi.org/10.1029/2007PA001552. Milankovitch, M., 1930. Mathematische klimalehre amd astronomishe Theorie der Torrent, J., Barrón, V., Liu, Q., 2006. Magnetic enhancement is linked to and pre- Klimasch-wankungen. In: Koppen, W., Geiger, R. (Eds.), Handbuch der Klimatologie. cedeshematite formation in aerobic soil. Geophys. Res. Lett. 33, L02401. vol. I(A). Gebruder Borntraeger, Berlin, pp. 1–76. Wang, P.X., 2009. Global monsoon in a geological perspective. Chin. Sci. Bull. 54, Mildenhall, D.C., 1989. Terrestrial palynology. In: Barrett, P.J. (Ed.), Antarctic Cenozoic 1113–1136. History From the CIROS-1 Drillhole, McMurdo Sound. DSIR Bulletin, vol. 245. pp. Wang, E.Q., Wan, J.L., Liu, J.Q., 2003a. Late Cenozoic geological evolution of the fore- 119–127. land basin bordering the West Kunlun range in Pulu area: constraints on timing of Miller, K.G., Fairbanks, R.G., Mountain, G.S., 1987. Tertiary oxygen isotope synthesis, sea uplift of northern margin of the Tibetan Plateau. J. Geophys. Res. Solid Earth 108 level history, and continental margin erosion. Paleoceanography 2, 1–19. (B8) (ETG l5-1). Miller, K.G., Wright, J.D., Fairbanks, R.G., 1991. Unlocking the ice house: Wang, P.X., Zhao, Q., Jian, Z., Cheng, X.R., Huang, W., Tian, J., Wang, J.L., Li, Q.Y., Li, Oligocene–Miocene oxygen isotopes, eustasy, and margin erosion. J. Geophys. Res. B.H., Su, X., 2003b. Thirty million year deep-sea records in the South China Sea. 96, 6829–6848. Chin. Sci. Bull. 48, 2524–2535. Nie, J.S., Song, Y.G., Wing, J.W., Fang, X.M., Heil, C., 2010. HIRM variations in the Wang, P.X., Clemens, S., Beaufort, L., Beaufort, L., Braconnot, P., Ganssen, G., Jian, Z.M., Chinese red-clay sequence: insights into pedogenesis in the dust source area. J. Asia Kershaw, P., Sarnthein, M., 2005. Evolution and variability of the Asian monsoon Earth Sci. 38, 96–104. system: state of the art and outstanding issues. Quat. Sci. Rev. 24, 595–629. Nie, J.S., Zhang, R., Necula, C., Heslop, D., Liu, Q., Gong, L., Banerjee, S., 2014. Late Wang, E.Q., Kamp, P.J.J., Xu, G.Q., Hodges, K.V., Meng, K., Chen, L., Wang, G., Luo, H., Miocene–early Pleistocene paleoclimate history of the Chinese Loess Plateau revealed 2015. Flexural bending of southern Tibet in a retro foreland setting. Sci Rep 5, 12076. by remanence unmixing. Geophys. Res. Lett. 41, 2163–2168. Wang, X.X., Song, C.H., Zattin, M., He, P.J., Song, A., Li, J.J., Wang, Q.Q., 2016. Cenozoic Nie, J.S., Garzione, C., Su, Q.D., Liu, Q.S., Zhang, R., Heslop, D., Necula, C., Zhang, S.H., pulsed deformation history of northeastern Tibetan Plateau reconstructed from fis- Song, Y.G., Luo, Z., 2017. Dominant 100,000-year precipitation cyclicity in a late sion-track thermochronology. Tectonophysics 672, 212–227. Miocene lake from northeast Tibet. Sci. Adv. 3, e1600762. White, A.F., Buss, H.L., 2014. Natural weathering rates of silicate minerals. In: Drever, J.I. Qiang, X., An, Z., Song, Y., Chang, H., Sun, Y., Liu, W., Ao, H., Dong, J., Fu, C., Wu, F., (Ed.), Surface and Ground Water, Weathering and Soils, Treatise on Geochemistry, 2011. New eolian red clay sequence on the western Chinese Loess Plateau linked to 2nd edition. vol. 7. Elsevier, pp. 115–155. onset of Asian desertification about 25 Ma ago. Sci. China Earth Sci. 54, 136–144. White, K., Walden, J., Gumey, S.D., 2007. Spectral properties, iron oxide content and Qiao, Y.S., Guo, Z.T., Hao, Q.Z., Yin, Q.Z., Yuan, B.Y., Liu, T.S., 2006. Grain-size features provenance of Namib dune sands. Geomorphology 86, 219–229. of a Miocene loess-soil sequence at Qinan: implications on its origin. Sci. China Ser. D Wright, J.D., Miller, K.G., Fairbanks, R.G., 1992. Early and middle Miocene stable iso- Earth Sci. 49, 731–738. topes: implications for deep water circulation and climate. Paleoceanography 7, Qiu, Z.X., Wang, B.Y., Deng, T., 2004. Mammal fossils from Yagou, Linxia Basin, Gansu, 357–389. and related stratigraphic problems. Vertebrat. PalAsiatic 42, 276–296. Xiao, G.Q., Abels, H.A., Yao, Z.Q., Dupont-Nivet, G., Hilgen, F.J., 2010. Asian aridifica- Quan, C., Liu, Y., Utescher, T., 2012. Eocene monsoon prevalence over China: a paleo- tion linked to the first step of the Eocene-Oligocene climate transition (EOT) in ob- botanical perspective. Palaeogeogr. Palaeoclimatol. Palaeoecol. 365–366, 302–311. liquity-dominated terrestrial records (Xining Basin, China). Clim. Past 6, 501–513. Raine, J.I., 1998. Terrestrial palynomorphs from Cape Roberts Project drillhole CRP-1, Xiong, Y., Li, Q.K., 1987. Chinese Soil. Science Press, Beijing. Ross Sea, Antarctica. Terra Antarct. 5, 539–548. Yin, A., Nie, S., Craig, P., Harrison, T.M., Ryerson, F.J., Qian, X.L., Yang, G., 1998. Late Raine, J.I., Askin, R.A., 2001. Terrestrial palynology of Cape Roberts Project drillhole Cenozoic tectonic evolution of the southern Chinese Tian Shan. Tectonics 17 (1), CRP-3, Victoria Land Basin, Antarctica. Terra Antarct. 8, 389–400. 1–27. Reading, H.G., 2009. Sedimentary Environments: Processes, Facies and Stratigraphy, 3rd Zachos, J.C., Pagani, M., Sloan, L., Thomas, E., Billups, K., 2001. Trends, rhythms, and ed. Wiley-Blackwell (704 pp.). aberrations in global climate 65 Ma to present. Science 292, 686–693. Retallack, G.J., 2001. Soil of the Past: An Introduction to Paleopedology. Blackwell Zachos, J.C., Gerald, R.D., Richard, E.Z., 2008. An early Cenozoic perspective on green- Science. house: warming and carbon-cycle dynamics. Nature 451 (17), 279–283.