Introduction

Variations in insolation due to orbital changes and solar output over decadal to millennial time scales (Wang et al. 2005) may result in seasonal changes of the land–sea atmospheric pressure distribution and may lead to monsoonal circulation. The Asian monsoon is caused by the thermal difference between the Asian landmass and adjacent oceans that is enhanced by the thermal and dynamic effects of the Tibetan Plateau (An et al. 2001). The Asian Monsoon system, which consists of two important sub-systems, i.e. the east Asian monsoon (EAM) and the Indian monsoon (IM), is an integral part of the global climatic system and plays a significant role in the global hydrological and energy cycles (Cai et al. 2010). The monsoon controls the atmospheric heat budget in the Northern Hemisphere and hence, changes in monsoon have a great potential in controlling global climatic change. Understanding the nature of Asian monsoon variability during the Holocene is essential both to understand the present climatic conditions and to predicting future climatic processes.

Recently, the relationship of paleoclimate changes to the EAM and the IM has attracted the attentions of scholars. Hong et al. (2005, 2010) considered an inverse phase relationship of climate variation between the two monsoons in China. However, Zhang et al. (2011) and Stebich et al. (2011) expressed arguments against such an anti-phase relationship. Thus, these inconclusive results indicate that further studies are still urgently needed for deciphering the forcing mechanisms responsible for the behaviors of the two monsoon systems.

The Nanling Mountains, which are the watershed between the Yangtze and Zhujiang River systems and act as a very important geographic division of the mid- and south-subtropical zones in China, are located within the area, and they are strongly influenced by the Asian Monsoon (Gao et al. 1962). The western Nanling Mountain (WNLM) region is located in the transitional belt between the EAM and IM systems (Fig. 1a; Qian et al. 2007; Zhang et al. 2011). Detailed paleoclimatic studies of the WNLM will develop the understanding of the influences of the Asian Monsoon systems on climate change in the monsoonal area of China. In this paper, we present a climatic record derived from a peat section within this region and use these data to explore the history of variations in the Asian monsoon during the late Holocene and investigate possible forcing mechanisms.

Fig. 1
figure 1

Climatic background of the subalpine Daping Swamp in the western Nanling Mountains. a Decadal means of the moisture (specific humidity) transports in July for 1979–1998 in the lower troposphere (redrawn from Qian et al. (2007); Zhang et al. (2011) with minor modifications). Dashed blue lines show the boundaries between westerly and the monsoon moisture. Solid blue lines indicate the rough boundaries between Indian summer monsoon and East ASM. Also shown are the locations of various Asian monsoon records in China cited in Fig. 4: (1) the location of Daping Swamp b and the core site (c, this study); (2) Huguangyan Maar Lake in South China (Yancheva et al. 2007); (3) Dongge Cave in southwest China (Wang et al. 2005); (4) Wanxiang Cave in west Qinling Mountains (Zhang et al. 2008); (5) Hongyuan peat on the eastern fringe of Tibetan Plateau (Hong et al. 2005); (6) Kusai Lake (Liu et al. 2009), and (7) Qinghai Lake (Liu et al. 2006) on the northern and eastern Qinghai-Tibetan Plateau. (Color figure online)

Full size image

Site description

The Daping Swamp is located in the Nanshan Pasture of the Chenbu Miao Nationality Autonomous County in Hunan Province in South China. This region is in the southern Bashili Grand Mountains, part of the Xuefeng Mountains in the WNLM (Fig. 1b). The present annual average temperature of the region is 10.9 °C, and the average temperatures in January and July are −0.5 and 19 °C respectively. The annual precipitation is about 2,000 mm and the annual evaporation is about 500 mm. Local humidity and temperature conditions mean that the original vegetation was likely dominated by evergreen and deciduous broad-leaved vegetation such as Machilus rehderi, Cyclobalanopsis, Fagus and Alnus, as well as shrub vegetation dominated by Enkianthus, Rhododendron and others (Xiao et al. 1986). However, nowadays, the natural vegetation of this region has been strongly decimated by human activity.

The Daping Swamp is about 300-m long and 150-m wide (Fig. 1c). Geologic investigation has revealed that the accumulation of buried peat accounts for more than 100,000 m3 (Sun and Zhang 1984). Granite bedrock formed a near-surface aquiclude that trapped water and established favorable conditions for peat development in the Daping Basin. A pioneering study on this swamp, performed by Sun and Zhang (1984), indicated that the peaty sediments have promising potential for reflecting past climate change.

Materials and methods

In September 2009, we performed field investigations in this swamp. We selected a 95-cm-thick section (designated section HNDP-02, 26°10.472′N, 110°07.223′E, ~1,620 m a.s.l; Fig. 1c) for our investigation of past climate. Samples were taken at 1-cm intervals for analyses of organic carbon isotopes, humification degree (HD) and loss on ignition. Measurements of dry bulk density (DD) were performed on samples collected at 2-cm intervals.

To exclude the reservoir effect of peat materials, we utilized the method proposed by Zhou et al. (2004). Based on an experiment for peaty sediments, they found that 14C dating for the bulk portion of the grain size range between 90- and 300-μm fractions can provide reliable dates that agree very well with the pollen accelerator mass spectrometry 14C ages. Seven bulk samples were collected for conventional radiocarbon dating. We used the Calib 6.0 program and the Intcal 09 data set (Reimer et al. 2009) to calibrate the radiocarbon ages. The chronology of the section was established based on the linear interpolation between neighboring age levels using the basic (non-Bayesian) age-depth modeling software developed by Blauuw (2010).

To measure organic carbon isotopes (δ13C), samples were acid-washed with 5 % HCl to remove traces of carbonate, rinsed with de-ionized water, and oven dried at 50 °C. About 20 mg of each sample were combusted in an excess of oxygen at 1,020 °C by an elemental analyzer. The δ13C of the produced CO2 was measured using a Finnigan MAT-253 Mass Spectrometer. The analytical error is 0.1 ‰. Results are reported in the δ notation relative to the international PDB standard in per mil (‰).

Dry bulk density was measured as the weight of the dry mass per unit volume (Janssens 1983). Dry mass was determined by oven drying at 50 °C until constant weight was achieved. After samples were ground and homogenized, the total organic matter (TOM) was determined by loss on ignition for one and half hours at 550 °C. The fraction below 60 μm was mixed uniformly for measuring the HD. After treatment with 0.1 mol L−1 NaOH and boiling for 1 hour to extract the humic acid, the solution was filtered and diluted. We used a UV-1901 spectrophotometer to measure the absorbance of the solution at a wavelength of 520 nm. The absorbance was expressed relative to the standard value, which was the absorbance of distilled water (defined as 0 %). The absorbance was used to express HD. In this study, we use calibrated HD (HDcal) to indicate the intensity of humification (Zhong et al. 2010b): HDcal = HD/(Msample × TOM), where HDcal is the corrected absorbance value, HD is the raw absorbance value, Msample is the weight of the sample for measurement, and TOM is the percentage content of TOM represented by loss on ignition. Higher or lower HDcal values represent stronger or weaker humification intensity.

To eliminate the impact of degradation of TOM, the stratigraphic trends of TOM and HDcal were removed using the original data minus their 3-order polynomial fits and the results referred to as TOMdetrended and HDcal-detrended.

Results

Lithology and chronology

The lithology of section HNDP-02 is composed primarily of peat (Fig. 2). Below the depth of 95 cm (DP-1), the sediments are dominated by grayish-yellow coarse sand and fine gravels (≤2 cm). The unit from 95–10 cm depth (DP-2) is composed of dark gray peat, mainly bearing herb plant remains. The material of the upper 10 cm in the section (DP-3) consists of black peat with many roots of modern plants and plant remains.

Fig. 2
figure 2

Lithology and relationship between age and depth of section HNDP-02 in Daping Swamp

Full size image

Radiocarbon dating results of the section are listed in Table 1. The bottom age was determined as 2960 cal. yr BP. The average resolution is about 35 years per sample. As illustrated in Fig. 2, the accumulation rates of the upper 50 cm of the section are generally higher than that of the lower part.

Table 1 Radiocarbon dating results for section HNDP-02 in Daping Swamp in western Nanling Mountains
Full size table

Variations of multi-proxy records

In section HNDP-02, the litho-unit DP-1 may originate from the underlying weathered granite crust. The δ13C values range from −30.3 to −28.5 % with a mean value of −29.5 %. The HDcal values vary between 21.3 and 69.4 % with an average value of 52.7 %. The DD values range from 0.07 to 0.24 g cm−3 with a mean value of 0.12 g cm−3, and the value of TOM fluctuates from 35.5 to 59.7 % with a mean value of 49.4 % (Fig. 3).

Fig. 3
figure 3

Variations of multi-proxy records in section HNDP-02 in Daping Swamp The light gray lines indicate the original data of TOM and humification degree The blue bars indicate two dry and cold periods that possibly correspond to the North Atlantic ice-rafting events (i.e., Bond events 1 and 0, Bond et al. 2001). (Color figure online)

Full size image

Discussion

Climatic implication of δ13C, HD, TOM and DD

C3 plants exhibit a large range of carbon isotope compositions (−20–37 ‰), generally reflecting a physiological response to aridity (anomalously high δ13C) and a combination of low light levels plus leaf litter recycling (anomalously low δ13C) (Farquhar et al. 1989). The δ13C values of all samples in HNDP-02 section fall within the range of δ13C values for C3 plants, suggesting that the peat is dominated by C3 plants. In this study, the influences of both the atmospheric CO2 and its δ13C on the δ13C values of C3 plants should be negligible (Feng and Epstein 1995). Investigation of modern C3 plants in north China revealed that a temperature increase of 1 °C causes a decrease in δ13C values of C3 plants of ~0.3 ‰ (Wang et al. 2002). However, it is unlikely that the amplitudes of variation of δ13C values in the section (1.73 ‰) are caused solely by temperature. A number of studies have demonstrated that the δ13C values of C3 plants are also sensitive to humidity or precipitation, the higher the humidity or precipitation, the lower the δ13C values (White et al. 1994; Hong et al. 2001; Hatté et al. 2001; Chen et al. 2006). A broad compilation of modern carbon isotope composition in all C3 plant types also shows a distinct but nonlinear increase in δ13C values with decreasing mean annual precipitation (Kohn 2010). In north China, an increase in precipitation by 100 mm corresponds to a decrease in the δ13C value of C3 plants by 0.5 ‰ (Wang et al. 2003). Although the exact relationship between the δ13C of C3 plants and temperature and precipitation in southern China remains unclear, we infer that values of δ13C in section HNDP-02 that are more negative would indicate relatively warmer and wetter conditions, whereas δ13C values that are more positive imply a cooler and drier climate.

HD is a proxy to describe the degree of decomposition of plant remains. As the process of decomposition of dead plants is related closely to local climatic conditions, this proxy is a useful indicator of the sedimentary environment and past climatic change (Aaby 1976; Chambers et al. 1997; Christopher and John 2000; Charman et al. 2001; Wang et al. 2010; Zhong et al. 2010). Some experimental studies have revealed a nonlinear influence of water and temperature conditions on the decomposition intensity of plants (Chai 1990). The ability of microbes to decompose plant remains is weak when the soil temperature is below 5 °C and when the soil humidity is less than 20 %. With increases in both the temperature and humidity of soil, the activity of microbes is rapidly enhanced and peaks when the soil temperature is about 30 °C and when the soil humidity is around 60 to 80 % (Chai 1990). Therefore, conditions that are warmer and more humid promote vegetation production (i.e., higher TOM) and provide more plant remains for decomposition, thus leading to an increase of humic acids in peat. In contrast, dry and cool conditions lead to a reduction in vegetation production (i.e., lower TOM) and weakened microbial decomposition, which results in a decrease in the humic acids in peat. Therefore, higher and lower values of HDcal-detrended reflect relatively wet and warm versus dry and cool conditions, respectively.

The DD record can provide information related to climatic conditions regarding the external input of clastic materials into the peat. Increased surface vegetation cover around the swamp in response to the wet and warm conditions would decrease surface erosion, limiting the input of clastic materials and resulting in a reduction of DD values. Conversely, dry and cool conditions would result in lower TOM and higher DD values.

Climate variability in the past ~3,000 years

As shown in Fig. 3, multi-proxy records show poor correlations before 2900 cal. yr BP. From ~2900 to 2700 cal. yr BP, lower δ13C and DD values as well as sharp increases in TOMdetrended and HDcal-detrended suggest a wetter and warmer climate, which may have resulted from an intensified summer monsoon. Low TOMdetrended and relatively higher DD values displayed between 2700 and 2500 cal. yr BP, together with two peak intervals of δ13C values, indicate drier and cooler conditions that were probably caused by a weakening of the Asian summer monsoon (ASM) intensity. Lower δ13C and DD values as well as higher TOMdetrended and HDcal-detrended between 2500 and 1700 cal. yr BP reflect enhanced precipitation caused by a strengthening in ASM intensity. This interval may correspond to the Roman Warm Period (Desprat et al. 2003). This warm period has also been detected in the Dongge stalagmite record (2300–2100 yr BP) in southwest China (Dykoski et al. 2005), Huguangyan Lake in Leizhou Peninsula of northern tropical China (~2500–1500 cal. yr BP, Yancheva et al. 2007), Qinghai Lake (~2500–1600 cal. yr BP, Liu et al. 2006), and Kusai Lake on the Tibetan Plateau (~2500–1800 cal. yr BP) (Liu et al. 2009; Fig. 4). A decadal resolution temperature reconstruction for the extratropical Northern Hemisphere (90–30°N) over the last two millennia indicates warm conditions from 1 AD to 300 AD, during which temperatures may have reached the 1961–1990 AD mean temperature level (Ljungqvist 2010).

Fig. 4
figure 4

Correlation of the δ13C record of Daping Swamp with the summer solar insolation at 30°N (Berger and Loutre, 1991), stalagmite δ18O records of Dongge Cave (Dykoski et al. 2005) and Wanxiang Cave (Zhang et al. 2008), the magnetic susceptibility of Huguangyan Maar Lake (Yancheva et al. 2007), TOC record of Kusai Lake (Liu et al. 2009), Alkenone-based temperature of Qinghai Lake (Liu et al. 2006), and δ13C in cellulose of Hongyuan peat (Hong et al. 2005) RWP, DACP, MWP and LIA represent the Roman warm period, the Dark Age cold period, the medieval warm period, and the Little Ice Age, respectively. The green line indicates the varying trend of Daping δ13C record. The red triangles denote three main periods of glacier advances in the southeastern Tibetan Plateau (Yang et al. 2008) The blue arrow lines indicate the intervals that correspond to the Bond Events 1 and 0 (Bond et al. 2001). (Color figure online)

Full size image

Distinct increase in DD (Fig. 3) since 1700 cal. yr BP afterwards and lower accumulation rate in the upper part of the section (~30 cm depth upward, Fig. 2) may result from more input of clastic materials possibly due to the drier and cooler conditions. Gradual increase in δ13C decreases in both TOMdetrended and HDcal-detrended suggest a dry and cool period (~1700–1000 cal. yr BP) that may have been caused by a weakening of the ASM intensity until ~1000 cal. yr BP. This period is characterized by two dry and cool periods (1700–1500 and 1300–1000 cal. yr BP) separated by a short wet and warm interval (1500–1300 cal. yr BP). Similar climatic conditions have also been revealed in the stalagmite δ18O records in Dongge Cave (~1700–900 yr BP, Dykoski et al. 2005) and Wanxiang Cave in the western Qinling Mountains (~200–950 AD, Zhang et al. 2008), the TOC record of Kusai Lake (~1700–1000 cal. yr BP; Liu et al. 2009), the temperature record of Qinghai Lake (~1600–1000 cal. yr BP, Liu et al. 2006), the Ti record of Huguangyan Lake (~ 1500–900 cal. yr BP; Yancheva et al. 2007), and the cellulose δ13C record of Hongyun peat (1800–1300 cal. yr BP, Hong et al. 2005; Fig. 4). This period may correspond to the Dark Age cold period.

A significant increase in HDcal-detrended and lower δ13C values between 1000 and 500 cal. yr BP indicate wetter and warmer conditions resulting from a strengthened ASM intensity. This agrees well with the results derived from the stalagmite δ18O record in Wanxiang Cave (Zhang et al. 2008), and the sediments of Qinghai Lake (~1000–600 cal. yr BP, Liu et al. 2006) and Kusai Lake (~1050–1300 AD, Liu et al. 2009). This warm and wet period correlates well with the medieval warm period. From 500–200 cal. yr BP, higher values of δ13C, lower values of HDcal-detrended, a significant increase in DD, and a decrease in TOMdetrended possibly imply a dry and cool interval, which correlates with the Little Ice Age.

The reconstruction of glacial advance on the southeast Tibetan Plateau also supports our interpretation. Three main periods of glacial advance around 1920–1400 AD, 1150–800 AD, and 600–200 AD (Yang et al. 2008) correspond to the three cool periods identified in section HNDP-02 (i.e., 1700–1500, 1300–1000, and 500–200 cal yr BP). The two cool periods 1700–1500 and 600–200 cal yr BP (within dating errors) show a possible correlation to the Holocene ice rafting events in the North Atlantic (i.e., Bond events 1 and 0, Bond et al. 2001). This suggests that late Holocene climatic events in the North Atlantic had counterpart events in the WNLM region.

Forcing mechanism of climatic variability

The increasing trend of δ13C values in HNDP-2 implies a trend towards dry and cool conditions, suggesting the influence of decreased solar insolation on an orbital scale (Fig. 5).

Fig. 5
figure 5

Correlation of the δ13C record of section HNDP-02 in Daping Swamp with the sunspot number record (Solanki et al. 2004) and the residual atmospheric 14C (△14C) record (Stuiver et al. 1998) Seven bulk 14C dates are shown with an uncertainty interval of 2σ. The blue bars indicate some multi-centennial correlations of the three records. (Color figure online)

Full size image

Many studies have demonstrated a link between the Asian monsoon intensity and solar activity (Hong et al. 2001; Fleitmann et al. 2003; Weber et al. 2004; Wang et al. 2005; Selvaraj et al. 2012; Zhang et al. 2008; Liu et al. 2009). Although the resolution of the Daping section is insufficient for direct comparisons on a short time-scale, we find that on a multi-centennial scale, the δ13C record generally shows an inverse and a positive correlation with the two proxies of solar activity, i.e., the △14C (Stuiver et al. 1998) and the sunspot records (Solanki et al. 2004), respectively (Fig. 5). In particular, the correlations are significant for the interval between 2960 and 2200 cal. yr BP. Using the Redfit program (Schulz and Mudelsee 2002), spectral analysis of the δ13C record reveals three statistically significant centennial cycles centered on 396, 110, and 102-yr (above 90 % confidence, Fig. 6). The 396-yr cycle is very similar to the 400 yr cycle indicated by the △14C record (Sonnett and Suess 1984). The 110-yr cycle is very close to the 113-yr cycle obtained from analysis of the total solar irradiance derived from sunspot activity (Lean et al. 1995; Agnihotri et al. 2002). The 102-yr cycle is very similar to the 104-yr cycle, which is widely present in Asian monsoon records and is considered of solar origin (Neff et al. 2001; Agnihotri et al. 2002; Fleitmann et al. 2003).

Fig. 6
figure 6

Estimated periodicities of climatic changes by Redfit analysis derived from the δ13C record of HNDP-02 section in Daping Swamp

Full size image

Conclusions

The evidence gathered for the study region shows a trend towards dry and cold conditions in the late Holocene. We believe that this is because of a decrease in solar insolation on an orbital scale. Three intensified ASM periods, represented by wet and warm conditions, were revealed at ~2900–2700, 2500–1700, and ~1000–600 cal. yr BP. Three weakened ASM periods, which were characterized by dry and cool conditions, were recorded at ~2700–2500, 1700–1000 and ~600–200 cal. yr BP. The climatic variability of this study is generally identical to various other Chinese monsoon records, reflecting similar multi-centennial monsoonal variation. This study demonstrates that solar irradiance may have played a role in the variation of ASM intensity. Increased solar irradiance favors strengthened ASM intensity, whereas decreased solar irradiance would result in weakened ASM intensity. Although the peat section is not suitable for high-resolution paleoclimatic study and it has not been possible to discern the individual influence of the two summer monsoon systems (i.e., the EAM and IM), this study presents a unique climatic history of the late Holocene in the WNLM region. Our findings show that in the future more precise dating of the sedimentary sequence in the Daping Swamp could improve our understanding of the variation of the Asian monsoon and its effect on global climatic change over geological time scales.