Holocene water-level changes inferred from a section of fluvio-lacustrine sediments in the southeastern Mu Us Desert, China

Quaternary International xxx (2017) 1e10

Contents lists available at ScienceDirect

Quaternary International journal homepage: www.elsevier.com/locate/quaint

Holocene water-level changes inferred from a section of fluviolacustrine sediments in the southeastern Mu Us Desert, China Xiaokang Liu a, b, c, Ruijie Lu a, b, c, *, 1, Feifei Jia d, Lu Chen a, b, c, Tengfei Li a, b, c, Yuzhen Ma a, b, c, Yongqiu Wu a, b, c a

MOE Key Laboratory of Environmental Change and Natural Disaster, Beijing Normal University, Beijing, 100875, China State Key Laboratory of Earth Surface Processes and Resource Ecology, Beijing Normal University, Beijing, 100875, China MOE Engineering Center of Desertification and Blown-sand Control, Beijing Normal University, Beijing, 100875, China d School of Urban and Environmental Sciences, Liaoning Normal University, Dalian, 116029, China b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 7 March 2016 Received in revised form 19 December 2016 Accepted 26 December 2016 Available online xxx

A natural exposure of fluvio-lacustrine sediments (the DSGL section) on the left bank of Salawusu River, located at the southeastern margin of the Mu Us Desert in North China, was studied in order to reconstruct Holocene water-level changes. A chronology was established based on 11 AMS 14C dates, and variations in the lithology and grain-size C-M variations were used to assess the forcing factors of local hydrological changes during the Holocene. A lake-swamp environment with high hydrodynamic energy occurred from 9.6 to 9.2 ka cal yr BP, which was succeeded by a full lake environment at around 9.6 ka cal yr BP; however, an interval of sand deposits and peat sediments from 9.2 to 8.6 ka cal yr BP indicates unstable hydrodynamic conditions. The Salawusu paleo-lake reached its maximum level between 8.4 and 6.5 ka cal yr BP and gradually shrank thereafter. After 1.2 ka cal yr BP a shallow lake environment returned, indicating a humid phase which may be correlative with the relatively warm and moist Medieval Warm Period (MWP). After 0.68 ka cal yr BP, broadly coeval with the Little Ice Age (LIA), a fluvial environment appeared. The variations in water level were mainly triggered by the East Asian monsoon; however, site-specific factors may have also have been important. © 2016 Elsevier Ltd and INQUA. All rights reserved.

Keywords: Holocene Fluvio-lacustrine sediments Mu Us Desert Grain-size C-M patterns Hydrological forcing conditions

1. Introduction The Mu Us Desert in Northern China is situated in the loessdesert transitional zone, on the northern margin of the East Asian monsoon (EAM). The area is climatically sensitive and is therefore suitable for studies of long-term paleoclimatic and paleoenvironmental changes (Sun et al., 1999; Zhou et al., 2002, 2009; Ding et al., 2005). In particular, the transitional area at the desert boundary provides an excellent record of phases of desert expansion and retreat (Sun et al., 1999; Yang et al., 2004), while variations in water level reflect changes in the strength of hydrodynamic forcing factors such as the EAM. The Mu Us Desert has accumulated a diverse range of sediments, including aeolian sand, paleosol, loess and peat, which are useful for reconstructing the history of desert

* Corresponding author. MOE Key Laboratory of Environmental Change and Natural Disaster, Beijing Normal University, Beijing, 100875, China. E-mail address: [email protected] (R. Lu). 1 This author contributed equally with the first author to this work.

development on local and regional scales (Li et al., 1998). Compared with aeolian sediments, fluvio-lacustrine sediments can provide more direct evidence of water-level changes; moreover, variations in hydrological conditions may dominate the sedimentary effects of aeolian activity in this region. Fluvio-lacustrine sediments are potentially useful archives of paleoenvironmental information because of their continuous accumulation and suitability for accurate dating (Wagner et al., 2000; Roberts et al., 2001; Chen et al., 2002; Wolfe et al., 2004; Eris, 2013). The multi-proxy indices typically analyzed in fluviolacustrine sediments include grain-size (Solohub and Klovan, 1970; Chen et al., 2004; Lu et al., 2011; Zhang et al., 2014), geochemistry (Leng and Marshall, 2004; Li et al., 2008; Wuennemann et al., 2010; Mishra et al., 2015), and pollen assemblages (Bigler et al., 2002; Shen et al., 2005; Brisset et al., 2015). Overall, fluvio-lacustrine sediments are sensitive to hydroclimatic changes. The Salawusu River Valley is located on the southeastern margin of the Mu Us Desert and the Salawusu Formation found in this

http://dx.doi.org/10.1016/j.quaint.2016.12.032 1040-6182/© 2016 Elsevier Ltd and INQUA. All rights reserved.

Please cite this article in press as: Liu, X., et al., Holocene water-level changes inferred from a section of fluvio-lacustrine sediments in the southeastern Mu Us Desert, China, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2016.12.032

2

X. Liu et al. / Quaternary International xxx (2017) 1e10

region has been regarded as a standard site for fluvio-lacustrine strata in China for the Upper Pleistocene (Dong et al., 1999) and since its discovery in the early 1920s numerous stratigraphic (e.g., Pei and Li, 1964; Yuan, 1978; Dong et al., 1983; Sun et al., 1996) and paleoclimatic (e.g., Gao et al., 1985; Li et al., 2005; Du et al., 2012; Rao et al., 2013) investigations have been carried out. However, previous work has mainly been focused on long-term climatic evolution and only a small number of studies (Li et al., 2012a; Liu and Lai, 2012) have emphasized water-level reconstruction during the Holocene based on fluvio-lacustrine deposits. Therefore, more work is needed to reconstruct water-level variations during the Holocene. Grain-size C-M patterns (Passega, 1957, 1964) provide insights into the hydrodynamic conditions of sedimentation and this information can potentially be used to infer changes in climatic controls on the sedimentary environment. In the present paper, we select a new section which is situated at Dishaogouwan on the left bank of Salawusu River and use grain-size C-M patterns, combined with sedimentary facies characteristics, to reconstruct changes in hydrodynamic conditions and the sedimentary environment, especially variations in water-level, during the Holocene.

232 samples were collected from the DSGL section at 2 cm or 5 cm intervals for grain-size analyses. The measurements were made using a Mastersizer 2000 with a measurement range of 0.02e2000 mm at the Key Laboratory of Environmental Change and Natural Disasters, Ministry of Education of China, Beijing Normal University. The pretreatment procedure is described by Jia et al. (2015).14 samples of organic sediments were collected, according to their lithological characteristics, for AMS 14C dating. The measurements were made at the Beta Analytic Radiocarbon Dating Laboratory and detail of the pretreatment procedure is at http:// www.radiocarbon.com/pretreatment-carbon-dating.htm#Washes. The AMS 14C dates were converted to calendar ages using the program Calib 7.02 based on the INTCAL 13 calibration (Reimer et al., 2013).

2. Regional setting

4. Results

The Mu Us Desert, with an area of about 39,000 km2, is situated on the margin of the East Asian monsoon (EAM), in the transitional area between the Ordos Plateau and the Loess Plateau (Fig. 1). The regional climate is dominated by the EAM. The mean annual rainfall is 250e440 mm and the mean annual temperature is 6.0e8.5
C. Warm and humid air brought by the East Asian summer monsoon (EASM) delivers more than 60% of the annual precipitation, which falls mainly from June to August. The modern vegetation consists mainly of Artemisia ordosica Krasch., Tamarix chinensis Lour., and Hippophae rhamnoides Linn. The regional geomorphic types are varied and include stabilized, semi-stabilized and active sand dunes, shallow lake basins and seasonally-flowing streams, and loess-mantled hills and sandy loess platforms (Zhou et al., 1996). A natural exposure of fluvio-lacustrine strata, the DSGL (Dishaogouwan Left bank) section (37
430 N, 108
310 E), 1300 m above

4.1. Chronology

sea level and on the left bank of the Salawusu River in the southeastern margin of the Mu Us Desert, was chosen for study. The sequence is 4.04 m thick (Fig. 1a and b). 3. Materials and methods

The AMS 14C dates are listed in Table 1. The ages are generally in chronological order, although there are some inversions: at the base of the section which probably result from involutions, possibly caused by freeze-thaw processes; and at the top of the section where there are three dates which were clearly anomalously old and were excluded from the final age model. The uppermost part of the section contains reworked material (section 4.2) and the three anomalous dates may be the result of the incorporation of ancient organic material. We established a final age model using piecewise linear fitting (Fig. 2) and the age of each sample was determined by linear interpolation. Viewed in the context of previous dating results (Fig. 3) from the adjacent region we are confident that chronological framework is reliable.

Fig. 1. (a) Location of the DSGL section on the southern margin of the Mu Us desert (orange shaded area). The green dashed line indicates the modern Asian summer monsoon limit and the inset map shows the location of the study area in China. (b) The location of DSGL section. (c) The involutions at the bottom of the section. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: Liu, X., et al., Holocene water-level changes inferred from a section of fluvio-lacustrine sediments in the southeastern Mu Us Desert, China, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2016.12.032

X. Liu et al. / Quaternary International xxx (2017) 1e10

3

Table 1 14 C dates for the DSGL section. Sample No.

Laboratory No.

Depth (cm)

13

C/12C Ratio

Radiocarbon age (yr BP)

Calendar year (2s, cal yr BP)

DSGL14C09 DSGL14C01 DSGL14C10 DSGL14C17 DSGL14C18 DSGL14C02 DSGL14C04 DSGL14C05 DSGL14C08 DSGL14C11 DSGL14C12 DSGL14C13 DSGL14C15 DSGL14C16

beta-387097 beta-387083 beta-387088 beta-387094 beta-387095 beta-387084 beta-387085 beta-387086 beta-387087 beta-387089 beta-387090 beta-387091 beta-387092 beta-387093

17e18 54.5e56.5 144e146 149e151 159e161 173e176 209e212 234e235 274e275 284e285 299e300 312e313 376e377 384e385

20.8‰ 22.3‰ 21.6‰ 24.4‰ 24.7‰ 25.1‰ 24.7‰ 23.9‰ 24.1‰ 24.5‰ 24.5‰ 24.2‰ 24.6‰ 24.8‰

6090 ± 30 5820 ± 30 7230 ± 30 490 ± 30 740 ± 30 710 ± 30 1310 ± 30 2200 ± 30 4890 ± 30 6090 ± 30 7140 ± 30 6770 ± 30 8530 ± 30 7480 ± 30

7015 ± 130 6615 ± 60 8065 ± 90 522.5 ± 17.5 680 ± 20 627.5 ± 57.5 1237.5 ± 57.5 2225 ± 95 5622.5±32.5 7015 ± 130 7970 ± 35 7625 ± 45 9515 ± 25 8290 ± 85

4.2.2. Stage B (389-328 cm, 9.6e8.6 ka BP) The lithology consists predominantly of silt and clay, reflected by the significant fining of the grain-size parameters and indicating the onset of the warm, moist conditions of the early Holocene. It is subdivided into the following two sub-stages.

Fig. 2. Age-depth model for the DSGL section, based on AMS 14C dating. The three dates shown as magenta triangles were excluded from the linear fitting (green lines). 2s calibrated age ranges are shown for each age and the red lines are the 95% confidence interval. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

4.2. Lithology and grain-size analyses The lithology and stratigraphic profiles of grain-size parameters are illustrated in Fig. 3 and Table 2 while representative grain-size frequency distribution curves, together with curves for samples from selected modern environments, are illustrated in Fig. 4. The grain-size distribution of lacustrine sediments is closely linked with a lake's hydrodynamic status (Li et al., 2012b) and in order to summarize the main phases of the site's hydrological evolution the record was divided into five stages, labeled A to E, which are summarized below.

4.2.1. Stage A (404-389 cm, prior to 9.6 ka BP) The lithology consists mainly of very pale brown fine sand which contains scattered yellow rust spots which may be indicative of post-depositional waterlogging. The grain-size distribution of a representative sample is unimodal (Fig. 4a) and similar to that of a modern active sand dune sample from the southeastern margin of the Mu Us Desert. Thus we infer that this part of the Mu Us Desert was an aeolian dune environment prior to 9.6 ka BP, resembling the modern active dune landscape.

4.2.2.1. Sub-stage B1 (389-366 cm, 9.6e9.2 ka BP). The lithology consists mainly of dark gray very fine sand, but with the proportion of clay and silt increasing significantly upwards. In contrast, MZ, D50 and modal grain-size exhibit a decreasing trend. A representative grain-size frequency distribution curve (Fig. 4b) is similar to that of a sample from a modern grassland-swamp environment. The grainsize C-M patterns (Fig. 5, B1) indicate that the primary mode of transport of the sediment is as a graded suspension with a rolling component. The strong ability to transport particles is one reflection of an unstable sedimentary environment, with fluctuations between high energy and lower energy conditions. Based on these characteristics we speculate that the paleoenvironment was a swamp or incipient swamp. 4.2.2.2. Sub-stage B2 (366-338 cm, 9.2e8.6 ka BP). The sediments mainly consist of pale brown very fine sand with an average mean grain-size size of 90 mm. The grain-size frequency distribution curve (Fig. 4c) is dissimilar to that of the aeolian sand of stage A and we infer a hydrodynamic influence. It is possible that the main water body was blocked by sand dunes. Both B1 and B2 are possibly affected by involutions (Figs. 1c and 3). 4.2.3. Stage C (338-255 cm, 8.6e4.0 ka BP) The clay content is relatively high and the MZ value remains at a low and stable level. The sediments consist primarily of lacustrine deposits and the hydrodynamic conditions were low in energy and stable. Four sub-stages are recognized: 4.2.3.1. Sub-stage C1 (338-328 cm, 8.6e8.4ka BP). The lithology consists of pale gray very fine sand and is a transitional layer between B2 and C2. 4.2.3.2. Sub-stage C2 (328-303 cm, 8.4e7.8 ka BP). The sediments consists of light gray very fine sand. Mz and D50 exhibit a gradually decreasing trend from C1 to C2. The grain-size C-M patterns change systematically from stage C1 to C2 (Fig. 5, C1 and C2) and the points plot towards the “SR” zone. These characteristics indicate a change from a graded suspension to a uniform suspension, thus indicating a decrease in hydrodynamic energy. We infer that sub-stages C1 and C2 represent a change in the sedimentary environment from a sandy, littoral lake margin to a significantly deeper-water

Please cite this article in press as: Liu, X., et al., Holocene water-level changes inferred from a section of fluvio-lacustrine sediments in the southeastern Mu Us Desert, China, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2016.12.032

4

X. Liu et al. / Quaternary International xxx (2017) 1e10

Fig. 3. Lithology, percentages of clay, silt and sand, selected grain-size parameters (Mz, D50 and mode), previous dating results (red - Liu and Lai (2012); dark blue- Su and Dong (1996)) and a photograph of the DSGL section. Horizontal shaded bands indicate possible abrupt climatic events recorded in the section. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Table 2 Lithological description of the DSGL Section. Unit number

Depth(cm)

Lithology description

Munsell color

A

404e389

Very pale brown (10YR 7/4)

B1 B2 C1 C2 C3 C4 D1 D2 D3 D4 E

389e366 366e338 338e328 328e303 303e281 281e255 255e238 238e209 163e209 163e147 147e0

Loose fine aeolian sand containing scattered yellow rust spots which may be indicative of waterlogging. Not bottomed. Limnetic very fine sand affected by involutions. Relatively loose very fine sand affected by involutions. Transitional layer from sand to lacustrine deposits. Very coarse lacustrine silt. Very coarse lacustrine, rich in yellow rust spots. Very coarse lacustrine silt, with abundant freshwater gastropod fossils. Transitional layer from lacustrine to peat deposits. Loose peat with crumb structure. Very coarse lacustrine silt and coarse silt. Very fine fluvial sand. Alluvial or diluvium deposits

environment closer to the lake centre.

4.2.3.3. Sub-stage C3 (303-281 cm, 7.8e6.5 ka BP). The sediments are dark in colour and consist mainly of silt, and there is a striking peak in the clay content. There is a dense distribution of grain-size C-M points (Fig. 5, C3). We infer that this was an interval of high and stable water level.

4.2.3.4. Sub-stage C4 (281-255 cm, 6.5e4.0 ka BP). The sediments are massive and contain freshwater gastropod fossils; in addition, the clay content tends to decrease and the silt content to increase. The two red dots and arrow in the C-M plot (Fig. 5, C4) indicate particle movement. Based on these characteristics we infer that the mode of particle movement changed from a graded suspension to a uniform suspension, which was triggered by continuously decreasing hydrodynamic energy, and that the sub-stage represents a transition from a deep water lake to a shallow lake environment, with the initiation of marshy conditions.

Dark gray (10YR4/1) Very pale brown (10YR 7/4) Light gray (10YR 7/1) Gray (10YR 6/1) Dark yellowish brown(10YR 4/4) Gray (10YR 5/1) Gray (10YR 6/1) Gray(10YR 5/1) Gray(10YR 6/1),gray(10YR 5/1) Brown (10YR 5/3) Pale brown (10YR 6/3), very pale brown (10YR 7/3)

4.2.4. Stage D (255-147 cm, 4.0e0.52 ka BP) The sediments mainly consist of gray limnetic deposits and there are significant variations in grain-size. Four sub-stages are recognized: 4.2.4.1. Sub-stage D1 (255-238 cm, 4.0e2.5 ka BP). This sub-stage is an extension of sub-stage C4 and consists predominantly of gray silt. There is a strong trend of decreasing MZ values, and the grainsize C-M distribution (Fig. 5, D1) indicates a uniform suspension. Based on these characteristics we infer a drastic change in the sedimentary environment with the onset of stable and low energy hydrodynamic conditions. 4.2.4.2. Sub-stage D2 (238-209 cm, 2.5e1.2 ka BP). The lithology consists of peat and the grain-size parameters exhibit low values. The distribution of points in the grain-size C-M plot (Fig. 5, D2) indicates a uniform suspension mode of particle transport, indicating stable and weak hydrodynamic conditions. The sub-stage represents a continuation of the trend of decreasing lake size

Please cite this article in press as: Liu, X., et al., Holocene water-level changes inferred from a section of fluvio-lacustrine sediments in the southeastern Mu Us Desert, China, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2016.12.032

X. Liu et al. / Quaternary International xxx (2017) 1e10

5

Fig. 4. Representative grain-size frequency distribution curves for the DSGL section. Plots a) and b) include grain-size frequency distribution curves for samples from modern active sand dune and grassland-swamp environments, respectively.

represented in sub-stage D1. At the end of the sub-stage the points (triangles) in the C-M plot indicate an increase in hydrodynamic energy, which may reflect a rising water level.

the shallow lake decreased in size and became a river. The age range of this sub-stage is AD 1270-AD1430, roughly coeval with the Little Ice Age (1280-1860AD).

4.2.4.3. Sub-stage D3 (209-163 cm, 1.2e0.68 ka BP). The lithology consists of grey silty and sandy lacustrine sediments. There are significant fluctuations in the grain-size parameters and the sand content increases. The grain-size C-M distribution begins at a transitional layer (green dots) representing a change from a swamp environment in D2 to a lake environment in D3. The distribution of green dots indicates a uniform suspension and weak hydrodynamic energy. Subsequently, as shown by the distribution of red dots, the hydrodynamic energy commences an increasing trend, and finally towards the end of the stage the C-M distribution indicates a change from a uniform to a graded suspension (black dots). We speculate that during sub-stage D3 a shallow lake existed and that it underwent minor fluctuations in level (Fig. 4d, D31-D33; Fig. 6D, navy blue curve). The age range of this sub-stage is AD750-AD1270, roughly coeval with the Medieval Warm Period (MWP, 6001280AD).

4.2.5. Stage E (147-0 cm) The lithology consists predominantly of brown silt containing reworked material. It is noteworthy that three inverted radiocarbon ages occur within this interval. The distribution of points in the C-M plot (Fig. 5E) is highly scattered; the distribution contains point clusters indicating a uniform suspension, a graded suspension and pelagic suspension. The sediments may consist of alluvium or diluvium.

4.2.4.4. Sub-stage D4 (163-147 cm, 0.68e0.52 ka BP). The lithology consists mainly of brown sand, with the sand content increasing sharply and with the grain-size parameters exhibiting significant fluctuations. The distribution of points in the C-M plot (Fig. 5, D4) is centered on the graded suspension region where most of the coarse-grained samples are located. The C and M values vary in proportion indicating that the particle motion consisted of saltation in an environment with high hydrodynamic energy. We infer that

4.3. Inferred changes in lake-level The alternations of aeolian sand, fluvio-lacustrine sediments and peat within the study section (Fig. 4d) clearly indicate significant changes in hydrological conditions. Aeolian sand beds often exhibit cross-bedding indicating active sand dunes and an absence of hydrodynamic influences. Fluvio-lacustrine sediments represent a relatively humid environment, with the occurrence of lake-level fluctuations. The development of peat requires suitable local moisture conditions and a low water-level. Based on the combination of the inferred relative water depth obtained from the grain-size C-M patterns and the characteristics of the sedimentary facies, we produced a semi-quantitative record of past water level fluctuations based on inferred changes in hydrological regime (Zheng, 1989). The record was constructed as follows. (i) The sand layers of stages A and B2 are assigned values of

Please cite this article in press as: Liu, X., et al., Holocene water-level changes inferred from a section of fluvio-lacustrine sediments in the southeastern Mu Us Desert, China, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2016.12.032

6

X. Liu et al. / Quaternary International xxx (2017) 1e10

Fig. 5. Grain-size C-M patterns for the various hydrological stages defined for the section. Passage (1964) distinguished several modes of sediment transport based on point distributions. ON: coarse particles transported by rolling; PO: coarse particles transported by both rolling and in suspension; QP: finer particles transported by both rolling and in suspension; RQ: graded suspension; SR: uniform suspension; T: pelagic suspension.

0 and 0.5, respectively, since the site was essentially a sub-aerial environment; (ii) during the swamp (B1, D1 and D2) and river phases the water-level was low and a value of 1 was assigned; (iii) shallow lake and deep lake phases were assigned values of 2 and 4, respectively; (iv) for the remaining phases, the relative water-level changes are based on analysis of grain-size C-M patterns in the previous section. The result is shown in Fig. 6D (blue curve). 5. Discussion 5.1. Changes in hydrological conditions during the Holocene and regional correlations The study area became an aquatic environment after ~9.6 ka BP, indicated by the appearance of limnetic sediments. The stratigraphic record from Midiwan, located about 40 km east of our studied section, records peat deposition after 10 ka BP, indicating a warm and moist early Holocene climate (Zhou et al., 1996). The

presence of a sand layer dated from ~9.2e8.6 ka suggests a fall in water-level. A detailed proxy precipitation record from Gonghai Lake in central-north China (Fig. 6C, blue curve), within the EAM region, also records a transient dry interval. The Salawusu paleolake reached its maximum level during 8.4e6.5 ka BP (Fig. 6D, brown curve) with relatively stable hydrological conditions. The lake level gradually declined (Fig. 6D, navy blue curve), which lasted until 4.0 ka BP, and subsequently the lake never attained its former level, even though a relatively high lake-level occurred during the MWP (Stage D3). Our evidence for an interval of relatively high lake-level from 8.4 to 4.0 ka BP is broadly consistent with that from adjacent monsoonal areas in China, such as the existence of a lake environment from 8.3 to 5 ka BP recorded in the Dagouwan section in the Salawusu River Valley (Li et al., 2012a), the occurrence of intensified but highly variable monsoonal precipitation from 7.9 to 3.1 ka BP at Daihai Lake, in north-central China(Peng et al., 2005), a precipitation peak at 8-3ka BP from Chinese loess deposits (Lu et al., 2013) and some evidences from aeolian

Please cite this article in press as: Liu, X., et al., Holocene water-level changes inferred from a section of fluvio-lacustrine sediments in the southeastern Mu Us Desert, China, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2016.12.032

X. Liu et al. / Quaternary International xxx (2017) 1e10

7

Fig. 6. Variations in estimated water-level of the DSGL section over the last 12 ka and selected environmental proxies. (A) Northern Hemisphere July insolation at 65
N (red curve) (Berger and Loutre, 1991); and summer air temperature for MS2008E in Central China (dark cyan curve) (Peterse et al., 2011). (B) d18O records from Sanbao Cave (SB26 purple curve; SB10 orange curve) (Wang et al., 2008) and Dongge Cave (green curve) (Wang et al., 2005). (C) Reconstructed annual precipitation from Gonghai Lake, North China (blue curve) (Chen et al., 2015a); reconstructed precipitation at the ZBT site (pink curve) (Chen et al., 2015b). (D) Relative water level (blue curve) and clay content (brown curve) for the DSGL section. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

sediments which record widespread dune stability and soil development from about 8.4 to 4.0 ka BP (Mason et al., 2009; Lu et al., 2015; Jia et al., 2015) in Northern China. An abrupt change in hydrological conditions occurred at around 4.0 ka BP, with the development of a transient swamp environment indicated from ~4.0e1.2 ka BP. An abrupt climatic change is also documented at Sanbao Cave (Fig. 6B purple) at around 4.0 ka. A dated record from the neighboring DSG section revealed that after ~3 ka BP the paleo-lake shrank further, indicated by the occurrence of a peat layer (Liu and Lai, 2012). Dry conditions also prevailed after 3.0 ka BP at Midiwan (Li et al., 2003) while pollen evidence from Daihai Lake, located about 450 km from the DSGL section, indicates that from 4.45 to 2.9 ka BP temperature, precipitation and the occurrence of woody plants decreased (Xiao et al., 2004). It is noteworthy that two stages within the DSGL section are correlative with the Medieval Warm Period (MWP, stage D3) and the Little Ice Age (LIA, stage D4). A shallow lake environment briefly

reappeared from 1.2 to 0.68 ka BP (stage D3), when minor environmental fluctuations were observed. A similar climatic oscillation based on high-amplitude fluctuations in TOC content at the Jingbian section, close to our research area, occurred after 1.5 ka BP (Xiao et al., 2002); and grassland/wetland conditions occurred during the Song and Yuan Dynasty (Hu et al., 2011). After 0.68 ka BP, the sedimentary characteristics of the DSGL section indicate significantly increased hydrodynamic energy suggesting that a fluvial environment occurred. However, the sedimentary environment was somewhat different at other sites, such as at the DSG section where the coeval layer is a paleosol (Liu and Lai, 2012). In other words, after stage D3, the shrinking of the lake led to the formation of many seasonal streams, similar to braided channels. All of these results relating to the MWP and LIA are in agreement with temperature reconstructions based on phenological records in China (Chu, 1973). From a spatial perspective, the distribution of lake or playas in the modern Mu Us Desert resembles a series of ‘beads-on-strings’

Please cite this article in press as: Liu, X., et al., Holocene water-level changes inferred from a section of fluvio-lacustrine sediments in the southeastern Mu Us Desert, China, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2016.12.032

8

X. Liu et al. / Quaternary International xxx (2017) 1e10

and in addition the variation of lake area and water-level is mainly controlled by the rainfall amount. Therefore, we suspect that similar cases occurred in the study area during the Holocene. During MIS 5 the Salawusu paleo-lake, and the corresponding sediments of the Salawusu Formation, was extensively developed but it did not consist of a continuous water body (Dong et al., 1983). In the Holocene the paleo-lake reappeared, but it did not regain its former size and instead developed in the form of scattered, discontinuous lake-swamp environments within inter-dune swales (Jia, 1996). Numerous field investigations have revealed that Holocene fluvio-lacustrine sediments are well exposed in our study area and we also observed several abrupt changes in sedimentary facies in locations close to the DSGL section. This explains why a ‘bead-shaped’ distribution of lakes existed in the study area during the Holocene Climatic Optimum period, similar to the modern status mentioned above.

5.2. Possible forcing mechanisms The DSGL section records signals of varying temporal scale, from centennial to millennial. In order to understand the geomorphic processes involved, we now consider the possible forcing mechanisms, based on the hydrological conditions. Lake level is influenced by runoff from the watershed and by evaporation. Increasing runoff generally produces an increase in lake level (Harrison et al., 1993). Lake-level variations in the study area are inevitably related to some extent to monsoonal activity. Orbital variations drive changes in summer insolation which lead to thermal contrasts between land and sea (An et al., 2000). This contrast influences the strength of the monsoon, in turn causing variations in rainfall. Ultimately, changes in rainfall amount cause the water-level fluctuations. The present study area lies near the northern limit of the EASM, which has a strong influence on the area and on the levels of the lakes. A stronger EASM delivers more precipitation, resulting in a denser vegetation cover; and conversely a weak EASM results in a drier and colder climate, resulting in a decrease in vegetation cover (An et al., 1991). Based on Fig. 6A, at the onset of Holocene, strong Northern Hemisphere insolation and higher temperatures resulted in an enhanced EASM intensity, which is also reflected by the speleothem d18O records shown in Fig. 6B. The monsoon intensity then peaked at around 9.0 ka BP, as indicated by the stalagmite records from Sanbao Cave (Fig. 6B, purple curve). Hence the EASM commenced a weakening trend after about 7.0 ka BP, and speleothem d18O values, precipitation and lake-levels declined gradually (Fig. 6B, C and D). The relatively stable Holocene lacustrine phase in the study area culminated at 8.4e4.0 ka BP, probably as a direct result of the intensity of monsoonal circulation and the consequent strong precipitation. Paleosol formation requires an interval with an optimum and stable climate (Lu et al., 2015), and therefore paleosol development in aeolian sections may not occur instantaneously in response to a strengthening of ASM. In contrast, the water-level in lake catchments is highly sensitive to EASM changes. These differences in response time may be the reason for the controversy which exists regarding whether the early Holocene climate of the Mu Us Desert was arid, as recorded by aeolian sand (Lu et al., 2005, 2011; Mason et al., 2009), or humid, as recorded by fluvio-lacustrine sediments (Li et al., 2003; Liu and Lai, 2012). Given the differences in topography of individual sites and the uncertainties in the proxy data, a greater number of internal and external driving factors need to be explored, including non-climatic factors (Gasse and Van Campo, 1994): for example, several watershed-scale factors are considered below.

5.3. Origin of the phases of swamp development There is evidence for two episodes of swamp development at the study site, the first at the beginning of the Holocene and the second at around 4.0 ka BP. These episodes are reflected by expansion and shrinkage of the water body and the two different types of swamp produced are significant for the interpretation of the fluctuations in hydrological conditions and climate change feedbacks. A key question is what factor(s) triggers the onset of a deep lake environment and the accumulation of limnetic sediments. A striking peak in precipitation occurred between 10 and 9.0 ka BP, as indicated by the reconstructed annual precipitation record (Fig. 6C) and this was directly responsible for the accumulation of limnetic sediments. In addition, the permafrost degradation has a considerable influence on the increased fluvial activity (Xu et al., 2015). At the early Holocene, followed by the increase in temperature and insolation (Fig. 6A, dark cyan curve),the effect of melting permafrost in areas to the north of the Loess Plateau (Cui et al., 2004) resulted in enhanced runoff. Moreover, locally improved drainage with permafrost degradation could also have increased the potential for sand deflation and fluvial activity (Xu et al., 2015). Rapid climatic changes can trigger abrupt geomorphic responses (Perry and Hsu, 2000). The aforementioned process of swamp development at the start of the Holocene was also a relatively rapid feedback response, which resulted in the deposition of the weaklydeveloped limnetic sediments. In contrast, the gradual transition from a lacustrine to a swamp environment from 6.2 to 4.0 ka BP represents a significant fall in the water level. The clay content of the section (Fig. 6D, brown curve) decreased dramatically at around 4.0 ka BP and this event was largely responsible for the disappearance of the lake, and in addition corresponded to the transition from the mid-Holocene Optimum to the more arid late Holocene (4.0 ka BP-present). In addition, it is important to note that the lacustrine deposits were the key for the subsequent initiation of peat development. As well as constituting a water-retaining ecosystem for peatland development, the lake sediments would also have provided a nutrient-rich substrate for vegetation growth as well as a habitat for gastropods. These factors may explain why the two types of peat sediment (sub-stage B1 and D2, Fig. 3) are broadly similar but also exhibit significant differences. In addition, differences in local temperature and moisture conditions and their duration would lead to differences in the degree of swamp development. Although climate may have been the main driver of the changes from a lake to a swamp environment, watershed-scale factors may also have been responsible, such as lake infilling caused by sediment accumulation. 5.4. Possible evidence for the 8.2ka event The 8.2 ka BP cooling event is thought to have been the coldest interval during the Holocene (Alley et al., 1997). Although there is no clear registration of the event in the study section, a layer of very fine sand is recorded at ~9.2e8.6 ka BP, which is disturbed by involutions (Fig. 1c). It is possible that that the sand layer is a registration of the 8.2 ka BP cooling event; however, further evidence is needed to test this speculation. 6. Conclusions Using a chronology based on radiocarbon dating, we have used variations in lithology and grain-size C-M patterns of a Holocene sedimentary sequence on the southern margin of the Mu Us desert to infer changes in hydrological forcing conditions and to reconstruct changes in water-level. The results indicate that a swamp

Please cite this article in press as: Liu, X., et al., Holocene water-level changes inferred from a section of fluvio-lacustrine sediments in the southeastern Mu Us Desert, China, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2016.12.032

X. Liu et al. / Quaternary International xxx (2017) 1e10

environment occurred during two intervals, from 9.6 to 9.2 ka BP and from 4.0 to 1.2 ka BP. A lacustrine environment occurred from 8.4 to 4.0 ka BP and from 1.2 to 0.68 ka BP, possibly a registration of the MWP. A fluvial environment may have occurred from 0.68 to 0.52 ka BP, possibly a registration of the LIA. During the early Holocene (9.6e8.4 ka BP) the site accumulated peat and sand and experienced unstable hydrological conditions and an overall low water-level. A sand layer dated to 9.2e8.6 ka BP, together with peat deposits, are indicative of unstable hydrodynamic conditions and confirm that the site was not a continuous lake environment throughout the Holocene. Lake sediments accumulated at the beginning of the mid-Holocene (8.4e4.0 ka BP). The Salawusu paleo-lake reached its maximum level from 8.4 to 6.5 ka BP and then gradually shrank. After 1.2 ka BP, a shallow lake returned, characterized by frequent fluctuations in lake level. From 0.68 to 0.52 ka BP, a fluvial environment with a relatively low water-level occurred. The water-level variations were mainly triggered by changes in the intensity of the East Asian summer monsoon, but they may also reflect the role of non-climatic factors. Acknowledgments We thank Professor Gao Shangyu and Tao Mingxin for detailed and constructive suggestions, Li Jinfeng for assistance in the field, and Jan Bloemendal for English improvement. We are grateful to Bai Qingyuan of Uxin Banner’s Cultural Heritage Bureau for the full supports. Special thanks to the anonymous reviewers who have significantly improved this paper. This study was supported by the National Natural Science Foundation of China (No.41330748), and the Beijing Higher Education Young Elite Teacher Project (YETP0261). References Alley, R.B., Mayewski, P.A., Sowers, T., Stuiver, M., Taylor, K.C., Clark, P.U., 1997. Holocene climatic instability: a prominent, widespread event 8200 yr ago. Geology 25 (6), 483e486. An, Z.S., Kukla, G.J., Porter, S.C., Xiao, J.L., 1991. Magnetic susceptibility evidence of monsoon variation on the Loess Plateau of central China during the last 130,000 years. Quat. Res. 36 (1), 29e36. An, Z.S., Porter, S.C., Kutzbach, J.E., Wu, X.H., Wang, S.M., Liu, X.D., Li, X.Q., Zhou, W.J., 2000. Asynchronous Holocene optimum of the East Asian monsoon. Quat. Sci. Rev. 19 (8), 743e762. Berger, A., Loutre, M.F., 1991. Insolation values for the climate of the last 10 million years. Quat. Sci. Rev. 10 (4), 297e317. Bigler, C., Larocque, I., Peglar, S.M., Birks, H.J.B., Hall, R.I., 2002. Quantitative multiproxy assessment of long-term patterns of Holocene environmental change from a small lake near Abisko, northern Sweden. Holocene 12 (4), 481e496. Brisset, E., Guiter, F., Miramont, C., Revel, M., Anthony, E.J., Delhon, C., Arnaud, F., Malet, E., Beaulieu, J.L.D., 2015. Lateglacial/Holocene environmental changes in the Mediterranean Alps inferred from lacustrine sediments. Quat. Sci. Rev. 110, 49e71. Chen, J.A., Wan, G.J., Wang, F.S., Zhang, D.D., Huang, R.G., Zhang, F., Schmidt, R., 2002. Environmental records of carbon in recent lake sediments. Sci. China Ser. D Earth Sci. 45 (10), 875e884. Chen, J.A., Wan, G.J., Zhang, D.D., Zhang, F., Huang, R.G., 2004. Environmental records of lacustrine sediments in different time scales: sediment grain size as an example. Sci. China Ser. D Earth Sci. 47 (10), 954e960. Chen, F.H., Xu, Q.H., Chen, J.H., Birks, H.J.B., Liu, J.B., Zhang, S.R., Jin, L.Y., An, C.B., Telford, R.J., Rao, Z.G., 2015a. East Asian summer monsoon precipitation variability since the last deglaciation. Sci. Rep. http://dx.doi.org/10.1038/srep11186. Chen, Y.Y., Lu, H.Y., Yi, S.W., Zhang, E.L., Xu, Z.W., Yu, K.F., Mason, J.A., 2015b. A preliminary quantitative reconstruction of precipitation in southern Mu Us sandy land at margin of Asian monsoon-dominated region during late Quaternary. J. Geogr. Sci. 25 (3), 301e310. Chu, K.C., 1973. A preliminary study on the climate fluctuations since the last 5,000 years in China. Sci. China ser A 2, 226e256. Cui, Z.J., Yang, J.Q., Zhang, W., Zhao, L., Xie, Y.Y., 2004. Discovery of a large area of ice-wedge networks in Ordos: implications for the southern boundary of permafrost in the north of China as well as for the environment in the latest 20 kaBP. Chin. Sci. Bull. 49 (11), 1177e1184. Ding, Z.L., Derbyshire, E., Yang, S.L., Sun, J.M., Liu, T.S., 2005. Stepwise expansion of desert environment across northern China in the past 3.5 Ma and implications for monsoon evolution. Earth Planet. Sci. Lett. 237 (1), 45e55.

9

Dong, G.R., Li, B.S., Gao, S.Y., 1983. The case study of the vicissitude of Mu Us sandy land since the late Pleistocene according to the Salawusu River strata. J. Desert Res. 3 (2), 9e14 (In Chinese with English abstract). Dong, G.R., Su, Z.Z., Jin, H.L., 1999. New views on age of the Salawusu Formation of late Pleistocene in northern China. Chin. Sci. Bull. 44 (7), 646e650. Du, S.H., Li, B.S., Chen, M.H., Zhang, D.D., Xiang, R., Niu, D.F., Wen, X.H., Ou, X.J., 2012. Kiloyear-scale climate events and evolution during the last interglacial, Mu Us desert, China. Quat. Int. 263, 63e70. Eris, K.K., 2013. Late Pleistocene-Holocene sedimentary records of climate and lakelevel changes in Lake Hazar, eastern Anatolia, Turkey. Quat. Int. 302, 123e134. Gao, S.Y., Dong, G.R., LI, B.S., Li, C.Z., 1985. Migration and accumulation of chemical elements in the Quaternary strata of the Salawusu River area in relation to climatic evolution. Geochimica 3, 269e276 (In Chinese with English abstract). Gasse, F., Van Campo, E., 1994. Abrupt post-glacial climate events in West Asia and North Africa monsoon domains. Earth Planet. Sci. Lett. 126 (4), 435e456. Harrison, S.P., Prentice, I.C., Guiot, J., 1993. Climatic controls on Holocene lake-level changes in Europe. Clim. Dyn. 4, 189e200. Hu, K., Mo, D.W., Wang, H., Zhang, Y.F., 2011. Environmental changes and human activities on both sides of sarah wusu River,in song(Xixia)and yuan dynasty. Acta Sci. Nat. Univ. Pekin. 47 (3), 466e474 (In Chinese with English abstract). Jia, T.F., 1996. Some problems about quaternary sediment and environment of sjara osso-gol river. J. Inn. Mong. Normal Univ. 4, 66e70 (In Chinese with English abstract). Jia, F.F., Lu, R.J., Gao, S.Y., Li, J.F., Liu, X.K., 2015. Holocene aeolian activities in the southeastern Mu Us desert, China. Aeolian Res. 19, 267e274. Leng, M.J., Marshall, J.D., 2004. Palaeoclimate interpretation of stable isotope data from lake sediment archives. Quat. Sci. Rev. 23 (7), 811e831. Li, X.Q., Zhou, W.J., An, Z.S., Dondson, J., 2003. The vegetation and monsoon variations at the desert-loess transition belt at Midiwan in northern China for the last 13 ka. Holocene 13 (5), 779e784. Li, B.S., Jin, H.L., Lv, H.Y., Zhu, Y.Z., Dong, G.R., Sun, D.H., Zhang, J.S., Gao, Q.Z., Yan, M.C., 1998. Processes of the deposition and vicissitude of Mu Us Desert, China since 150 ka B.P. Sci. China, Ser. D: Earth Sci. 41 (3), 248e254. Li, B.S., Zhang, D.D., Wen, X.H., Dong, Y.X., Zhu, Y.Z., Jin, H.L., 2005. A multi-cycle climatic fluctuation record of the last interglacial period: typical stratigraphic section in the Salawusu River Valley on the Ordos Plateau, China. Acta Geol. Sin. Ed. 79 (3), 398e404. Li, M.H., Kang, S.C., Zhu, L.P., You, Q.L., Zhang, Q.G., Wang, J.B., 2008. Mineralogy and geochemistry of the Holocene lacustrine sediments in Nam Co, Tibet. Quat. Int. 187, 105e116. Li, S.H., Sun, J.M., Li, B., 2012a. Holocene environmental changes in central Inner Mongolia revealed by luminescence dating of sediments from the Sala Us River valley. Holocene 22 (4), 397e404. Li, Y., Wang, N.A., Morrill, C., Anderson, D.M., Li, Z.L., Zhang, C.Q., Zhou, X.H., 2012b. Millennial-scale erosion rates in three inland drainage basins and their controlling factors since the Last Deglaciation, arid China. Palaeogeogr. Palaeoclimatol. Palaeoecol. 365, 263e275. Liu, K., Lai, Z.P., 2012. Chronology of Holocene sediments from the archaeological Salawusu site in the Mu Us desert in China and its palaeoenvironmental implications. J. Asian Earth Sci. 45, 247e255. Lu, H.Y., Miao, X.D., Zhou, Y.L., Mason, J.A., Swinehart, J., Zhang, J.F., Zhou, L.P., Yi, S.W., 2005. Late Quaternary aeolian activity in the Mu Us and Otindag dunefields (north China) and lagged response to insolation forcing. Geophys. Res. Lett. 32 (21), 365e370. Lu, H.Y., Mason, J.A., Stevens, T., Zhou, Y.L., Yi, S.W., Miao, X.D., 2011. Response of surface processes to climatic change in the dunefields and loess plateau of north China during the late quaternary. Earth Surf. Process. Landforms 3, 1590e1603. Lu, H.Y., Yi, S.W., Liu, Z.Y., Mason, J.A., Jiang, D.B., Cheng, J., Stevens, T., Xu, Z.W., Zhang, E.L., Jin, L.Y., Zhang, Z.H., Guo, Z.T., Wang, Y., Otto-Bliesner, B., 2013. Variation of East Asian monsoon precipitation during the past 21 k.y. and potential CO2 forcing. Geology 41, 1023e1026. Lu, R.J., Jia, F.F., Gao, S.Y., Shang, Y., Li, J.F., Zhao, C., 2015. Holocene aeolian activity and climatic change in qinghai lake basin, northeastern QinghaieTibetan Plateau. Palaeogeogr. Palaeoclimatol. Palaeoecol. 430, 1e10. Mason, J.A., Lu, H.Y., Zhou, Y.L., Miao, X.D., Swinehart, J.B., Liu, Z.Y., Goble, R.J., Yi, S.W., 2009. Dune mobility and aridity at the desert margin of northern China at a time of peak monsoon strength. Geology 37 (10), 947e950. Mishra, P.K., Anoop, A., Schettler, G., Prasad, S., Jehangir, A., Menzel, P., Naumann, R., Yousuf, A.R., Basavaiah, N., Deenadayalan, K., Wiesner, M.G., Gaye, B., 2015. Reconstructed late quaternary hydrological changes from lake tso moriri, NW Himalaya. Quat. Int. 371, 76e86. Passega, R., 1957. Texture as characteristic of clastic deposition. AAPG Bull. 41 (9), 1952e1984. Passega, R., 1964. Grain size representation by CM patterns as a geological tool. J. Sediment. Res. 34 (4), 830e847. Pei, W.C., Li, Y.H., 1964. Some tentative opinions on the problem of Sjara-Osso-Gol Series. Vertebr. Palasiat. 8 (2), 99e118 (In Chinese). Peng, Y.J., Xiao, J.L., Nakamura, T., Liu, B.L., Inouchi, Y., 2005. Holocene East Asian monsoonal precipitation pattern revealed by grain-size distribution of core sediments of Daihai Lake in Inner Mongolia of north-central China. Earth Planet. Sci. Lett. 233, 467e479. Perry, C.A., Hsu, K.J., 2000. Geophysical, archaeological, and historical evidence support a solar-output model for climate change. Proc. Natl. Acad. Sci. 97 (23), 12433e12438.

Please cite this article in press as: Liu, X., et al., Holocene water-level changes inferred from a section of fluvio-lacustrine sediments in the southeastern Mu Us Desert, China, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2016.12.032

10

X. Liu et al. / Quaternary International xxx (2017) 1e10

Peterse, F., Prins, M.A., Beets, C.J., Troelstra, S.R., Zheng, H.B., Gu, Z.Y., Schouten, S., , J.S.S., 2011. Decoupled warming and monsoon precipitation in East Asia Damste over the last deglaciation. Earth Planet. Sci. Lett. 301, 256e264. Rao, W.B., Chen, J., Tan, H.B., Weise, S.M., Wang, Y.X., 2013. NdeSr isotopic and REE geochemical compositions of Late Quaternary deposits in the deserteloess transition, north-central China: implications for their provenance and past wind systems. Quat. Int. 334e335, 197e212. Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C., Plicht, J., 2013. IntCal13 and Marine13 radiocarbon age calibration curves 0-50,000 years cal BP. Radiocarbon 55 (4), 1869e1887. lu, C., Fontugne, M., Bertaux, J., Woldring, H., Roberts, N., Reed, J., Leng, M., Kuzucuog lu, M., 2001. The tempo of Holocene Bottema, S., Black, S., Hunt, E., Karabiyikog climatic change in the eastern Mediterranean region: new high-resolution crater-lake sediment data from central Turkey. Holocene 11 (6), 721e736. Shen, J., Liu, X.Q., Wang, S.M., Matsumoto, R., 2005. Palaeoclimatic changes in the Qinghai Lake area during the last 18,000 years. Quat. Int. 136, 131e140. Solohub, J.T., Klovan, J.E., 1970. Evaluation of grain-size parameters in lacustrine environments. J. Sediment. Res. 40 (1), 81e101. Su, Z.Z., Dong, G.R., 1996. Recent progress in the study on Salawusu strata in north China. Sci. Geogr. Sin. 16 (3), 266e273 (In Chinese with English abstract). Sun, J.M., Ding, Z.L., Yuan, B.Y., Liu, T.S., 1996. Stratigraphic division of the Sala Us Formation and the inferred sedimentary environment. Mar. Geol. Quat. Geol. 16 (1), 31e47 (In Chinese with English abstract). Sun, J.M., Ding, Z.L., Liu, T.S., Rokosh, D., Rutter, N., 1999. 580,000-year environmental reconstruction from aeolian deposits at the Mu Us Desert margin, China. Quat. Sci. Rev. 18 (12), 1351e1364. Wagner, B., Melles, M., Hahne, J., Niessen, F., Hubberten, H.-W., 2000. Holocene climate history of geographical society Ø, east Greenland-evidence from lake sediments. Palaeogeogr. Palaeoclimatol. Palaeoecol. 160 (1), 45e68. Wang, Y.J., Cheng, H., Edwards, R.L., He, Y.Q., Kong, X.G., An, Z.S., Wu, J.Y., Kelly, M.J., Dykoski, C.A., Li, X.D., 2005. The Holocene Asian monsoon: links to solar changes and north Atlantic climate. Science 308 (5723), 854e857. Wang, Y.J., Cheng, H., Edwards, R.L., Kong, X.G., Shao, X.H., Chen, S.T., Wu, J.Y., Jiang, X.Y., Wang, X.F., An, Z.S., 2008. Millennial-and orbital-scale changes in the East Asian monsoon over the past 224,000 years. Nature 451 (7182), 1090e1093. Wolfe, A.P., Miller, G.H., Olsen, C.A., Forman, S.L., Doran, P.T., Holmgren, S.U., 2004.

Geochronology of High Latitude Lake Sediments. Long-term Environmental Change in Arctic and Antarctic Lakes. Springer, Netherlands, pp. 19e52. Wuennemann, B., Demske, D., Tarasov, P., Kotlia, B.S., Reinhardt, C., Bloemendal, J., Diekmann, B., Hartmann, K., Krois, J., Riedel, F., Arya, N., 2010. Hydrological evolution during the last 15 kyr in the Tso Kar lake basin (Ladakh, India), derived from geomorphological, sedimentological and palynological records. Quat. Sci. Rev. 29 (9), 1138e1155. Xiao, J.L., Nakamura, T., Lu, H.Y., Zhang, G.Y., 2002. Holocene climate changes over the desert/loess transition of north-central China. Earth Planet. Sci. Lett. 19 (1), 11e18. Xiao, J.L., Xu, Q.H., Nakamura, T., Yang, X.L., Liang, W.D., Inouchi, Y., 2004. Holocene vegetation variation in the Daihai Lake region of north-central China: a direct indication of the Asian monsoon climatic history. Quat. Sci. Rev. 23 (14), 1669e1679. Xu, Z.W., Lu, H.Y., Yi, S.W., Vandenberghe, J., Mason, J.A., Zhou, Y.L., Wang, X.Y., 2015. Climate-driven changes to dune activity during the Last Glacial Maximum and deglaciation in the Mu Us dune field, north-central China. Earth Planet. Sci. Lett. 427, 149e159. Yang, X.P., Rost, K.T., Lehmkuhl, F., Zhu, Z.D., Dodson, J., 2004. The evolution of dry lands in northern China and in the Republic of Mongolia since the Last Glacial Maximum. Quat. Int. 118e119, 69e85. Yuan, B.Y., 1978. Sedimentary environment and stratigraphical subdivision of sjara osso-gol formation. Sci. Geol. Sin. 3, 220e234 (In Chinese with English abstract). Zhang, Z.Q., Xing, W., Lv, X.G., Wang, G.P., 2014. The grain-size depositional process in wetlands of the Sanjiang Plain and its links with the East Asian monsoon variations during the Holocene. Quat. Int. 349, 245e251. Zheng, H., 1989. Late Pleistocene fluvo-lacustrine deposits and aeolian loess in North China. Geochimica 4, 343e351 (In Chinese with English abstract). Zhou, W.J., Donahue, D.J., Porter, S.C., Jull, T.A., Li, X.Q., Stuiver, M., An, Z.S., Matsumoto, E., Dong, G.R., 1996. Variability of monsoon climate in East Asia at the end of the last glaciation. Quat. Res. 46 (3), 219e229. Zhou, W.J., Dodson, J., Head, M.J., Li, B.S., Hou, Y.J., Lu, X.F., Donahue, D.J., Jull, A.J.T., 2002. Environmental variability within the Chinese desert-loess transition zone over the last 20000 years. Holocene 12 (1), 107e112. Zhou, Y.L., Lu, H.Y., Zhang, J.F., Mason, J.A., Zhou, L.P., 2009. Luminescence dating of sand-loess sequences and response of Mu Us and Otindag sand fields (North China) to climatic changes. J. Quat. Sci. 24 (4), 336e344.

Please cite this article in press as: Liu, X., et al., Holocene water-level changes inferred from a section of fluvio-lacustrine sediments in the southeastern Mu Us Desert, China, Quaternary International (2017), http://dx.doi.org/10.1016/j.quaint.2016.12.032