Mid-Late Quaternary loess-paleosol sequence in Lantian’s Yushan, China: An environmental magnetism approach and its paleoclimatic significance

Article Geology

September 2010 Vol.55 No.26: 2989–3000 doi: 10.1007/s11434-010-3212-6

SPECIAL TOPICS:

Mid-Late Quaternary loess-paleosol sequence in Lantian’s Yushan, China: An environmental magnetism approach and its paleoclimatic significance WU Yi1,4, ZHU ZhaoYu1*, RAO ZhiGuo2, QIU ShiFan1,4 & YANG Tian3 1

Key Laboratory of Marginal Sea Geology, Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, Guangzhou 510640, China; Key Laboratory of Western China’s Environment Systems (Ministry of Education), Lanzhou University, Lanzhou 730000, China; 3 School of Geography and Planning, Sun Yat-sen University, Guangzhou 510275, China; 4 Graduate University of Chinese Academy of Sciences, Beijing 100049, China 2

Received December 12, 2009; accepted March 18, 2010

The application of rock magnetism methods to investigating the variations of magnetic minerals in the sediments is an important approach to the reestablishment of paleoclimate evolution. Thus we performed fine magnetic measurements on the loess-paleosol sequence (from L15 upwards to S5, in which L is short for Loess, S is short for Paleosol, the same hereinafter) of Yushan stratigraphic section, which is on the southeastern margin of Chinese Loess Plateau, in Lantian County of China’s Shaanxi Province, and the thickness of which is ca. 40 m. Our study shows that the primary magnetic carriers of loess and paleosol in this section are magnetite, maghemite, hematite and goethite. Thermomagnetic analyses on the samples of representative horizons show that the higher pedogenesis degree of the sediments, the smaller variations of magnetization there will be before and after heating, probably related to the pedogenic alteration of loess sediments. Analyses of several magnetic parameters show a significant discrepancy between the paleoclimatic conditions recorded in the strata from the loess unit L15 upward to the paleosol unit S5 in the study area and those recorded in the relative strata of other sections on the Chinese Loess Plateau, and those recorded in marine sediments, indicating the great impact of regional geological background. Similarly, the rapid and intensive change recorded in the segment from L15 to S9-1, and the significant difference between the paleoclimate evolutions of the two periods before and after the change (from L15 to S9-1, and from L9 to the base of S6) indicate the strong alteration of magnetic carriers in the study area as a result of the alternations of summer and winter monsoons in East Asia. Loess Plateau, loess, paleosol, magnetic mineral, paleoclimate change Citation:

Wu Y, Zhu Z Y, Rao Z G, et al. Mid-Late Quaternary loess-paleosol sequence in Lantian’s Yushan, China: An environmental magnetism approach and its paleoclimatic significance. Chinese Sci Bull, 2010, 55: 2989−3000, doi: 10.1007/s11434-010-3212-6

Climate change has been an important issue with global concern in recent years. The great need of understanding and predicting future climate changes urges scientists to pay a close attention to the climate changes since 65 Ma ago, because the climate evolution of this geological period is closely related to future climate change. The Cenozoic global climate has experienced a long-term coldness and aridity after EECO (the early Eocene Climatic Optimum), although with occasional warm events after EECO [1]. *Corresponding author (email: [email protected]) © Science China Press and Springer-Verlag Berlin Heidelberg 2010

More frequent fluctuations occurred in the global climate of Late Cenozoic, with long-term and multi-cycled alternations of glacial and interglacial periods being recorded in marine sediments [2], polar ice cores [3] and Chinese loesspaleosol sediments [4,5]. The accumulation area of loess in East Asia has been a focus of paleoclimate research for Quaternary scientists as one of the most important sites for a long time. Heller and Liu [6] suggested that the Quaternary loess in Luochuan began to accumulate at least in the age of ca. 2.4 Ma, according to their magnetostratigraphical analyses. Ever since their pioneering work, various researches csb.scichina.com

www.springerlink.com

2990

WU Yi, et al.

Chinese Sci Bull

have been done on the Chinese Loess Plateau, and the age of the aeolian sediments has been put forward again and again. Guo et al. [7] showed that the Tertiary aeolian sediments in Qin’an, Gansu, on the west Loess Plateau began to accumulate at least in the age of ca. 22 Ma, much earlier than that of the previously oldest red clay in Jiaxian, on the north Chinese Loess Plateau, the age of which is 8.35 Ma [8]. All these studies are of great significance on the researches of the Cenozoic arid and monsoonal environments and the tectonic-climate in Asia. Relatively fine investigation at local area is very necessary under the background of long-term climate research on the Loess Plateau, since it can improve our understanding the response of regional climate records to global changes. Meanwhile, it is a good try as well as a good approach to the reestablishment of paleoclimate change with various techniques. The past 20 more years has seen a lot of work done on the reestablishment of paleoclimate records based on the Quaternary loess-paleosol sequence, a very important part of which includes the initial stratigraphical identifications and geochronological studies on the loess-paleosol sequence. Many parallel comparisons of stratigraphic sections and their global comparisons with marine sediments are based on magnetic susceptibility [4,9–12]. The overall cold-arid trend of Quaternery climate is definitely indicated by several magnetic curves in recent excellent reports by Deng et al. [13,14] on the enviromagnetic records of the Jiaodao section in the middle Loess Plateau and the Jingbian section in the north Loess Plateau, adjacent to the Mu Us Desert. Fine investigations on the magnetoclimatological records of the aeolian sediments in the southeast Loess Plateau, which are very sensitive to paleomonsoon climate changes, are very necessary on the complicated background of East Asian monsoon system. The study area is adjacent to the north piedmont of Qinling Mts., thus the underlying surface of the aeolian sediments are subject to the direct or indirect impact of frequent and intensive tectonic movements [15–17], and the response of magnetic record is more complicated. Situated in the vicinity of Yushan town, Lantian county of China’s Xi’an, the recently found Yushan stratigraphic section is selected as the research object in our study (Figure 1(a)). We reestablish the mid-late Quaternary climate evolution recorded in the loess-paleosol strata on the basis of various rock magnetic measurements, and try to understand the magnetoclimatic record model of the aeolian sediments in good hydrothermal conditions.

1 Geological setting and sampling The Lantian area is dominated by a semi-humid continental monsoon climate of warm temperate zone, with a mean annual temperature of ca. 13.1°C and a mean annual precipitation of ca. 620 mm concentrating from July to September. The Yushan stratigraphic section (39.2°N, 109.5°E) is lo-

September (2010) Vol.55 No.26

cated along the north Bahe River (an important branch of Weihe River that originates from the north piedmont of the Qinling Mts.) bank, with an outcrop of nearly horizontal strata about 40 m thick. Three red beds of paleosols on top of the section are clearly visible during field investigations [19]. They are attributed to the strongly developed paleosol unit S5, the horizon overlying the paleosol unit S6 on which the fossil mandible of Lantian Homo erectus was found in the Chenjiawo section [20], which is transversely comparable in many places. There is a yellow-brown silt unit nearly 6 m thick in the mid-upper Yushan section, and a similar yellow-brown silt unit nearly 3 m thick at the bottom of the Yushan section, determined by field investigation and subsequent measurements as the upper silty loess unit L9, and the lower silty loess unit L15, respectively (Figure 1(b)). Fifteen paleosol units are recognized in the strata above the L15 horizon, in light of the color differences and the soil grain features of the outcrop of the section. A stratigraphic segment nearly 0.5 m thick from the base of the L15 unit to the bottom of the trench is beyond our discussion because of the potential effect of fluviation on the magnetic properties of this segment. We collected oriented samples for paleomagnetic measurements in the total section, except the nearly 0.6-m-thick layer of the cultivated soil and weathering earth on top of the section. Reliable geochronological sequence is established on the basis of fine paleomagnetic measurements. Paleomagneitic data are not involved in the present paper in detail, only the Brunhes/Matuyama (B/M) boundary, and the lower and upper boundaries of the Jaramillo (J) subchron are illustrated (Figure 1(b)). We performed fine rock magnetism measurements on a total of 388 powder samples (numbered from YS-1 to YS-388) collected at an average interval of 10 cm in a thickness of 38.8 m.

2 Experimental methods Magnetic susceptibility values, χlf and χhf, were measured on air-dried and weighed samples using a Bartington MS2 meter at a low frequency of 470 Hz and a high frequency of 4700 Hz, respectively. Two measures of frequencydependent magnetic susceptibility (χfd, defined as χlf–χhf, and χfd%, defined as (χlf–χhf)/χlf×100%) were calculated from these measurements. Anhysteretic remanent magnetization (ARM) was imparted in a 100 mT peak alternating field (AF) and a 0.05 mT constant biasing field with an AGICO AF demagnetizer, yielding the χARM values. Isothermal remanent magnetization (IRM) and saturation isothermal remanent magnetization (SIRM) measurements were performed using a MMPM10 pulse magnetizer. All remanences were measured on the samples using a Molspin spinner magnetometer, after the samples were exposed to peak alternating fields of 20, 50, 100, 300, and 1000 mT, respectively, then demagnetized at reversed fields of 20 and 300 mT, respectively. Moreover, SIRM measurements were

WU Yi, et al.

Chinese Sci Bull

2991

September (2010) Vol.55 No.26

Figure 1 (a) A sketch map of Chinese Loess Plateau and several important aeolian sediment sites (adapted from [18]). (b) Stratigraphic classification of the middle-Late Quaternary loess-paleosol sequence in the Lantian section with an χlf curve of powder samples. In (b), major loess and paleosol horizons are marked with Li and Si, respectively. The weak soil in the lower S6 is marked with S6-1 and the weak soil in the middle-lower L9 is marked with L9-S. The two sub-soils within S9 are marked with S9-1 and S9-2, respectively.

performed at 2.5 T on 6 representative samples selected on the basis of field stratigraphic investigations and magnetic susceptibility measurements. After that, the 6 samples were demagnetized at a range of reversed fields. Subsequently, temperature-dependent magnetization and hysteresis parameters (including saturation magnetization Ms, saturation remanence Mrs, coercivity Bc, and coercivity of remanence Bcr) were measured in air on the 6 weighed samples, using a VFTB Curie balance. All the above-mentioned magnetic parameters have been normalized by the sample masses. Other magnetic parameters employed in our study are defined as follows [21–23]: S–0.3T = 0.5×([–IRM–0.3T/SIRM]+1) ×100% ,

Table 1

Magnetic data of the samples from representative horizons

Samples from χlf (10–7 χfd (10–7 χfd% representative horizons m3 kg–1) m3 kg–1)

ΔM%a)

IRM0.3T /IRM2.5T%

S5, YS-27

26.85

3.63

13.51

3.65

95.99

L9, YS-158

2.81

0.16

5.62

29.08

87.01

L9-S, YS-184

4.10

0.26

6.29

25.79

87.44

S12, YS-261

20.07

2.29

11.39

13.30

97.62

L15, YS-367

2.53

0.08

3.03

28.92

86.09

L15, YS-374

2.83

0.15

5.44

26.01

85.14

a) ΔM% represents the magnetization difference before and after heating.

HIRM = 0.5×(SIRM+IRM–0.3T), in which IRM–0.3T represents the IRM value measured at two peak alternating fields, first at a normal peak alternating field of 1 T, yielding the SIRM, then at a reversed peak alternating field of 0.3 T. The magnetic data of the 6 selected representative samples are listed in Table 1. All the SIRM values of samples were obtained at a normal peak alternating field of 1 T, except those of the 6 representative samples, which were obtained at a normal peak alternating field of 2.5 T, yielding the SIRM value defined as IRM2.5T in Figure 2 and Table 1.

3 Results 3.1

Temperature-dependent magnetization

An obvious decrease of magnetization on the 6 representative samples occurs around 585°C, indicating that the primary magnetic minerals in the loess and paleosol samples are magnetite (Figure 3). The magnetizations of the representative loess samples YS-158, YS-367 and YS-374, and the weak soil sample YS-184 decrease slightly in the heating period from 585 to 700°C, indicating the presence of

2992

WU Yi, et al.

Chinese Sci Bull

September (2010) Vol.55 No.26

Figure 2 SIRM acquisition (a) and demagnetization (b) curves of typical samples, in which loess samples YS-158, YS-184, YS-367 and YS-374 are shown in solid lines, paleosol samples YS-27 and YS-261 are shown in dashed lines. Notably, the curve crossing the horizontal axis between –50 and –60 is attributed to the weak weathered loess sample YS-184.

hematite (Figure 3(b),(c),(e),(f)). But the presence of hematite, either from a detrital origin or generated by heating [23], is unidentified by a nearly horizontal curve of magnetizations for the representative paleosol samples, YS-27 and YS-261 above 600°C (Figure 3(a),(d)). All the heating curves of the 6 samples show a dent around 100°C, indicating the presence of goethite [24]. The heating curves of the loess samples YS-158, YS-367 and YS-374, and the weak weathered loess sample YS-184 show a bump at 250°C, and a dent at 350°C, as is illustrated in Figure 3. The dent is due to the transformation of unstable ferrimagnetic maghemite into weak magnetic hematite by heating, indicating the presence of maghemite [23]. 3.2 SIRM and reversed demagnetization of representative samples The SIRM curves of the 6 samples are largely consistent, with their IRM values at 300 mT accounting for 85.14% to 97.62% of their SIRM values, indicating that most IRM of the samples is carried by low-coercivity magnetic minerals, whereas only a small amount of remanence is obtained in the strong field above 300 mT, indicating the presence of high-coercivity magnetic minerals in these samples (Figure 2(a), Table 1). Analyses of the heating curves in Figure 3 show that the magnetism carriers of typical paleosol samples are dominated by magnetite and maghemite, but the presence of hematite is insignificant, and that the magnetism carriers of loess samples are dominated by ferrimagnetic magnetite and maghemite, and antiferromagnetic hematite and goethite. The remanence coercivities of both the paleosol samples YS-27 and YS-261 are ca. 25 mT, and ca. 55 mT for that of the weak soil sample YS-184, whereas the remanence coercivities of all the 3 typcial loess samples are over 60 mT (Figure 2(b)). The remanence coercivity of the weak soil

sample YS-184 is between those of the loess samples and those of the paleosol samples, indicating that its highcoercivity magnetic mineral content is between those of the paleosol and those of the loess samples. The variations of high or low-coercivity magnetic mineral contents indicated by remanence coercivity values from paleosol to weak soil, and to loess are probably due to the significant alteration of secondary pedogenesis on the composition of magnetic minerals. 3.3

Hysteresis loops

Figure 4 illustrates the hysteresis loops of 2 representative samples. Obviously, the hysteresis loops of the paleosol sample YS-27 is narrower than that of the loess sample YS-367, probably due to the higher content of lowcoercivity magnetic mineral in the paleosol than that in the loess [18,25]. The hysteresis loop of the loess sample YS-367 is still open at 1 T, equally indicating the much higher content of high-coercivity magnetic mineral in the loess sample YS-367 than that in the paleosol sample, such as YS-27 [21,26]. The Day diagrams of Mrs/Ms and Bcr/Bc for the 6 samples illustrate that the mean magnetic mineral grain size is attributed to pseudo single domain (PSD) (Figure 5). 3.4 Depth-dependent magnetic parameter features and correlation analyses on some of the parameters The χlf, χfd, χARM and SIRM curves illustrated in Figure 6 are roughly consistent and well comparable, with valleys corresponding to the loess deposition in glacial periods, and peaks corresponding to the paleosol development during interglacial periods, displaying the cyclicities of magnetic mineral content and grain size. The χARM/SIRM and χARM/ χlf ratios are usually introduced as the parameters reflecting

WU Yi, et al.

Chinese Sci Bull

September (2010) Vol.55 No.26

2993

Figure 3 Normalized temperature-dependent magnetization curves for the samples of some typical horizons. Arrows indicate the heating and cooling processes. These curves were obtained with VFTB in air using a magnetic field of 100 mT.

Figure 4

Hysteresis loops of representative samples from the paleosol S5 (a) and the lower silty loess L15 (b).

the variation of magnetic mineral grain size. Previous studies have shown that the grain size distributions of magnetic minerals both in paleosol and loess are similar, and they are independent of the pedogenesis degree [29–31], which means that the χARM/SIRM and χARM/χlf ratios illustrated in Figure 6 largely indicate the variations in the content of fine magnetic minerals. The variations of both ratios are roughly consistent with those of χlf, χfd, χARM and SIRM in the Yu-

shan section. The alternations of loess deposition during glacial periods and paleosol development during interglacial periods are still definitely indicated by the χARM/SIRM and χARM/χlf ratios, but there is a marked difference between the peak/valley values of the χARM/SIRM and χARM/χlf curves, and those of the 4 magnetic parameters for the whole section, particularly in the period when the paleosol unit S5 developed, the χARM/χlf ratio is unusually lower than that of

2994

WU Yi, et al.

Chinese Sci Bull

Figure 5 Hysteresis ratios plotted on a Day diagram [27,28] for representative samples. SD, Single domain; MD, multidomain.

the less-developed paleosol unit below the S5 horizon. Furthermore, a significant transition is recorded by the magnetic parameters of all the curves in Figure 6 except for a slight difference of HIRM in the middle of the segment from the lower silt loess unit L15 to the paleosol unit S6, approximating 20 m in depth, between S9-1 and L9, and all the curves generally display low values in the segment from L9 to S6. S–0.3T serves as a good indicator for the relative contents of ferrimagnetic and antiferromagnetic minerals, with its value ranging from 0 (indicating pure hematite/goethite) to 100% (indicating pure magnetite/maghemite) [22]. The S–0.3T value of the Yushan section ranges from 57.13% to 95.41%, and averages 82.05%, indicating that the primary magnetic mineral of the section is ferrimagnetic (mainly magnetite). Very significantly, all the S–0.3T values of the paleosol unit S5 generally are over 90%, and the peak S–0.3T value of the paleosol unit S13 equally reaches 92.00% (Figure 6). HIRM indicates the absolute content of highcoercivity magnetic minerals [21]. Comprehensive analyses of HIRM and S–0.3T can reveal the antiferromagnetic mineral contents more clearly. The HIRM curve shows that the content of high-coercivity magnetic minerals in the paleosol is generally higher than that in the loess, but the relative content of high-coercivity magnetic minerals recorded by S–0.3T is higher than that in the paleosol. Generally, χlf indicates the total content of ferromagnetic minerals and χfd is introduced to indicate the presence of SP grains generated in the pedogenesis. The major carriers of ARM are SD and PSD magnetic grains [32–34]. SIRM differs from χlf in that SIRM is not subject to the influence of paramagnetic and diamagnetic minerals, it indicates the contribution of ferrimagnetic minerals and partial antiferromagnetic minerals, both with a grain size over the SP/SD boundary (ca. 20 μm for magnetite grains, for example) [21]. The linear correlation coefficients of χfd, χARM and

September (2010) Vol.55 No.26

SIRM with χlf for all the 388 samples are 0.9799, 0.9465, and 0.8992, respectively (Figure 7). The high correlation between χfd and χlf indicates the marked contribution of SP magnetic mineral grains (with a grain size ranging from ca. 20 to 25 nm) to the magnetic susceptibility. The χfd% roughly rises with the increase of χlf within a certain range, but without a linear correlation between them. χfd% is sensitive to the magnetic grains with a grain size around the SP/SD boundary [11]. Figure 7(b) shows that the correlation between χfd% and χlf will be changed if χfd% is above 10%, because χfd% will gradually saturate under such condition [12]. The highest χlf value of the paleosol unit S5 is 28.11×10–7 m3 kg–1 in the Yushan section, and the χfd% curve nearly reaches a platform averaging ca. 12% at this horizon. χfd% has a wide range of 0.18% to 14.88% in the total Yushan section (Figures 6 and 7), but many studies indicate that χfd% generally is not above 13% [13,31,35,36]. χfd% is the second highest in the segment from S12 to S9-1, ranging from 7.69 to 11.83% and averaging 9.77%. But χfd% is relatively low in the segment from L9 to the base of S6, with a wide range of 0.10 to 11.07% and an average of 5.16%. Comprehensive analyses of χfd%, χARM and χlf for the loess-paleosol sequence in the Yushan section indicate that the primary contributors of magnetic susceptibility and remanence are SP, SD and small PSD ferrimagnetic grains [11–13,26,37].

4 Discussion 4.1

Thermomagnetic behaviors of maghemite

The magnetization variations (ΔM%) of the 6 representative samples before and after heating are listed in Table 1. As is shown in the table, pedogenesis degree is almost negatively correlated with ΔM%. Usually, new magnetic minerals will be produced by the thermal treatments of loess and paleosol samples in the laboratory, either in air or in argon [18,23,26]. As is mentioned above, the heating curves in Figure 3 indicate the transformation of maghemite into hematite. Due to the irreversibility of this transformation, the magnetization of samples will decrease significantly after cooling [23]. The transformation of maghemite is insignificant on the heating curves of typical paleosol samples YS-27 and YS-261 (Figure 3 and Table 1), which is similar to the paleosol samples from the Sanmenxia area on the southeast Loess Plateau [18]. Moreover, the cooling curves of normalized magnetization-temperature that we obtained did not occur above the heating curves [38]. The magnetization variations before and after heating indicated by the thermal demagnetization curves of loess and paleosol samples from the Luochuan section were attributed by Liu et al. [39] to the aeolian origin of maghemite, whereas several magnetic analyses by Guo et al. [40] on the loess and paleosol samples of 3 sections from the north to the south of

WU Yi, et al.

Chinese Sci Bull

September (2010) Vol.55 No.26

2995

Figure 6 Variations of the magnetic parameters for the loess-paleosol sequence in the Yushan section by comparison with MIS curve. Grey bars indicate the paleosol horizons. Representative horizons are compared with the MIS curve “LR04” [2] (warm marine oxygen isotope stages are odd numbered) and the stacked Chinese loess grain size curves “Chiloparts” [5].

2996

Figure 7

WU Yi, et al.

Chinese Sci Bull

September (2010) Vol.55 No.26

The relationship between χfd versus χlf (a), χfd% versus χlf (b), χARM versus χlf (c), and SIRM versus χlf (d).

the Loess Plateau suggested that the relationship between the relative content of maghemite and the secondary pedogenesis degree might be subject to the critical conditions of the regional climate. Recent rock magnetism studies on loess and paleosol tend to believe that the occurrence of maghemite in sediments is largely due to secondary pedogenesis [12,29,41,42], and the maghemite content related to secondary pedogenesis degree was even recommended as an indicator of climate change [43]. In light of such recognition, the magnetization-temperature curves of paleosol samples should display more obvious valley shapes than those of loess samples if no disturbance results from the thermal transformations of other matters. Actually, the thermomagnetic features of loess and paleosol obtained in our studies seem to be contrary to that, however. Similar thermomagnetic behaviors in the loess of the Sanmenxia area were attributed by Wang et al. [18] to the more common transformations of magnetite/ maghemite into hematite. Then how to interpret the difference of thermomagnetic behaviors between paleosol and loess? Carbon isotope analyses on the Duanjiapo section of the Lantian area indicated that the abundance of C4 plant biomass was higher in paleosol than in loess, and spatially

decreased from the southeast to the northwest of the Loess Plateau, obviously characterized by the temperature and precipitation of the paleoclimate under the control of summer monsoon [9]. Land cover difference between loess and paleosol in geological history leads to the difference of organic matter content in them, and glacial and interglacial alternations would lead to the alteration of silicates as well as the compositional and structural alterations of clay minerals within the loess and paleosol [19]. Significantly, systematic thermomagnetic analyses by Deng et al. [26] on the Jiaodao loess section suggested that the ferrous source in paleosol that can be changed into new magnetite by heating in experiments may be scarce, but no evaluation on the scarcity of such ferrous matters is available at present. We believe that the difference of environments under which loess deposited and paleosol developed may have contributed to the difference of ferrimagnetic minerals generated by heating, with more newly-generated ferrimagnetic minerals in paleosol than in loess, and that the newly-generated ferrimagnetic minerals may have compensated the magnetization decline of maghemite as a result of thermal transformation, thus no magnetite transformation is visible on the heating curves of paleosol. Consequently, the higher

WU Yi, et al.

Chinese Sci Bull

the pedogenesis degree is, the smaller the ΔM% will be, and the magnetization variations of the paleosol samples from the S5 unit with the highest pedogenesis degree approximates zero. 4.2 Long-term magnetoclimate changes of the Yushan section and its characteristic periods A simple comparison between the sedimentary rates of loess for some sections in the Loess Plateau is shown in Table 2. Being situated near the Mu Us Desert, the Jingbian section has a sedimentary rate much higher than those of other sections. The sedimentary rates of the Jiaodao and Luochuan loess sections in the middle Loess Plateau are similar, and little difference occurs between those of the Weinan section in the southeast Loess Plateau and the Caocun section in the Sanmenxia area. Situated on the 5th terrace of Weihe River, the Yanyu section has a relatively low sedimentary rate, probably due to the distance to the source area and the slope runoff. Besides the influence of these two factors, the obviously low sedimentary rate of the Yushan section may be closely related to the adjacence of the Weihe River valley to the north piedmont of the Qinling Mts. Striking from west to east, the Qinling Mts. do not only block the northward warm-wet summer monsoon, but also have a strong influence on the regional circulation of winter monsoon. Differences of source distance, migration capability and the underlying surface of sediments lead to the difference between the sedimentary rate of the loess in the study area and those of other areas on the Loess Plateau, thus equally resulting in the big difference of sedimentary rates between the Weinan, Duanjiapo and Yushan sections in the Weihe River valley. Analyses on the Quaternary magnetic records in the loess-paleosol sequence of the Jiaodao section after isolating the detrital signals from the secondary pedogenesis signals revealed that magnetic grains gradually coarsened and winter monsoons step-intensified in Quaternary [13]. Subsequently, analyses by Deng et al. [14] on the hematite of the loess-paleosol sequence in the Jingbian section adjacent to the Mu Us Desert showed the general trend of coldness and aridity in Quaternary climate. But no overall and consistent trend is indicated by the content and grain size of magnetic minerals recorded by the magnetic parameters of the Yushan section in our study. The total rock grain size of the aeolian sediments of the Jingbian loess section shows that multi-phased southward invasions of the Mu Us Desert Table 2

2997

September (2010) Vol.55 No.26

occurred in 2.6, 1.2, 0.7 and 0.2 Ma since 3.5 Ma, indicating the step-weakened summer monsoon in East Asia [44]. In addition, the climate transitions of 1.2 and 0.7 Ma recorded in this section correspond to the loess units L14 and L7 in the sedimentary sequence. A rapid transition is definitely indicated by all the magnetic parameters during the transition from S9-1 to L9, except the smooth transition of HIRM, as is illustrated in Figure 6. The generally small variation of HIRM may just have demonstrated the stability of the absolute content of the coarse-grained antiferromagnetic minerals mainly of a detrital origin. By comparison, the significant transition has not been detected in other contemporaneous parallel records, such as the related magnetic records of the Jingbian [14], Jiaodao [13], Luochuan [45] and Sanmenxia [47] loess sections, and even the Weinan loess section [46]. Only the contemporaneous magnetic record of the Duanjiapo loess section [45] is comparable with that of the Yushan section. Nor has it been detected in the contemporaneous Marine Isotope Stages (MIS) (Figure 6), indicating that such a rapid transition recorded in the Yushan section is under the control of regional factors. There is a significant difference of magnetic records from L15 to S9-1 in the Yushan section, but the ferrimagnetic mineral content indicated by S–0.3T shows an increasing trend in this segment, it follows that the general hydrothermal conditions were favorable for pedogenesis, thus indicating an intensified summer monsoon in East Asia, which is closely related to the climate of this area during this period. Such a situation does not seem to be consistent with the paleomonsoon evolution reestablished by the magnetic records of the Jingbian and Jiaodao sections, nor is it consistent with the global cooling trend indicated by marine oxygen isotope ratios (Figure 6). Bloemendal and Liu [45] did a lot of magnetic and geochemical investigations on the Duanjiapo section in the Lantian area and the representative Luochuan section, using the difference values of several parameters of the two sections to reestablish the differences of weathering and pedogenesis between the two sections, and two intensified summer monsoon events were indicated by their analyses, one was after 1.2 Ma (corresponding to L14), the other was after 0.6 Ma (corresponding to the upper L6). Therefore, the climatic transition recorded in the Yushan section is not simply a regional climate change signal. Actually, the climate transition records of the Yushan and Duanjiapo sections and even the Luochuan section [48] are comparable. An et al. [10] showed that the variability of

Loess sedimentary rates of some sections on the Loess Plateau a)

Section

Jingbian [14, 44]

Jiaodao [13]

Luochuan [15, 45]

Weinan [46]

Duanjiapo [45]

Yushan (This study)

Yanyu [16]

Caocun [47]

Selected range

L15–S5

L15–S5

L15–S5

L13–S5

L15–S5

L15–S5

L15–S5

L13–S5

Sedimentary 13.52 6.76 6.38 6.77 5.99 4.72 5.87 6.33 rate (cm ka–1) a) The chronological data of each section used in this table is cited from “Chiloparts” [5]. There is a slight difference of selected range between the Weinan and Caocun sections.

2998

WU Yi, et al.

Chinese Sci Bull

East Asian summer monsoon increased with its strength being decreased, whereas East Asian winter monsoon progressively intensified 2.6 Ma ago. The differences of climatic trend and transitions recorded in the Lantian loess and Jingbian loess may indicate the complexity of the East Asian monsoon climate evolution on the Loess Plateau. Situated on the southeast margin of the Loess Plateau, the Yushan section is influenced by the summer monsoon earlier and more intensively than other dust accumulation areas on the Loess Plateau. Estimations of mean annual temperature and mean annual precipitation on the Xifeng, Changwu and Weinan loess-paleosol sections indicate that the loess units L9 and L15 deposited in a semi-desert environment, and the Loess Plateau was under the control of an intensified winter monsoon. Although the warm-wet summer monsoon could reach the southern margin of the Loess Plateau, but seldom moved northward or northwestward, or throughout the whole Loess Plateau [49]. Carbon isotope analyses on the Duanjiapo section indicate that C4 plants widely spread from 1.3 to 0.9 Ma, indicating an intensified summer monsoon [9]. Our previous study showed that the loess unit L15 in the study area probably was not altered by pedogenesis [50]. But after the loess unit L15 deposited, the southeast margin of the Loess Plateau, where the Yushan section is located, was subject to a warm-wet summer monsoon, although the Loess Plateau was dominated by an intensive winter monsoon. Under such a circumstance, progressive changes of magnetic mineral content and grain size occurred as a result of the intensified secondary pedogenesis. From the L9 unit upward to the base of the S6 unit in the Yushan section, the magnetic parameters of this segment are totally different from those of the segment below the L9 unit. Compared with those of the underlying segment, the magnetic parameters of this segment generally decreased with low values, and the ferrimagnetic mineral content indicated by S–0.3T decreases significantly with a relatively weakened pedogenesis recorded in this segment. All these conditions indicate that the aeolian sediments have experienced a relatively stable environment of coldness and aridity during this period. Significant differences still remain between loess and paleosol in the χARM/SIRM and χARM/χlf ratios, the parameters of magnetic mineral grain size, but generally their background values are relatively low. Similarly and comparatively, such stable characteristics have not been revealed in other contemporaneous parallel records, such as the related magnetic records of the Jingbian [14], Jiaodao [13], Luochuan [45], Sanmenxia [47], and Duanjiapo loess sections [45]. The χlf values of the Weinan loess section [46] may equally indicate that the contemporaneous paleosol units did not experience a strong pedogenesis. The oxygen isotope values increase during the contemporaneous period from MIS25 to MIS17 (Figure 6), indicating a cold and arid climate trend. The paleosol unit S6 developed at the age of 0.7 Ma, which was previously envisaged as an

September (2010) Vol.55 No.26

important period when the Qinhai-Tibet Plateau, the Loess Plateau and the Qinling Mts. were stage-uplifted by the neotectonic movement [17]. The Qinghai-Tibet Plateau reached an average altitude of over 3000 m after the Kunlun-Yellow River tectonic movement from 1.1 to 0.6 Ma, thus largely rising into the cryosphere (as demonstrated in [51]). The extensive landform alteration and land-cover change during this period inevitably have a great impact on the general situation of the East Asian climate, most probably resulting in an intensified East Asian monsoon system. From the L9 unit to the base of the S6 unit, little difference of magnetic mineral types is revealed by S–0.3T, thus indicating the climatic stability of dust source area under the control of a strong winter monsoon. By comparison, the small variations in the parameters of magnetic mineral grain-size, which are controlled by the alternations of glacial and interglacial periods, may indicate the alteration of summer monsoon on the dust sediments to varying degrees. But such an alteration was not very intensive even in an interglacial climate, probably due to a weakened summer monsoon. Being adjacent to the north piedmont of the Qinling Mts., the underlying surface of the dust sediments in the study area is subject to frequent alterations caused by the neotectonic movement, thus raising the uncertainties in the application of sediments to reestablish the climate evolution in this area. The paleosol units S12, S13 and S14 correspond to the marine isotope stages MIS33, MIS35 and MIS37, respectively. The peak values of oxygen isotope in these three marine isotope stages decrease successively, but consistent changes did not occur in the magnetic parameters of the three paleosol units. The L15, L9 and S5 units usually occur at the bottom of the aeolian sediments on the river terrace in the Fen Wei Graben, on the edge of the Qinling orogen, where the Yushan section is located. The ages of these important loess-paleosol horizons, approximating 1.2, 0.9 and 0.65 Ma, are regarded as the important periods of the accelerated northward movement of India Plate in Quaternary [16]. The paleosol unit S5 is the most strongly developed soil in the loess-paleosol sequence throughout the Loess Plateau, corresponding to a climate optimum [19,49]. The reticulated red clay in South China’s red bed has been compared with the paleosol unit S5 on the Loess Plateau as a contemporaneous horizon [52]. The paleosol unit S5 developed in a period of climate conditions better than any other periods during which other earlier Quaternary paleosol units developed. The climate of the south Loess Plateau during this period was probably similar to a subtropical climate [49,53]. As is mentioned above, all the magnetic parameter curves of the paleosol unit S5 in the Yushan section approximate a platform, indicating a strong pedogenesis in the loess of the study area during this period, which is inconsistent with other sections mentioned above. The magnetic parameter curves of the paleosol unit S5 in these sections generally display a narrow peak. Compared with

WU Yi, et al.

Chinese Sci Bull

other magnetic parameters, the peak value of the χARM/χlf ratio for S5 is not significant, and even lower than those of the paleosol units S6, S13 and S14, probably due to the abundant SP magnetic grains generated in the pedogenesis [47]. As is indicated by the magnetic records, the climate of the study area probably became warm and wet since S6 developed, after experiencing a long-term cold-arid climate during the interval of L9 to S6, and a rapid climate change may occur during the interval of L6 to S5.

5

Conclusions

Our analyses of fine rock magnetism measurements on the loess-paleosol sequence from L15 to S5 for the Yushan section in Shaanxi’s Lantian area have come to the following conclusions. Firstly, the primary magnetic carriers of the sediments in the Yushan section are ferrimagnetic magnetite and maghemite, and antiferromagnetic hematite and goethite. The SP, SD and small PSD grains of ferrimagnetic magnetite and maghemite are the major contributors of the increased χlf, χARM and SIRM. Secondly, the magnetization variations before and after heating indicated by the thermomagnetic curves are probably due to composition differences, including the differences of organic matter content, and the difference in the contents and structures of silicates and clay minerals, as a result of the pedogenesis difference between loess and paleosol. Finally, several magnetic parameter records indicate that the regional tectonic movement during the period from L15 to S5 has a significant impact on the deposition and secondary alteration of the aeolian sediments in the Yushan section, with a rapid and extensive transition in the interval of S9-1 to L9 as the most important consequence. As is indicated by the magnetic records of L15 to S9-1, the study area was dominated by a progressively intensified summer monsoon during this period, which is closely related to the situation of the Yushan section on the southeast margin of the Loess Plateau, and being in the frontier of summer monsoon, as well. A weakened summer monsoon and progressively intensified winter monsoon are indicated by the magnetic records of the segment from L9 to the base of S6. All the magnetic parameters were measured in the Key Laboratory of Western China’s Environmental Systems, Lanzhou University, except the χlf and χhf measurements. Thanks are given to Prof. Liu Xiuming for his kind help in our magnetic measurements. We thank Dr. Han Jiangwei for his hard work in the field sampling. We are also grateful to the two reviewers for their constructive advice which improves our paper substantially. This work was supported by the National Basic Research Program of China (2004CB720200 and 2010CB833405), the Knowledge Innovation Program of Chinese Academy of Sciences (KZCX2-SW-133) and the National Natural Science Foundation of China (40872111). This is contribution No. IS-1175 from GIGCAS.

September (2010) Vol.55 No.26

1 2

3 4 5

6 7 8 9

10 11 12 13

14

15 16

17 18

19 20 21 22

23

24

2999

Zachos J, Pagani M, Sloan L, et al. Trends, rhythms, and aberrations in global climate 65 Ma to present. Science, 2001, 292: 686–693 Lisiecki L, Raymo M E. A Pliocene-Pleistocene stack of 57 globally distributed benthic δ 18O records. Paleocenography, 2005, 20, PA1003, doi:10.1029/2004PA001071 Petit J R, Jouzel J, Raynaud D, et al. Climate and atmospheric history of the past 420000 years from the Vostok ice core, Antarctica. Nature, 1999, 399: 429–436 Liu T S, Ding Z L. Chinese loess and the paleomonsoon. Annu Rev Earth Sci, 1998, 26: 111–145 Ding Z L, Derbyshire E, Yang S L, et al. Stacked 2.6-Ma grain size record from the Chinese Loess based on five sections and correlation with the deep-sea δ18O record. Paleocenography, 2002, 17, doi: 10. 1029/2001PA000725 Heller F, Liu T S. Magnetostratigraphical dating of loess deposits in China. Nature, 1982, 300: 431–433 Guo Z T, Ruddiman W F, Hao Q Z, et al. Onset of Asian desertification by 22 Myr ago inferred from loess deposits in China. Nature, 2002, 416: 159–163 Qiang X K, Li Z X, Powell C McA, et al. Magnetostratigraphic record of the late Miocene onset of the East Asian monsoon, and Pliocene uplift of northern Tibet. Earth Planet Sci Lett, 2001, 187: 83–93 An Z S, Huang Y S, Liu W G, et al. Multiple expansions of C4 plant biomass in East Asia since 7 Ma coupled with strengthened monsoon circulation. Geology, 2005, 33: 705–708 An Z S, Kutzbach J E, Prell W L, et al. Evolution of Asian monsoons and phased uplift of the Himalaya-Tibetan Plateau since late Miocene times. Nature, 2001, 411: 62–66 Zhou L P, Oldfield F, Wintle A G, et al. Partly pedogenic origin of magnetic variations in Chinese loess. Nature, 1990, 346: 737–739 Liu Q S, Deng C L, Torrent J, et al. Review of recent developments in mineral magnetism of the Chinese loess. Quat Sci Rev, 2007, 26: 368–385 Deng C L, Vidic N J, Verosub K L, et al. Mineral magnetic variation of the Jiaodao Chinese loess/paleosol sequence and its bearing on long-term climatic. J Geophys Res, 2005, 110, B03103, doi: 10.1029/ 2004JB003451 Deng C L, Shaw J, Liu Q S, et al. Mineral magnetic variation of the Jingbian loess/paleosol sequence in the Northern Loess Plateau of China: Implications for Quaternary development of Asian aridification and cooling. Earth Planet Sci Lett, 2006, 241: 248–259 Sun J M, Liu T S. Stratigraphic evidence for the uplift of the Tibetan Plateau between ~1.1 and ~0.9 Myr ago. Quat Res, 2000, 54: 309–320 Sun J M. Long-term fluvial archives in the Fen Wei Graben, central China, and their bearing on the tectonic history of the India-Asia collision system during the Quaternary. Quat Sci Rev, 2005, 24: 1279– 1286 Zhu Z Y, Ding Z L. The Climatic and Tectonic Evolution in the Loess Plateau of China During the Quaternary (in Chinese). Beijing: Geological Publishing House, 1994. 1–226 Wang X S, Løvlie R, Yang Z Y, et al. Remagnetization of Quaternary eolian deposits: A case study from SE Chinese Chinese Loess Plateau. Geochem Geophys Geosyst, 2005, 6, Q06H18, doi:10.1029/2004GC000901 Liu T S. Loess and the Environment. Beijing: China Ocean Press, 1985. 1–251 An Z S, Ho C K. New magnetostratigraphic dates of Lantian Homo erectus. Quat Res, 1989, 32: 213–221 Thompson R, Oldfield F. Environmental and Magnetism. London: George Alien & Unwin, 1986 Bloemendal J, King J W, Hall F R, et al. Rock magnetism of late Neogene and Pleistocene deep-sea sediments: Relationship to sediment source, diagenetic process, and sediment lithology. J Geophys Res, 1992, 97: 4361–4375 Evans M E, Heller F. Magnetic enhancement and palaeoclimate: Study of a loess/palaeosol couplet across the Loess Plateau of China. Geophys J Int, 1994, 117: 257–264 Fukuma K, Torri M. Variable shape of magnetic hysteresis loops in the Chinese loess-paleosol sequence. Earth Planets Space, 1998, 50:

3000

25

26

27

28

29

30

31

32

33

34 35

36

37

38

39

WU Yi, et al.

Chinese Sci Bull

9–14 Zhu R X, Laj C, Mazaud A. The Matuyama-Brunhes and Upper Jaramillo transitions recorded in a loess section at Weinan, north-central China. Earth Planet Sci Lett, 1994, 125: 143–158 Deng C L, Zhu R X, Verosub K L, et al. Mineral magnetic properties of loess/paleosol couplets of the central Loess Plateau of China over the last 1.2 Ma. J Geophys Res, 2004, 109, B01103, doi:10.1029/ 2003JB002532 Day R, Fuller M, Schmidt V A. Hysteresis properties of titanomagnetites: Grain-size and compositional dependence. Phys Earth Planet Inter, 1977, 13: 260–267 Dunlop D J. Theory and application of the Day plot (Mrs/Ms versus Hcr/Hc): 1. Theoretical curves and tests using titanomagnetite data. J Geophys Res, 2002, 107, B32056, doi:10.1029/2001JB000486 Liu Q S, Jackson M J, Yu Y, et al. Grain size distribution of pedogenic magnetic particles in Chinese loess/paleosols. Geophys Res Lett, 2004, 31, L22603, doi:10.1029/2004GL021090 Liu Q S, Torrent J, Maher B A, et al. Quantifying grain size distribution of pedogenic magnetic particles in Chinese Loess and its significance for pedogenesis. J Geophys Res, 2005, 110, B11102, doi: 10.1029/2005JB003726 Deng C L. Paleomagnetic and mineral magnetic investigation of the Baicaoyuan loess-paleosol sequence of the western Chinese Loess Plateau over the last glacial-interglacial cycle and its geological implications. Geochem Geophys Geosyst, 2008, 9, Q04034, doi: 10. 1029/2007GC001928 Banerjee S K, King J, Marvin J. A rapid method for magnetic granulometry with application to environmental studies. Geophys Res Lett, 1981, 8: 333–336 King J, Banerjee S K, Marvin J, et al. A comparison of different magnetic methods of determining the relative grain size of magnetite in natural materials: some results from lake sediments. Earth Planet Sci Lett, 1982, 59: 404–419 Maher B A. Magnetic properties of some synthetic sub-micron magnetites. Geophys J, 1988, 94: 83–96 Nie J S, King J W, Fang X M. Enhancement mechanisms of magnetic susceptibility in the Chinese red-clay sequence. Geophys Res Lett, 2007, 34, L19705, doi:10.1029/2007GL031430 Zhang W G, Yu L Z, Lu M, et al. Magnetic properties and geochemistry of the Xiashu loess in the present subtropical area of China, and their implications for pedogenic intensity. Earth Planet Sci Lett, 2007, 260: 86–97 Florindo F, Zhu R X, Guo B, et al. Magnetic proxy climate results from the Duanjiapo loess section, southernmost extremity of the Chinese Loess Plateau. J Geophys Res, 1999, 104: 645–659 Liu X M, Liu T S, Paul H, et al. Two pedogenic models for paleoclimatic records of magnetic susceptibility from Chinese and Siberian loess. Sci China Ser D-Earth Sci, 2008, 51: 284–293 Liu X M, Hesse P, Rolph T. Origin of maghaemite in Chinese Loess

September (2010) Vol.55 No.26

40

41

42

43

44

45

46

47

48

49

50

51 52

53

deposits: aeolian or pedogenic? Phys Earth Planet Inter, 1999, 112: 191–201 Guo B, Zhu R X, Bai L X, et al. Rock magnetic properties of a loess-paleosol couple along an N-S transect in the Chinese Loess Plateau. Sci China Ser D-Earth Sci, 2001, 44: 1099–1109 Liu Q S, Banerjee S K, Jackson M J, et al. New insights into partial oxidation model of magnetites and thermal alteration of magnetic mineralogy of the Chinese Loess in air. Geophys J Int, 2004, 158: 506–514 Deng C L, Liu Q S, Pan Y X, et al. Environmental magnetism of Chinese Loess-Plateau sequences (in Chinese). Quat Sci, 2007, 27: 193–209 Deng C L, Zhu R X, Jackson M J, et al. Variability of the temperature-dependent susceptibility of the Holocene eolian deposits in the Chinese Loess Plateau: A pedogenesis indicator. Phys Chem Earth, Part A, 2001, 26: 873–878 Ding Z L, Debryshire E, Yang S L, et al. Stepwise expansion of desert environment across northern China in the past 3.5 Ma and implications for monsoon evolution. Earth Planet Sci Lett, 2005, 237: 45–55 Bloemendal J, Liu X M. An assessment of magnetic and geochemical indicators of weathering and pedogenesis at two contrasting sites on the Chinese Loess Plateau. Palaeogeogr Palaeoclimatol Palaeoecol, 2008, 257: 152–168 Pan Y X, Zhu R X, Liu Q S, et al. Geomagnetic episodes of the last 1.2 Myr recorded in Chinese loess. Geophys Res Lett, 2002, 29: 1282, doi:10.1029/2001GL014024 Wang X S, Yang Z Y, Løvlie R, et al. Environmental magnetism and paleoclimatic interpretation of the Sanmenxia loess-paleosol sequence in the southeastern extremity of the Chinese Loess Plateau. Chinese Sci Bull, 2006, 51: 2755–2762 Bloemendal J, Liu X M. Rock magnetism and geochemistry of two plio-pleistocene Chinese loess-palaeosol sequences-implications for quantitative palaeoprecipitation reconstruction. Palaeogeogr Palaeoclimatol Palaeoecol, 2005, 226: 149–166 Guo Z T, Liu T S, Fedoroff N, et al. Climate extremes in Loess of China coupled with the strength of deep-water formation in the North Atlantic. Glob Planet Change, 1998, 18: 113–128 Wu Y, Zhu Z Y, Rao Z G, et al. L15 at Gongwangling Homo erectus site in Lantian, Shaanxi Province: Magnetic fabrics and its implications (in Chinese). Quat Sci, 2008, 28: 1175–1176 Li J J, Fang X M. Uplift of the Tibetan Plateau and environmental changes. Chinese Sci Bull, 1999, 44: 2117–2124 Qiao Y S, Guo Z T, Hao Q Z, et al. Loess-soil sequences in southern Anhui Province: Magnetostratigraphy and paleoclimatic significance. Chinese Sci Bull, 2003, 48: 2088–2093 Zhao J B, Gu J, Du J. Climate and soil moisture environment during development of the fifth palaeosol in Guanzhong Plain. Sci China Ser D-Earth Sci, 2008, 51: 665–676