Relationship between pollen assemblages and organic geochemical proxies and the response to climate change in the Zhuye Lake sediments

Yu Li , XueHua Zhou, ChengQi Zhang, ZhuoLun Li, Yue Wang, NaiAng Wang

College of Earth and Environmental Sciences/Center for Hydrologic Cycle and Water Resources in Arid Region, Lanzhou University, Lanzhou, Gansu 730000, China

1 Introduction

Organic geochemical proxies and pollen assemblages are paleoenvironmental indicators which are often used in lake sediment research (Cai and Wei, 1997; Digerfeldtet al., 2000; Wagneret al., 2000; Luet al., 2010; Jianget al., 2011). Total Organic Carbon (TOC), C/N ratio and Organic Carbon Isotope (δ13C) are commonly used as organic geochemical proxies in lake sediments, which can reflect primary productivity and vegetation types in a lake basin, and the conservation state of organic matter(Krishnamurthy and Bhattacharya, 1986; Meyers, 1997;Meyers and Vergès, 1999). Lacustrine pollen assemblages have been widely used to reconstruct vegetation and climatic changes during geological and historical periods(Xiaoet al., 2007; Maet al., 2008; Liet al., 2009). Although these two proxies are frequently used in the reconstruction of paleovegetation and paleoecology, there are some differences between them.

Pollen assemblages emphasis details of paleovegetation while organic geochemical proxies reflect integral changes of paleoproductivity and paleovegetation. In theory, a comparative study between these two indicators would be useful to profoundly comprehend paleoenvironmental changes. Therefore, this paper explores the relationship between pollen assemblages and organic geochemical proxies and the response to climate change.This relationship will provide helpful evidence for reconstruction of past global changes, and useful in understanding environmental significance and changing mechanisms of paleoenvironmental proxies.

This paper uses Zhuye Lake, the terminal lake of the Shiyang River drainage basin, as an example to discuss the relationship between the aforementioned two indica-tors. Records of TOC, C/N ratio and δ13C come from the QTH01 section, while pollen records come from the QTH02 section at Zhuye Lake.

2 Study region

The Shiyang River drainage basin (37°02'N–39°17'N,100°57'E–104°57'E) is located in the eastern Hexi Corridor, central Gansu Province, Northwest China. This basin has an area of 41,600 km2(Figure 1), and the river is one of three major inland rivers in the Hexi Corridor. The basin, which is in a marginal area of the Asian monsoon,is affected both by the Asian monsoon system and westerly winds (Zhao, 1983; Editorial Board of China’s Physical Geography, 1984). The basin belongs to a temperate continental arid climate, and is divided roughly into three climatic zones from south to north: The first zone is Qilian Mountains alpine semi-arid southern humid region. The elevation is 2,000–5,000 m a.s.l., mean annual temperature is lower than 0 °C, annual precipitation is 200–700 mm and annual evaporation capacity is 700–1,200 mm. The vegetation includes alpine cushion,alpine meadow, alpine shrub, forest and piedmont meadow steppe. The second zone is a moderate arid region. The elevation is 1,500–2,000 m a.s.l., annual precipitation is 150–300 mm, annual evaporation capacity is 1,300–2,000 mm and mean annual temperature is 7.8 °C.This region is a desert steppe zone. The last zone is northern warm arid region. The elevation is 1,300–1,500 m a.s.l., annual mean temperature is 6–8 °C, annual precipitation is less than 150 mm and annual evaporation capacity exceeds 2,600 mm. The vegetation is entirely desert vegetation (Chen and Qu, 1992; Huang, 1997; Guo and Li, 2010). The Shiyang River drainage basin has faced multiple ecological problems in recent decades,such as water shortages, falling water tables, water quality degradation and serious desertification (Zhanget al.,2011). Zhuye Lake is a terminal lake of the Shiyang River Basin, which historically had a high lake level (Feng,1963; Li, 1993), but the lake has been completely dried since the early 1960s. Pachuret al.(1995), Zhanget al.(2001, 2004), and our research team (Wanget al., 2011)have studied lake geomorphic evolution since the Late Pleistocene. According to their findings, we know that Zhuye Lake had a higher lake level in the early- and mid-Holocene; however, the lake level is low in the Late Holocene.

3 Materials and methods

The QTH01 and QTH02 sections are located in Zhuye Lake of the Shiyang River. The geographic coordinates are 39°03'N, 103°40'E at an altitude of 1,309 m(Figure 1), and are taken at depths of 6.92 m and 7.36 m,respectively. Liet al.(2009) described the lithology and dates of the two sections (Figure 2). QTH01 section was sampled at 2 cm intervals in the lake sediment layer and at 5 cm intervals otherwise, resulting in 292 samples for analyses of organic geochemical proxies (TOC, C/N,δ13C). QTH02 was sampled at 10 cm intervals, yielding 74 samples for pollen analysis. The two sections were sampled continuously from top to bottom.

Figure 1 The topography around Shiyang River drainage basin

Percent total nitrogen, organic carbon and total hydrogen of QTH01 section were measured by a Vario-3 elemental analyzer, and the experimental methods are reported in Liet al.(2011) (Figure 2). According to lithology of QTH02 section, the sandy samples were sampled at 80 g, the rest were sampled at 40 g for pollen extraction. Fossil pollen was extracted by treating with acid and alkali, then flotation with heavy liquid (KI + HI + Zn)at a gravity of approximately 2.0 g/cm3for 5 minutes or screening after hydrofluoric acid treatment to get rid of silica (Faegri and Iversen, 1989). Fossil pollen was identified and counted at 400 X magnification with an optical microscope made by Olympus.

4 Results

Liet al.(2011) studied the relationship among organic geochemical proxies in the QTH01 section from Zhuye Lake; therefore, this paper will focus on the relationship between pollen records and organic geochemical proxies. In the QTH02 sedimentary section, 74 samples were selected for pollen analysis, and more than 50 families and genera were identified. In this study, we selected eight pollen types to analyze the relationship between organic geochemical proxies and pollen assemblages for different sedimentary phases. These pollen types are dominant in the whole section and commonly seen in the region, includingPinus,Picea,Quercus,Ephedra,Nitraria,Artemisia, ChenopodiaceaeandTypha.

Figure 2 Dates and lithology at QTH01 and QTH02 sections

According to organic geochemical proxies records in QTH01 section, this section can be divided into five phases (Figure 3), including Phase A (602–692 cm),Phase B (447–495 cm), Phase C (495–602 cm), Phase D(199–447 cm) and Phase E (0–199 cm).

Phase A (from the bottom of the section to approximately 13 cal. a B.P.): TOC, C/N ratio and δ13C values are relatively low, and their mean values are 0.07%, 0.65 and -26.76‰ respectively, while total pollen concentration value is also low (Table 1). The relatively low values of TOC and total pollen concentration show that primary productivity, basin-wide effective moisture and vegetation coverage are low correspondingly. In regard to pollen percentages,Artemisiaand Chenopodiaceae are relatively low while tree percentages are relatively high, and there is noTypha(Table 2). The low δ13C value may be related to impacts of cold environments. From the variation range of δ13C (-28.10‰ to -25.50‰), organic matter may originate from C3plants (-21‰ to -33‰), CAM plants (-10‰ to -30‰) and emergent aquatic plants(-24‰ to -30‰). Pollen records reflect that the sources of organic matter are extraneous, and they are contributed by herbs and a small amount of trees, which are carried by wind and come from upstream mountains.

Figure 3 Lithology and TOC, C/N, δ13C values for the QTH01 section in Zhuye Lake

Phase B (approximately 13.0 to approximately 7.7 cal.a B.P.): TOC and C/N ratio values are still low, and their mean values are 0.22% and 3.46, respectively. However,the δ13C value is relatively high with a mean value of-24.89‰. Total pollen concentration increases slightly,but it is still low (Table 1). These records indicate that vegetation types and coverage increase slightly during this phase. In regard to pollen percentages, trees increase rapidly, at the same time, the mean percentages ofPiceaandPinusare 13.85% and 5.26%, respectively, and both reach the highest levels during the Holocene. Meanwhile,percentages of Chenopodiaceae pollen, which stands for an arid environment, reach the minimum value (average 4.49%) (Table 2). From the variation range of δ13C(-27.10‰ to -22.00‰), we know that organic matter may originate from C3plants, CAM plants and emergent aquatic plants. Pollen records reflect that the main sources of organic matter is extraneous, contributed byArtemisiaand trees. Besides, a small amount ofTyphaappearance suggests that the lake level reached the location of this section.

Phase C (approximately 7.7 to approximately 7.4 cal. a B.P.): Compared with the previous phase, TOC and C/N ratio values in Phase C decrease slightly, and their mean values are 0.10% and 1.50, respectively.δ13C value is high (average -23.82‰), while total pollen concentration is low (Table 1). Thus, this phase may be a short-term drought interval, and the input of organic matter is low. In regard to pollen percentages,herbs increase significantly (average 95%), but percentages of trees reduce rapidly, such asPinusandPicea(Tables 1 and 2). From the variation range of δ13C(-25.50‰ to -22.50‰), organic matter may originate from C3plants (-21‰ to -33‰), CAM plants (-10‰to -30‰) and emergent aquatic plants (-24‰ to-30‰). According to the pollen records, the sources of organic matter are mainly extraneous and they are contributed by herbs.

Table 1 The average values and ranges of the TOC, C/N, δ13C at the QTH01 section and corresponding major pollen indices at the QTH02 section in Zhuye Lake

Table 2 The average percentages and ranges of major pollen indices at the QTH02 section in Zhuye Lake

Phase D (approximately 7.4 to approximately 1.1 cal.a B.P.): TOC and C/N radio values reach the highest levels in the Holocene, and their average values are 1.20%and 10.12, respectively. Total pollen concentration also reaches the maximum value; however, δ13C value is low(average -26.68‰) (Table 1). These proxies indicate that primary productivity, basin-wide effective moisture, pollen types and vegetation coverage are correspondingly high during this phase. Percentages of Chenopodiaceae andArtemisiapollen are high (Table 2); meanwhile,some aquatic plants, such asTypha,grow in the lake and the surroundings. The average value of C/N radio is 10.12 (4.22 to 17.62), and that indicates a mixture of aquatic and terrestrial organic matter. As the mean value of δ13C is -26.68‰, and the variation range of δ13C is from -31.00‰ to -23.00‰, we know that organic matter may originate from C3plants (-21‰ to -33‰), CAM plants (-10‰ to -30‰) and emergent aquatic plants(-24‰ to -30‰). Based on pollen records, the sources of organic matter are mainly extraneous herbs, and the contribution of aquatic plants increases relatively.

Phase E (approximately 1.1 to 0 cal. a B.P.): TOC and C/N radio values decrease rapidly, and their mean values are 0.29% and 3.18, respectively. Compared with the previous phase, the δ13C value increases to an average value of -25.01‰ (Table 1), while total pollen concentration declines sharply (average 897 grains/g). In regard to pollen percentages, Chenopodiaceae pollen increases relatively, and its average value is 25.32%.Artemisiapollen decreases to 42.80%, and xerophytic vegetation likeEphedraandNitrariaincreases in this phase (Table 2). Besides, sedimentary facies also change from lacustrine to alluvial deposits, and finally turn to terrestrial sediments. From the variation range of δ13C (-27.40‰ to-24.20‰), organic matter may originate from C3plants,CAM plants and emergent aquatic plants. Pollen records reflect that the sources of organic matter are mainly extraneous, and they are mostly contributed by drought herbs and shrubs, such as Chenopodiaceae,NitrariaandEphedra.

5 Discussions

Environmental reconstruction results of organic geochemical proxies (TOC, C/N and δ13C) and pollen assemblages are different due to their different responses and sensitivity to climate change.

Regarding reconstruction of paleovegetation, values of organic geochemical proxies (TOC, C/N and δ13C)reflect organic matter contents and their sources in lacustrine sediments (Krishnamurthy and Bhattacharya, 1986;Bowen, 1991; Digerfeldtet al., 2000; Wang and Liu,2000). From these values, we can infer vegetation types and primary productivity in a lake basin. The results reconstructed by organic geochemical proxies for different sedimentary phases in Zhuye Lake sediments show that the vegetation types in this drainage basin include C3,CAM and emergent plants. At the same time, pollen assemblages and concentration also indicate various sedimentary phases for QTH02 section. For example, percentages ofPiceapollen reach the highest level during phase B (approximately 13.0 to approximately 7.7 cal. a B.P.), and herb pollen increases rapidly during phase C(approximately 7.7 to approximately 7.4 cal. a B.P.) (Table 2). These changes can reflect detailed variations of vegetation types and quantities for different phases.

As presented in figure 3, basin-wide effective moisture is relatively high before approximately 7.4 cal. a B.P.Correspondingly, TOC and C/N ratio values are high during phase B (approximately 13.0 to approximately 7.7 cal. a B.P.) compared with phases A and C. Organic matter of the terminal lake comes from different parts of the drainage basin; therefore, from a single organic geochemical proxy we cannot infer spatial variability in effective moisture for the entire drainage basin. During phase B, the results of pollen assemblages show that tree pollen percentages reach the highest values in the Holocene as demonstrated by percentages ofPiceathat grows in upstream mountains. But total pollen concentration, TOC (%), C/N values are still low during this phase. These data indicate that effective moisture obviously increases at the upstream mountains, and it is not obvious at middle and lower reaches of the drainage basin. In short, pollen records reflect various effective moisture changes in different parts of the drainage basin,and it is difficult to arrive at this conclusion from organic geochemical proxies. Therefore, the combination of these two indicators provides detailed paleoenvironmental information.

In defining sources of organic matter for lacustrine sediments, the two proxies also have some differences.Organic matter of a lake is divided into endogenous and exogenous. Exogenous organic matter comes from terrestrial plants and aquatic plants surrounding the lake;endogenous organic matter is contributed by phytoplankton, emergent aquatic plants and submerged plants.C/N values reflect the ratio of aquatic organic matter to terrestrial organic matter. Phytoplankton have low C/N values (<10), whereas terrestrial higher plants have high C/N values (14–23) (Meyers, 1997). Plants with different photosynthetic pathways have different δ13C values: δ13C value variation range of C3, C4and CAM plants are respectively from -21‰ to -33‰, from -9‰ to -21‰and from -10‰ to -30‰, and phytoplankton have a low δ13C value (about -35.5‰) (Bowen, 1991; Meyers,1997). The changes of C/N radio in Zhuye Lake sediments do not accurately reflect vegetation types due to a special composition of organic matter in arid environments. δ13C value of the QTH01 section is between-22‰ to -31‰, which indicates that organic matter originates from C3and CAM plants and emergent plants,including endogenous organic matter as well as exoge-nous organic matter. To distinguish the sources of organic matter using pollen assemblages, we need to understand the habitat of each plant.

According to organic geochemical proxies and pollen assemblages in Zhuye Lake sediments, the relatively low values of TOC, C/N ratio and total pollen concentration correspond to relatively high δ13C values during the phases of B, C and E, when the basin is relatively dry;meanwhile, precipitation, primary productivity and vegetation coverage are all low. Therefore, TOC, C/N ratio and total pollen concentration are low. At the same time,δ13C (‰) value is relatively high because of low precipitation, which leads to stomatal closure in plants in order to reduce evaporation; thus CO2concentration declines in the plant leaves and δ13C production by photosynthesis is increased (Tieszen and Button, 1989). During phase D(approximately 7.4 to 1.1 cal. a B.P.), the values of TOC,C/N ratio and total pollen concentration are relatively high when δ13C values are relatively low. Basin-wide primary productivity, vegetation types and coverage are high because of the humid climate during this phase, and that obviously increases TOC, C/N ratio and total pollen concentrations; at the same time, δ13C values are relatively low. The major reasons are as follows: on the one hand, previous studies have found that δ13C values of C3plants in Northwest China are likely to decrease when there is an increasing effective moisture (Tieszen and Button, 1989; Wang and Han, 2001; Wanget al., 2002;Liu, 2011); on the other hand, the increase in phytoplankton leads to δ13C values decreasing because of increasing effective moisture.

In conclusion, organic geochemical proxies and pollen assemblages have their own characteristics in paleoenvironmental reconstruction. Organic geochemical proxies (TOC, C/N and δ13C) are suitable for analyzing macro trends of paleoclimatic changes. However, we must take full consideration of sedimentary environment of organic matter which can affect proxies values, such as lake-water temperature, salinity, plant growth and diagenesis. Pollen analysis is appropriate to find out the delicate difference in macro trends, which will provide some detailed paleoenvironmental change information. Nevertheless, in regard to pollen records in lake sediments, the change of spatial distribution of vegetation in any part of a basin will affect the pollen assemblages in sediments,so we should pay more attention to these influencing factors in research. In a word, using organic geochemical proxies and pollen analysis together are advantageous by reflecting paleoenvironmental changes more accurately and comprehensively.

6 Conclusions

According to a comparative study between organic geochemical proxies and pollen assemblages in the sediments of Zhuye Lake, we found that these two indicators both have their own characteristics. Organic geochemical proxies are good at reflecting the overall trends of paleoenvironment, while pollen assemblages provide detailed information of environmental change.

From results of this study, we know that relatively high values of TOC, C/N ratio and total pollen concentration in lacustrine sediments correspond to low δ13C values; besides, the relatively low values of TOC, C/N ratio and total pollen concentration in sand layers correspond to high δ13C values.

It is the combination of these two indicators that can help us to accurately interpret paleoenvironmental information.

We thank the editor and reviewers for their constructive comments and suggestions for improving our paper. This research was supported by the National Natural Science Foundation of China (Grant Nos. 41371009 and 41001116) and the Fundamental Research Fund for the Central Universities (Grant Nos. lzujbky-2013-127 and lzujbky-2013-129).

Bowen R, 1991. Isotopes and Climates. Elsevier Applied Science,London, pp. 128–131.

Cai MT, Wei MJ, 1997. Climate change during the Penultimate interglacial in Luochuan area. Journal of Desert Research, 29: 536–543.

Chen LH, Qu YG, 1992. Water-land Resources and Reasonable Development and Utilization in the Hexi region. Science Press, Beijing,China, pp. 6–46.

Digerfeldt G, Olsson S, Sandgren P, 2000. Reconstruction of lake-level changes in lake Xinias central Greece during the last 40000 years.Palaeogeography, Palaeoclimatology, Palaeoecology, 158: 65–82.

Editorial Board of China’s Physical Geography (Chinese Academy of Sciences), 1984. Physical Geography of China (climate). Science Press, Beijing, China, pp. 1–30.

Faegri K, Iversen J, 1989. Textbook of Pollen Analysis. Blackwell,New York, pp. 69–89.

Feng SW, 1963. The evolution of the drainage system of the Minchin Oasis. Acta Geographica Sinica, 29: 241–249.

Guo CL, Li FM, 2010. Ecological problem and its rehabilitation countermeasures in Shiyang River basin. Journal of Desert Research, 3:608–613.

Huang DX, 1997. Gansu Vegetation. Gansu Science and Technology,Lanzhou, China, pp. 163–176.

Jiang XY, Li ZZ, Chen XL,et al., 2011. Late Holocene climate and environment changes inferred from pollen recorded in Takelmukul Desert in Yili valley of Xinjiang, China. Journal of Desert Research,31: 855–861.

Krishnamurthy RV, Bhattacharya SK, 1986. Palaeoclimatic changes deduced from δ13C/12C and C/N ratios of Karewa lake sediments India. Nature, 323: 150–152.

Li BC, 1993. A study on the Zhuye Lake and its historical evolution.Acta Geographica Sinica, 48: 55–59.

Li Y, Wang N, Cheng HY,et al., 2009. Holocene environmental change in the marginal area of the Asian monsoon: a record from Lake Zhuye, NW China. Boreas, 38: 349–361.

Li Y, Wang N, Li ZL,et al., 2011. The relationships among organic geochemical proxies and their palaeoenvironmental significances in the Zhuye Lake sediments. Journal of Glaciology and Geocryology,33: 334–341.

Liu XZ, 2011. Relationship between temperature and δ13C values of C3herbaceous plants and its implications of WUE in farming-pastoral zone in North China. Acta Ecologica Sinica, 31: 123–136.

Lu RJ, Wang YJ, Zhang DS, 2010. Climate changes and desert evolution of Muus Desert since 15 ka BP. Journal of Desert Research, 30:273–277.

Ma CM, Zhu C, Zhen CG,et al., 2008. High-resolution humification record for since the Late Glacial climate change at the eastern mountains of China. Science in China (Series D: Earth Sciences), 38:1078–1091.

Meyers PA, 1997. Organic geochemical proxies of paleoceanographic paleolimnologic and paleoclimatic processe. Organic Geochemistry,27: 213–250.

Meyers PA, Vergès EL, 1999. Lacustrine sedimentary organic matter records of Late Quaternary paleoclimates. Journal of Paleolimnology, 21: 345–372.

Pachur HJ, Wunnemann B, Zhang HC, 1995. Lake evolution in the Tengger Desert northwest China during the last 40000 years. Quaternary Reasearch, 44: 171–180.

Tieszen LL, Boutton TW, 1989. Stable carbon isotopes in terrestrial ecosystem research. In: Stable Isotopes in Ecological Research.Springer, New York, pp. 167–195.

Wagner B, Melles M, Hahne J, 2000. Holocene climate history of geographical socienty east greenland-evidence from lake sediments.Palaeogeography, Palaeoclimateology, Palaeoecology, 160:45–68.

Wang G, Han JM, 2001. δ13C variations of C3plants in dry and rainy seasons. Marine Geology and Quaternary Geology, 21: 43–48.

Wang GA, Han JM, Zhou LP, 2002. The relationship between carbon isotope ratios of C3plants and average annual temperature in Northwest China. Geology in China, 29: 55–57.

Wang JQ, Liu JL, 2000. Amino acids and stable carbon isotope distributions in Taihu Lake, China, over the last 15000 years, and their palaeoecological implications. Quaternary Research, 53:223–228.

Wang N, Li ZL, Cheng HY,et al., 2011. Re-investigation of Late Quaternary high lake levels and lager lake at the Alashan Plateau. Chinese Science Bulletin, 56: 1367–1377.

Xiao JY, Lü HB, Zhou WJ,et al., 2007. Since the Last Glacial pollen and environmental change in Jiangxi Great Lakes. Science in China(Series D: Earth Sciences), 37: 789–797.

Zhang HC, Ma YZ, Li JJ,et al., 2001. Palaeolake evolution and abrupt climate changes during last glacial period in NW China. Geophysical Research Letters, 28: 3203–3206.

Zhang HC, Peng JL, Ma YZ,et al., 2004. Late Quaternary palaeolake levels in Tengger Desert, NW China. Palaeogeography, Palaeoclimatology, Palaeoecology, 211: 45–48.

Zhang KX, Liu PX, Huo HL, 2011. Spatio-temporal change characteristics of sunshine hour in Shiyang River basin in recent 50 years.Journal of Desert Research, 31: 517–520.

Zhao SQ, 1983. A new scheme for comprehensive physical regionalization in China. Journal of Geographical Science, 38: 1–10.