• Nie Znaleziono Wyników

Late Glacial and Holocene vegetation and climate history of an alpine wetland on the Qinghai-Tibetan Plateau

N/A
N/A
Protected

Academic year: 2022

Share "Late Glacial and Holocene vegetation and climate history of an alpine wetland on the Qinghai-Tibetan Plateau"

Copied!
8
0
0

Pełen tekst

(1)

Late Gla cial and Ho lo cene veg e ta tion and cli mate his tory of an al pine wet land on the Qinghai–Ti betan Pla teau

Demei LIU1, 2, *, Guichen CHEN2, Zhongping LAI3, 4, Haicheng WEI3, Guoying ZHOU2 and Min PENG2

1 Uni ver sity of Chi nese Acad emy of Sci ences, 19 Yuquan Rd., Beijing 100049, China

2 North west In sti tute of Pla teau Bi ol ogy, Chi nese Acad emy of Sci ences, 23 Xinning Rd., Xining 810008, China

3 CAS Key Lab o ra tory of Salt Lake Re sources and Chenistry, Qinghai In sti tute of Salt Lakes, Chi nese Acad emy of Sci - ences, 18 Xinning Rd., Xining 810008, China

4 Key Lab o ra tory of Qinghai-Ti betan Pla teau En vi ron ment and Re source (MOE), School of Life and Geo graphic Sci ence, Qinghai Nor mal Uni ver sity, 38 West Wusi, Xining 810008, China

Liu D., Chen G., Lai Z., Wei H., Zhou G. and Peng M. (2013) Late Gla cial and Ho lo cene veg e ta tion and cli mate his tory of an al pine wet land on the Qinghai–Ti betan Pla teau. Geo log i cal Quar terly, 57 (2): 261–268, doi: 10.7306/gq.1096

This study pro vides a veg e ta tion and cli mate his tory of an al pine wet land in the cen tre of China’s Qinghai–Ti betan Pla teau (QTP) over the last 14.8 ka. High res o lu tion cli mate prox ies in clude veg e ta tion (pol len), geo chem i cal (Ba, Rb, d13C) and phys i cal (par ti cle size and mag netic sus cep ti bil ity) data; op ti cally stim u lated lu mi nes cence (OSL) and AMS ra dio car bon dat - ing were used to re con struct the en vi ron men tal changes in the wet land eco sys tem. Based on phys i cal and geo chem i cal prox ies, the sucession di vided into five strati graphic units en com pass ing time in ter vals of 14.76–14.31, 14.31–13.17, 13.17–4.47, 4.47–0.51 and 0.51–0 ka. The av er age sed i men ta tion rate of all these units was 0.098 mm/a. The study showed that this eco sys tem was sen si tive to cli mate change.

Keywords: Late Gla cial, Ho lo cene, Qinghai–Ti betan Pla teau, al pine wet land, op ti cally stim u lated lu mi nes cence, pol len anal y sis.

INTRODUCTION

The Qinghai–Ti betan Pla teau (QTP) has im por tant ef fects on both lo cal and global cli mate. Many palaeoenvironmental stud ies have con strains the na ture of cli mate change since the last inter gla cial pe riod across the QTP (Zhu et al., 2002; Wu et al., 2007; Wang et al., 2009; Zhu et al., 2009, 2010). How ever, a gen eral and in ter na tion ally ac cept able frame work has still not been es tab lished and the mech a nism of cli mate change re - mains un clear (Lis ter et al., 1991; Wang, 2005; Herzschuh et al., 2006; Wu et al., 2007; Zhao et al., 2008; Liu et al., 2010;

Bian et al., 2011). Wetlands serve as ar chives for at mo spheric chem i cal de po si tion and may pro vide use ful re cords of spa tial and tem po ral pat terns of en vi ron men tal and con strain rates and amounts of de po si tion (Mesnage et al., 2002; Kim and Kim, 2010; Margielewski et al., 2011). Al pine wet land cov ers a to tal area of 47,000 km2 on the QTP (Zhao, 1999).

The past cli mate can be re con structed us ing a com bi na - tion of dif fer ent types of proxy re cords (such as ice cores, tree rings, fos sil pol len, bore holes, cor als, and lake and ocean sed -

i ments). Car bon ate in late Ho lo cene marsh de pos its is formed un der iso to pic equi lib rium and re flects the iso to pic com po si - tion of the en vi ron men tal wa ter, and thus the d13C in marsh de - pos its is a proxy of con ti nen tal cli mate. To tal Or ganic Car bon (TOC) is an in dex of or ganic pro duc tiv ity and is sup ported by the Mg/Ca ra tio, which is re lated to pre cip i ta tion and tem per a - ture. The ob jec tive of this study was to eval u ate an al pine wet - land as an ar chive of past en vi ron men tal con di tions by at - tempt ing to re con struct the veg e ta tion and cli mate his tory from mul ti ple sed i men tary prox ies. In the cur rent study, Op ti cally Stim u lated Lu mi nes cence (OSL) and the Ac cel er a tor Mass Spec trom e ter (AMS) tech niques were com bined to es tab lish the chro nol ogy and to de rive sed i ment ac cu mu la tion rates. A suite of geo chem i cal (Rb, Ba, d13C) and phys i cal (par ti cle size, mag netic sus cep ti bil ity) prox ies were mea sured to as sess past cli mate con di tions.

MATERIALS AND METHODS

STUDY AREA

The study area is lo cated in Maqin County (Fig. 1) of the Quoluo Zang Au ton o mous Pre fec ture of Qinghai Prov ince, China. The study site (N34°28’, E100°09’) has an al ti tude of 3760 m above sea level and an an nual av er age air tem per a ture of ~–4°C with no frost-free sea son. The av er age an nual pre cip i -

* Corresponding author: liudemei83@163.com

Received: April 19, 2012; accepted: January 9, 2013; first published online: May 21, 2013

(2)

ta tion is 513 mm and the av er age an nual pan-evap o ra tion is ap prox i mately 3000 mm. A to tal of 20 spe cies of higher plants be long ing to 9 fam i lies and 18 gen era in the sam ple plots has been iden ti fied. The dom i nant fam i lies in clude Gramineae (3 gen era, 3 spe cies), Cyperaceae (4 gen era, 4 spe cies), Compositae (2 gen era, 4 spe cies), Ranunculaceae (3 gen era, 3 spe cies). Scrophulariaceae, Liliaceae, Umbelliferae and Rosaceae in clude one ge nus and one spe cies, re spec tively.

The main char ac ter is tics of the dom i nant spe cies (dom i nance

>5%) are sum ma rized in Ta ble 1.

MATERIALS AND PREPARATION TECHNIQUES SAMPLING

The sed i ment push-core sam ple (150 cm long) was col - lected with an iron spade from a wet land at Maqin in Oc to ber 2010. OSL sam ples were col lected at depths of 24, 54, 85, 113 and 144 cm. The lithostratigraphy of the de pos its in the core is shown in Fig ure 2. Based on sed i men tary tex ture and ver ti cal color vari a tion, three sed i men tary units are iden ti fied: 0–52 cm, red loam (young soil); 52–90 cm, gravel layer (flu vial de pos its);

90–150 cm, gray soil (old wet land soil). Core subsamples were sliced into 3 cm sec tions and stored in plas tic bags for fur ther anal y sis.

LABORATORY ANALYSES

TOC was de ter mined with an or ganic car bon an a lyt i cal in - stru ment (JingDao 5000A, Ja pan), the to tal ni tro gen con tent (TN) was mea sured with an el e men tal an a lyzer (Vario EL III,

Ger many), the mag netic sus cep ti bil ity (´ 10–6 m3/kg, SI sys tem) was mea sured us ing a mag netic sus cep ti bil ity in stru ment (MS2, UK), and the TOC/TN was cal cu lated by mass. The car - bon sta ble iso topes were mea sured us ing a Finnigan MAT Delta Plus mass spec trom e ter cou pled with a Flash EA 1112 el - e men tal an a lyzer. Sta ble iso tope re sults were ex pressed as per mil (‰) de vi a tion rel a tive to the VPDB stan dard and the reproducibility of du pli cate anal y ses was ±1%.

For ra dio car bon dat ing, a sam ple of plant ma te rial re main (at 144 cm depth ) was cho sen. AMS 14C dat ing of the bulk sam - ple was per formed at the Heavy Ion In sti tute of Beijing Uni ver - sity in China.

In OSL dat ing, raw sam ples were treated with 38% H2O2, then with 10% HCl to re move or ganic ma te ri als and car bon - ates. Grains be tween 38 and 63 mm were se lected by dry siev - ing. The 38–63 mm par ti cle frac tion was treated with 35%

H2SiF6 acid for about two weeks to re move feld spars (Lai, 2010), then the quartz grains were washed with 10% HCl and then with wa ter.

Af ter in fra red light check ing, sam ples that showed ob vi ous IRSL sig nals un der in fra red light were re-treated with H2SiF6, and checked again with IRSL to avoid age un der es ti ma tion (Lai et al., 2008). Fi nally, the pure quartz ex tracts were mounted in the cen tral parts (~0.5 mm in di am e ter) of stain less discs (10 mm in di am e ter) with sil i cone oil.

All lu mi nes cence mea sure ments, beta ir ra di a tion and pre - heat treat ments were car ried out on a TL/OSL reader (RisÝ TL/OSL DA-20) cou pled with a 90Sr/90Y beta source. The OSL sig nal was de tected with a 7.5mm thick Hoya U-340 glass fil ter.

OSL sig nals were mea sured for 40 s at a sam ple tem per a ture of 130°C. The quartz OSL was stim u lated with blue di odes Fig. 1. Lo ca tion of the study re gion and site

A – lo ca tion of the study re gion (black square) in China; B – lo ca tion of the study re gion (black square) on the Qinghai–Ti betan Pla teau; C – lo ca tion of the study site (white dot) in Maqin County

Dom i nant spe cies d13C [‰]

Cov er age [%]

Height [cm]

Dom i nance [%]

Kobresia tibetica –27.9 ± 0.003 41.67 ± 10.5 51.83 ± 5.5 24.43 ± 3.8 Puccinellia distans –26.6 ± 0.002 45 ± 8 18.5 ± 4.3 20.06 ± 4.4 Carex tristachya –27.6 ± 0.005 5.17 ± 1.2 17.33 ± 6.2 8.16 ± 2.1 Caltha scaposa Hook –23.7 ± 0.003 2.5 ± 0.8 5.5 ± 1.4 5.47 ± 1.9 Ranunculus pulchellus –25.3 ± 0.004 1.5 ± 0.3 13.17 ± 2.3 5.75 ± 1.4 Potentila anserina –26.2 ± 0.005 6.16 ± 2.1 14.15 ± 2.7 6.55 ± 1.5

T a b l e 1 Char ac ters of mod ern plants com mu ni ties in al pine wet land

(3)

(l = 470 ± 20 nm). Equiv a lent dose (De) was de ter mined with sin gle aliquot re gen er a tive-dose (SAR) pro to col (Murray and Wintle, 2000; Lai et al., 2007). Af ter back ground sub trac tion, sig nals from the first 0.64 s stim u la tion were in te grated for growth curve con struc tion. An a-value of 0.035 ± 0.003 was adopted for quartz (Lai et al., 2007).

The con tents of U, Th and K were mea sured us ing neu tron ac ti va tion anal y sis (NAA). The cos mic ray dose was es ti mated for each sam ple as a func tion of depth, al ti tude and geo mag - netic lat i tude (Prescott and Hutton, 1994).

POLLEN ANALYSIS

Pol len sam ples were sieved, and then were treated with 10% HCL and 10% KOH, for floa ta tion in a heavy liq uid so lu tion (KI + HI + Zn) with a spe cific grav ity of about 1.9 to ex tract pol - len from sed i ments (Wei et al., 2010). Pol len taxa were iden ti - fied un der an op ti cal mi cro scope at 400 ´ mag ni fi ca tion. The pol len per cent ages and in flux di a grams were con structed with Tilia soft ware. The re sults of the pol len anal y sis were pre sented as a per cent age pol len di a gram. The to tal pol len sum com - prises all the pol len counted (at least 300 pol len grains per sam - ple). The pol len con cen trates were mounted in glyc erol gel.

RESULTS

CHRONOLOGY

PREHEATING PLATEAU AND OSL SATURATION

To de ter mine ap pro pri ate pre heat ing con di tions for De es ti - ma tion, De pre heat ing pla teau tests were car ried out. De val ues

were ob tained us ing the SAR method at pre heat ing tem per a tures of 160, 180, 200, 220, 240, 260 and 280°C for 10 s. Four aliquots were mea sured at tem per - a tures. De is in de pend ent of pre heat ing tem per a ture in the range of 160–280°C. There fore, pre heat ing of 260°C for 10 s was em ployed. The re gen er a tive-dose growth curve and OSL de cay curves are il lus trated in Fig ure 3.

DOSE RECOVERY TESTS

In or der to test the re li abil ity of the SAR pro to col, dose re cov ery tests were de signed for sam ple gls-3 and gls-4. The nat u ral OSL sig nal from a sam ple was first re - moved com pletely by ex pos ing the aliquot to blue light in the OSL reader at room tem per a ture. It was then given a known beta dose be fore any mea sure ments. This given dose is treated as an un known nat u ral dose, and then mea sured us ing the SAR. Fig ure 4 shows that the re - sults of dose re cov ery test and the mea sured doses are in agree ment within er ror lim its with the given dose. This sug gests that the SAR pro to col is suit able for these sam ples.

EQUIVALENT DOSE AND AGES

The De av er ages of all mea sured aliquots and all dat ing re sults are listed in Ta ble 2. The age span of all sam ples was from 14.77 to 0.72 ka, and the er ror ranged from 7 to 11.2%. The OSL ages are in strati - graphic or der (Fig. 2).

AMS 14C AGES

At the depth of 144 cm, the ra dio car bon age was 8.500 ± 30 ka BP, and the cal i brated 14C age was 9.511 ± 17 cal. ka BP.

The ra dio car bon age was youn ger than the OSL age (14.77 ± 1.12 Ka) for a sam ple (at the depth of 144 cm), the rea son might be that a youn ger plant frag ment had en tered into the core. Al - though the core com prised peat, pres ent plant roots can reach the bot tom of the core, which can re sult in the con tam i na tion for 14C dat ing. In an en vi ron ment which even tu ally leads to a peat de posit, the en ergy in the flu vial pro cess is low, so that the sed i ment can have enough ex po sure time to light, and the OSL sig nal can be re - set com pletely. As a re sult, OSL ages are cred i ble. In this study only OSL ages were used to es tab lish the chro nol ogy.

The age of ev ery 3 cm in ter val of the core was es ti mated us - ing lin ear in ter po la tion or ex trap o la tion be tween the near est OSL dates. The bot tom (old est) date was taken from the sam - ple at 144 cm (14.77 ± 1.12 ka). The sed i men ta tion rates (Fig. 5) were con structed; the mean sed i men ta tion rate for the en tire core was 0.098 mm/a. Ac cord ing to the five ob tained dates, five in ter vals can be iden ti fied: 0–24, 24–54, 54–85, 85–113 and 113–144 cm. The sed i men tary rates were 0.33, 0.06, 0.04, 0.23 and 0.52 mm/a, re spec tively.

POLLEN ANALYSIS

All sam ples in the five units were mi cro scop i cally ana lysed for pol len (Fig. 6). Pol len pres er va tion was poor. More than 300 pol - len grains were counted for each sam ple. Nine teen taxa were iden ti fied, in clud ing the ma jor taxa pres ent in the com plete pro - file: Cyperaceae, Chenopodiaceae, Ar te mi sia, Betula, Picea, Fig. 2. Lithostratigraphy of the de pos its in the core

(4)

Pinus, Com positae, Ranunculaceae, Rosa ceae, Polygonaceae, Poaceae, Ephedra, Elaeagnaceae, Legumino sae, Lilia ceae, Cruciferae, Thalictrum, Gen tian a ceae and Ge ran iaceae.

Among the 47 sam ples ana lysed, there were three types of tree pol len, 14 types of her ba ceous pol len, and two types of

dwarf shrub. The main (<1%) tree pol len were Pinus, Picea, and Betula;

the main her ba ceous pol len (93%) were Cyperaceae (83%), Gramineae (0.7%), Composi tae (4.6%), Cheno - podiaceae (0.7%), Ar te mi sia (1.7%), Ranunculaceae (1.1%), Rosaceae (0.6%), Polygona ceae (0.7%), and Geraniaceae (0.1%); the main dwarf shrub pol len (3%) were Elaeagna - ceae and Ephe dra. Ac cord ing to pol - len con cen tra tion and spe cies per - cent ages, five lo cal pol len as sem - blage zones (nam ely 1–5) could be dis tin guished from the bot tom to the top of the core (Fig. 6).

In Pol len Zone 1 (14.3–14.8 ka), the to tal pol len con cen tra tion was rel - a tively high. Cyperaceae was dom i - nant, other her ba ceous spe cies such as Chenopodiaceae, Ar te mi sia, Com - positae, Elaeagnaceae and Poaceae be ing pres ent in rel a tively high per - cent ages.

In Pol len Zone 2 (14.3–13.2 ka), the to tal pol len con cen tra tion was at its high est. Her ba ceous spe cies such as Elaeag naceae, Ar te mi sia and Compositae were pres ent in rel a - tively high per cent ages. To tal pol len con cen tra tion and the pro por tion of Cyperaceae ini tially in creased and then reached a peak around 13.8 ka, while Ranunculaceae, Liliaceae and Cruciferae pol len de creased. There was al most no Rosaceae or Gentianaceae pol len in this as sem blage, and per cent ages of tree pol len were rel a tively low.

In Pol len Zone 3 (4.5–13.20 ka), the to tal pol len con cen tra - tion was at its low est. Tree pol len showed a pro por tional in - crease in the se quence. The Picea and Pinus pol len per cent - ages were rel a tively higher than in other pol len zones, and more her ba ceous spe cies were pres ent in this as sem blage, such as Rosaceae, Gentianaceae, Ephedra and Thalictrum. The pro - por tion of Chenopodiaceae, Elaeagnaceae, Ar te mi sia and Compositae re mained rel a tively high, and lo cally reached their max i mum val ues.

In Pol len Zone 4 (0.5–4.5 ka), the to tal pol len con cen tra tion started to in crease. The pro por tion of the Compositae, Rosaceae, and Leguminosae pol len grad u ally de creased, while the pro por tion of Elaeagnaceae in creased.

In Pol len Zone 5 (0–0.5 ka), the to tal pol len con cen tra tion con tin ued to in crease. The pro por tion of tree pol len grad u ally de creased, al though Cyperaceae pol len was still dom i nant.

Among the her ba ceous pol len, Compositae, Rosaceae and Leguminosae kept a sta ble pro por tion. Pol len of Elaeagnaceae and Ephedra was not ob served.

PROXY ANALYTICAL RESULTS

The val ues of TOC, d13C, and the ra tios of TOC/TN ranged from 6 to 175 mg/g, –26.92 to –25.12‰ and 2.99 to 316, re - spec tively. The val ues of Rb, Ba, mag netic sus cep ti bil ity (MS) and grain size ranged from 98.2–119.8 mg/g to 477.6–409 mg/g, Fig. 4. The re sults of dose re cov ery tests for sam ple

gls-3 and gls-4

The dashed line is the ra tio of mea sured dose/given dose

0 10000 20000 30000 40000

0 1 2 3 4 5

Stimulation time [s]

0.0 2.0 4.0 6.0

0 40 80

Laboratory dose [Gy]

0 1000 2000 3000 4000

0 1 2 3 4 5

Stimulation time [s]

67Gy

N TD =

0Gy 11.9Gy

45Gy

TD = 11.9Gy N

0Gy

d

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0

0 20 40 60 80 100 120

Labortory dose [Gy]

A B

C D

Corected OSLCorected OSL Counts/0.16 s Counts/0.16 s

Fig. 3. OSL de cay curves and growth curves for sam ple gls-2 (A, B) and gls-4 (C, D)

(5)

0.49 to 0.92, 29.8 to 71.59 ´ 10 –8 m3/kg (SI sys tem) and 35.5 to 64.5 mm, re spec tively. Five sed i men tary units, from Units 1 to 5 in as cend ing or der, were dis tin guished based on sed i men tary tex tures. The val ues of TOC, d13C, the ra tios of TOC/TN, grain size dis tri bu tion, MS, Ba and Rb vari a tion are show in Fig ure 7.

Unit 1: 14.3–14.8 ka (116–147 cm). The val ues of TOC de - crease and reach their min i mum value at 116 cm (14.3 ka); the d13C and MS, de crease first, and then in crease. The ra tio value of TOC/TN shows the op po site trend. The value of Ba shows mul ti ple fluc tu a tions. Two abrupt changes oc cur at 128 cm (14.5 ka) and 137 cm (14.6 ka).

Unit 2: 13.2–14.3 ka (86–116 cm). The val ues of TOC ini - tially in crease and then be came sta ble. The val ues of d13C and Ba first de crease then in crease. The MS, Rb and grain size de - crease slightly. Higher TOC and lower d13C val ues were found at the depth of 107 cm (13.9 ka).

Sam - ple no.

Depth [cm]

Wa ter con tent

[%]

K [%]

Th [ppm]

U [ppm]

Dose rate [Gy/ka]

De

[Gy]

Age

[ka] Method

gls-1 24 15 1.79 ± 0.07 6.91 ± 0.19 1.75 ± 0.13 2.62 ± 0.15 1.98 ± 0.13 0.72 ± 0.07 OSL gls-2 54 15 1.66 ± 0.07 9.1 ± 0.24 2.33 ± 0.17 2.73 ± 0.13 16.43 ± 0.74 5.62 ± 0.46 OSL gls-3 85 15 1.57 ± 0.06 9.40 ± 0.24 2.47 ± 0.17 2.70 ± 0.13 37.43 ± 1.13 12.95 ± 0.96 OSL gls-4 113 30 1.67 ± 0.06 9.71 ± 0.26 2.77 ± 0.18 2.84 ± 0.14 32.88 ± 0.46 14.19 ± 0.98 OSL gls-5 144 40 1.64 ± 0.06 10.45 ± 0.27 2.5 ± 0.19 2.78 ± 0.14 31.07 ± 1.05 14.77 ± 1.12 OSL

gls-6 144 – – – – – – 8.500 ± 30 AMS14C

All of the sam ples were dated us ing quartz OSL; sam ple gls-5 was dated us ing AMS14C; the el e va tion re fers to that of the sam ple lo ca tion;

for each sam ple, 12 aliquots were mea sured for De de ter mi na tion

T a b l e 2 En vi ron men tal ra di a tion mea sure ments and OSL ages

Fig. 5. Sed i men ta tion rates based on OSL dat ing and depth of the al pine wet land sed i ment core

Fig. 6. Pol len per cent ages of ma jor taxa in 47 pol len sam ples col lected from the al pine wet land Pink bands in di cate abrupt warm cli mate events; blue bands in di cate abrupt cold cli mate events

(6)

Unit 3: 4.469–13.167 ka (47–86 cm). The value of TOC is the low est in the suc ces sions, but MS is higher and reaches its high est value in this se quence at 59 cm (7.0 ka). Grain size de - creases and also reaches its low est value at 59 cm (7.0 ka) in this se quence, im plied stron ger cur rents, more rapid min eral mat ter weath er ing, and higher pre cip i ta tion at the cor ing site.

The Rb con tent show fluc tu ates. An abrupt de crease of d13C val ues, higher MS and Ba con tents is found at 65 cm (8.2 ka).

Unit 4: 0.51–4.4 ka (17–47 cm). Com pared to the pre vi ous unit, the value of TOC in creases and reach its high est value at 47 cm (4.469 ka). The d13C val ues also in crease with age. MS de creases while grain size in creases. The trends of Rb and Ba are as in Unit 3. The d13C val ues abruptly in crease and higher MS con tent is at 32 cm (2.0 ka).

Unit 5: 0–0.51 ka (0–17 cm). The value of TOC first de - creases and then starts to in crease at 8 cm (0.24 ka). The d13C val ues and Ba con tent grad u ally de crease while the grain size and Rb in crease. MS stays sta ble. An in flec tion point in TOC was re corded at 8 cm (0.24 ka).

DISCUSSION AND CONCUSIONS

The com bined sed i ment prox ies of grain size, mag netic sus cep ti bil ity, TOC, d13C, Rb, Ba, the ra tios of TOC/TN, and pol len were used to re con struct past en vi ron men tal change (the last ~14.8 ka) on the QTP in China. The re sults are in ter - preted and dis cussed be low, from the old est sed i ments to the youn gest ones, within lithological units 1 through 5.

Unit 1 (14.3–14.8 ka): the in crease in grain size and Ba con - tent, and the MS, TOC and Rb de crease, sug gests a rel a tively cold and wet en vi ron ment, con sis tent with other proxy time-se - ries on the north east ern QTP (Wischnewski et al., 2011). The mois ture-lov ing tree Betula and her ba ceous Cyperaceae were pres ent in rel a tively large pro por tions de spite lower pol len con - cen tra tions, sug gest ing high rel a tive hu mid ity pre vailed. How - ever, Cyperaceae pol len showed a sharp de crease around 14,300 year, im ply ing a sud den cold or arid event.

Unit 2 (13.2–14.3 ka): over all high per cent ages of or ganic ma te rial and low d13C value could be in ter preted as in dic a tive of

warmer and wet ter con di tions at this site, as pol len pres er va tion and abun dance in creased in Unit 2. Lower per cent ages of de - graded pol len and higher pol len in flux were re corded, sug gest - ing a shift to low en ergy depositional pro cesses as well as an in - crease in veg e ta tion. The wet land taxon Cyperaceae dom i - nated in this unit. The re corded ris ing in flux of her ba ceous, namely Amaranthaceae, Compositae, Poaceae and Polygo - naceae, sug gest higher bi o log i cal pro duc tiv ity for the al pine wet land and sur round ing up lands, in di cat ing an op ti mum pe riod for wet land plant growth.

Unit 3 (4.47–13.17 ka): at the time of the low est value of TOC, when the d13C value and grain size in creased first and then de creased, the sed i men tary rate was 0.04 mm/a. The sed i ment proxy re sults showed that sed i men tary de po si tion nearly came to a halt. The low est num bers of pol len grains and TOC value of the pro file were found in these sand lay ers. There fore, there are two pos si ble cli ma tic sce nar ios for this in ter val in the QTP re gion: ut - most cold and dry or ut most wet and warm. How ever, the in - crease of MS in di cates that the sand lay ers re corded a shift from wet land depositional sys tem to a flu vial depositional sys tem.

Poor pol len pres er va tion in this unit is also an in di ca tor of mov ing wa ter, that could pre vent their de po si tion and pres er va tion (Sol o - mon et al., 1982). The wet land taxa Cyperaceae and Compositae, Poaceae, are well-rep re sented in the pol len re cord.

The up lands were likely a mix of grasses and woody shrubs.

Unit 4 (0.51–4.47 ka): an in crease in the num ber of pol len grains, pol len pres er va tion, organics, and d13C value in a red loam layer in di cated a change to a rel a tively low-en ergy depositional en vi ron ment (Figs. 3 and 4). The qui eter wa ter dy - nam ics (com pared with the pre vi ous unit) re sulted in the ac cu - mu la tion of fine or ganic sed i ment, in di cat ing a shift in pre cip i ta - tion. The proxy re sults in di cated colder and dryer con di tions.

The Ar te mi sia, Polygonaceae, Ephedra, Elaeagnaceae, and Liliaceaepollen show a trend of in crease and were prob a bly as - so ci ated with dry soil and less pre cip i ta tion.

Unit 5 (0–0.56 ka): this unit re corded a drop in d13C and in - crease in grain size, in di cat ing a higher air tem per a ture. The dom i nant taxon was also Cyperaceae, but Compositae, Rosaceae, Leguminosae in creased. The com bined pol len and other prox ies also in di cate that hab i tat shift from a wet land to a Fig. 7. Strati graphic col umns and vari a tions of TOC con tent, d13C val ues, MS, grain size, Rb, Sr, Ba, and TOC/TN ra tios

(7)

meadow en vi ron ment took place in re sponse to dry ing and warm ing pe ri ods dur ing the last 0.5 ka. As an ef fect of warmer and drier sce nario on the suc ces sion of plant com mu ni ties, spe - cies rich ness was changed (Henderson-Sell ers et al., 1995). In the on go ing warm ing and dry ing fu ture (IPCC, 2007), plants of the Cy per aceous in al pine wetlands might be un der threat.

This study showed that six ma jor vegetational and cli mate char ac ters could be dis tin guished, in clud ing (1) an al pine wet - land that started to de velop at 14.8 ka in the QTP re gion; (2) a shift at 13.2 ka to con di tions of higher ef fec tive mois ture and a stepwise warmer cli mate sus tained from 13.2 ka to 4.5 ka; (3) a shift to a sig nif i cantly colder and drier phase at 4.4 ka; (4) sev - er ally warm and dry cli mate over the last 0.5 ka; (5) abrupt cold cli mate events hap pened at 14.6, 8.2, 7.0, 2.0 and 0.2 ka, and abrupt warm cli mate events hap pened at 14.5 and 13.9 ka (Figs. 6 and 7); and (6) the Cyperaceae were dom i nant through out. With a sig nif i cant pos i tive cor re la tion be tween the Cyperaceae pol len and the TOC val ues.

Some re search ers have con cluded that the high est lev els of in un da tion were reached be tween the 13,000 yr BP and 4000 yr BP (Kashiwaya et al., 1995) while the first half (10.8–4.4 cal. ka BP) of the Ho lo cene was char ac ter ized by warm and wet con di tions. Then cli mate be came colder and drier in the sec ond half (4.4–0 ka) of the Ho lo cene (Shi et al., 1993; Fleitmann et al., 2003; Morrill et al., 2003; Wang et al., 2005; Herzschuh, 2006). Our stud ies of this al pine wet land sug - gests that there were more large flu vial dis charge events dur ing mid-Ho lo cene than dur ing the late Ho lo cene. These re sults were con sis tent with pre vi ous stud ies (Berg and Loutre, 1991;

Wu et al., 2007) and suc cess fully ex plain the in flu ence of the In - dian sum mer mon soon dur ing the early Ho lo cene and of the SE Asian sum mer mon soon and the wes ter lies dur ing the mid-Ho - lo cene in the cen tre of the QTP (e.g., An et al., 2000; Fleitmann et al., 2003; Wang et al., 2005; Herzschuh, 2006; Chen et al., 2008; Yang and Scuderi, 2010).

Some re search ers have re cog nised cold events be tween 8.7 and 8.3 cal. ka BP (Hu et al., 1999), 7.0 cal. ka BP, 2.2 cal.

ka BP (Wischnewski et al., 2011), based on a pol len re cord from Lake Zigetang on the Cen tral Ti betan Pla teau. The most pro nounced cold events (with tem per a ture be low those of the pres ent) oc curred on the Cen tral Ti betan Pla teau be tween 8.7–8.3 and ~7.4 cal. ka BP (Van Cam po and Gasse, 1993;

Gasse et al., 1996) and the 8.2 ka cold event was re corded in Green land ice cores (Al ley et al., 1997). Dry events hap pened at about 4 ka BP and 2 ka BP to the south of the Qinghai–Ti - betan Pla teau (Zhao, 2009). There is ev i dence of warm or cold/wet or dry events. How ever, our study has not de tected pe ri ods of abrupt cli mate change, which might be in flu enced by the microcirculation sys tems from a dif fer ent re gion (An et al., 2006; Yang et al., 2008).

Multiproxy stud ies of al pine wet land on China’s QTP are rare. In this study, OSL dat ing, pol len, or ganic con tent, grain size, and MS have been used to re con struct the past en vi ron - men tal and cli mate his tory of the QTP. These pe ren ni ally wet eco sys tems proved to be ef fec tive ar chives of past vegetational and cli mate change. Sed i ment prox ies from the study site re - corded an un steady warm ing ten dency in early Ho lo cene, warm-hu mid con di tions in the mid Ho lo cene, and cold and dry con di tions in the late Ho lo cene.

Ac knowl edge ments. We thank Doc tors Xiangjun Liu, Luchun Yang and Guangping Xu for their as sis tance in the lab - o ra tory. This study was spon sored by The Na tional Key Tech - nol ogy R&D Pro gram (Grant No: 2007BAC30) and One-hun - dred Tal ent Pro ject of CAS(A0961). We are grate ful to Prof.

Zhuo Zheng and one anon y mous re viewer as well as to the ed i - tors W. Granoszewski and T. Peryt for valu able com ments and sug ges tions that helped to im prove the manu script.

REFERENCES

Al ley R.B., Mayewski P., Sow ers T., Stuiver M., Tay lor K.C., Clark P.U. (1997) Ho lo cene cli mate in sta bil ity: a prom i nent, wide - spread event 8200 yr ago. Ge ol ogy, 25: 463–486.

An C.B., Feng Z.D., Barton L. (2006) Dry or hu mid? Mid-Ho lo cene hu mid ity changes in arid and semi-arid China. Qua ter nary Sci - ence Re views, 25 (3–4): 351–361.

An Z.S., Por ter S.C., Kutzbach J.E., Zhou W.J. (2000) Asyn chron - ous Ho lo cene op ti mum of the East Asian mon soon. Qua ter nary Sci ence Re views, 19 (8): 743–762.

Berger A., Loutre M.F. (1991) In so la tion val ues for the cli mate of the last 10 mil lion years. Qua ter nary Sci ence Re views, 10 (4):

297–317.

Bian Y., Jian Z., Weng C., Kuhnt W., Bolliet T., Holbourn A. (2011) A palynological and palaeoclimatological re cord from the south - ern Phil ip pines since the Last Gla cial Max i mum. Chi nese Sci - ence Bul le tin, 56: 2359–2365.

Chen F.H., Yu Z.C., Yang M.L., Ito E., Wang S.M., Madsen D.B., Huang X.Z., Zhao Y., Sato T., Birks H.J.B., Boom er I., Chen J.H., An C.B., Wünnemann B. (2008) Ho lo cene mois ture evo - lu tion in arid cen tral Asia and its out-of-phase re la tion ship with Asian mon soon his tory. Qua ter nary Sci ence Re views, 27:

351–364.

Fleitmann D., Burns S., Mudelsee M. (2003) Ho lo cene forc ing of In dian mon soon re corded in a sta lag mite from south ern Oman.

Sci ence, 300: 1737–1739.

Gasse F., Fontes J.C., Van Cam po E., Wei K. (1996) Ho lo cene en vi - ron men tal changes in Bangong Co Ba sin (West ern Ti bet). Palaeo - ge ogra phy, Palaeoclimatology, Palaeo ec ol ogy, 120: 79–92.

Herzschuh U. (2006) Palaeo-mois ture evo lu tion in mon soonal Cen - tral Asia dur ing the last 50,000 years. Qua ter nary Sci ence Re - views, 25: 163–178.

Herzschuh U., Win ter K., Wünnemann B., Li S. (2006) A gen eral cool ing trend on the cen tral Ti betan Pla teau through out the Ho - lo cene re corded by the Lake Zigetang pol len spec tra. Qua ter - nary In ter na tional, 154–155: 113–121.

Henderson-Sell ers A., McGuffie K. (1995) Global cli mate mod els and “dy namic” veg e ta tion changes. Global Change Bi ol ogy, 1:

63–76.

Hu F., Slawinski D., Wright H.E., Boedigheimer A. (1999) Abrupt changes in North Amer i can cli mate dur ing Early Ho lo cene times. Na ture, 400: 437–440.

IPCC (In ter gov ern men tal Panel on Cli mate Change) (2007) Cli - mate Change 2007 – The Phys i cal Sci ence Ba sis: Work ing

(8)

Group I Con tri bu tion to the Fourth As sess ment Re port of the IPCC. Cam bridge Uni ver sity Press, Cam bridge.

Kashiwaya K., Masuzawa T., Morinaga H., Yaskawa K., Yuan B.Y., Liu J.G., Gu Z.Y. (1995) Changes in hy dro log i cal con di - tions in the cen tral Qing-Zang (Ti betan) Pla teau in ferred from lake bot tom sed i ments. Earth and Plan e tary Sci ence Let ters, 135: 31–39.

Kim H., Kim J. (2010) A 2000-year en vi ron men tal his tory of the Upo Wet land on the Ko rean Pen in sula. Jour nal of Paleolimnology, 44: 189–202.

Lai Z.P. (2010) Chro nol ogy and the up per dat ing limit for loess sam - ples from Luochuan sec tion in the Chi nese Loess Pla teau us ing quartz OSL SAR pro to col. Jour nal of Asian Earth Sci ences, 37:

176–185.

Lai Z.P., Wintle A.G., Thomas D.S.G. (2007) Rates of dust de po si - tion be tween 50 ka and 20 ka re vealed by OSL dat ing at Yuanbao on the Chi nese Loess Pla teau. Palaeo ge ogra phy, Palaeoclimatology, Palaeo ec ol ogy, 248: 431–439.

Lai Z.P., Zöller L., Fuchs M., Brückner H. (2008) Al pha ef fi ciency de ter mi na tion for OSL of quartz ex tracted from Chi nese loess.

Ra di a tion Mea sure ments, 43: 767–770.

Lis ter G.S., Kelts K., Zao C., Yu J., Niessen F. (1991) Lake Qinghai, China: closed-ba sin lake lev els and the ox y gen iso tope re cord for ostracoda since the lat est Pleis to cene. Palaeo ge - ogra phy, Palaeoclimatology, Palaeo ec ol ogy, 84: 141–162.

Liu F., Zhang Y., Feng Z., Hou G., Zhou Q., Zhang H. (2010) The im pacts of cli mate change on the Neo lithic cul tures of Gansu-Qinghai re gion dur ing the late Ho lo cene Megathermal.

Jour nal of Geo graph ical Sci ences, 20: 417–430.

Margielewski W., Ko³aczek P., Michczyñski A., Obidowicz A., Pazdur A. (2011) Re cord of the Meso- and Neoholocene palaeoenvironmental changes in the Jesionowa land slide peat bog (Beskid S¹decki Mts., Pol ish Outer Carpathians).

Geochronometria, 38: 138–154.

Mesnage V., Bonneville S., Laignel B., Lefebvre D., Dupont J.P., Mikes D. (2002) Fill ing of a wet land (Seine es tu ary, France):

nat u ral eutrophication or anthropogenic pro cess? A sedimentological and geo chem i cal study of wet land or ganic sed i ments. Hydrobiologia, 475–476 (1): 423–425.

Morrill C., Overpeck J.T., Cole J.E. (2003) A syn the sis of abrupt changes in the Asian sum mer mon soon since the last deglaciation. Ho lo cene, 13: 465–476.

Murray A.S., Wintle A.G. (2000) Lu mi nes cence dat ing of quartz us - ing an im proved sin gle-aliquot re gen er a tive-dose pro to col. Ra - di a tion Mea sure ments, 32: 57–73.

Prescott J.R., Hutton J.T. (1994) Cos mic ray con tri bu tion to dose rates for lu mi nes cence and ESR dat ing: large depths and long-term time vari a tions. Ra di a tion Mea sure ments, 23: 497–500.

Shi Y.F., Kong Z.Z., Wang S.M. (1993) Mid-Ho lo cene cli mates and en - vi ron ments in China. Global and Plan e tary Change, 7: 233–291.

Sol o mon A.M., Blasing T.J., Sol o mon J.A. (1982) In ter pre ta tion of floodplain pol len in al lu vial sed i ments from an arid re gion. Qua - ter nary Re search, 18: 52–71.

Van Cam po E., Gasse F. (1993) Pol len- and di a tom-in ferred cli ma - tic and hy dro log i cal changes in the Sumxi Co Ba sin (west ern Ti - bet) since 13,000 yr BP. Qua ter nary Re search, 39: 300–313.

Wang J., Zhu L., Nishimura M., Testsuya M. (2009) Spa tial vari - abil ity and cor re la tion of en vi ron men tal prox ies dur ing the past 18,000 years among mul ti ple cores from Lake Pumoyum Co, Ti - bet, China. Jour nal of Paleolimnology, 42: 303–315.

Wang M. (2005) A Study on Ecol ogy and En vi ron ment Pro cess of Ce no zoic Qinghai-Ti bet Pla teau Hin ter land (in Chi nese with Eng lish ab stract). Un pub lished Ph.D. dis ser ta tion, Chengdu Uni ver sity of Tech nol ogy.

Wang Y., Cheng H., Ed wards R.L. (2005) The Ho lo cene Asian Mon soon: links to so lar changes and North At lan tic cli mate. Sci - ence, 308: 854–857.

Wei H.C., Ma H.Z., Zheng Z., Pan A.D., Huang K.Y. (2010) Mod ern pol len as sem blages of sur face sam ples and their re la tion ships to veg e ta tion and cli mate in the north east ern Qinghai-Ti betan Pla teau, China. Re view of Palaeobotany and Palynology, 162:

168–182.

Wischnewski J., Mischke S., Wang Y., Herzschuh U. (2011) Re - con struct ing cli mate vari abil ity on the north east ern Ti betan Pla - teau since the last Lateglacial multi-proxy, dual-site ap proach com par ing ter res trial and aquatic sig nals. Qua ter nary Sci ence Re vues, 30: 82–97.

Wu Y., Andreas L., Bernd W., Li S., Wang S. (2007) Ho lo cene cli - mate change in the Cen tral Ti betan Pla teau in ferred by lac us - trine sed i ment geo chem i cal re cords. Sci ence in China Se ries D:

Earth Sci ences, 50: 1548–1555.

Yang X., Scuderi L. (2010) Hy dro log i cal and cli ma tic changes in deserts of China since the Late Pleis to cene. Qua ter nary Re - search, 73: 1–9.

Yang X., Zhu B., Wang X., Lu Y. (2008) Late Qua ter nary en vi ron - men tal changes and or ganic car bon den sity in the Hunshandake Sandy Land, east ern In ner Mon go lia, China.

Global and Plan e tary Change, 61: 70–78.

Zhao C.H. (2009) Cli mate Evo lu tion of Ho lo cene in East ern Aisa and Its Quan ti ta tive Anal y sis (in Chi nese with Eng lish ab stract).

Nanjing Uni ver sity.

Zhao K. (1999) Marshes and swamps of China: a com pi la tion (in Chi nese). Beijing, Sci ence Press of China.

Zhao Z., Qiao Y., Wang S., Jiang F. (2008) Geo log i cal char ac ter is - tics and evo lu tion of the east ern Qinghai-Ti betan pla teau since the late Ce no zoic. Fron tiers of Earth Sci ence, 2 (2): 209–216.

Zhu L., Zhen X., Wang J., Lü H., Xie M., Kitagawa H., Possnert G.

(2009) A ~30,000-year re cord of en vi ron men tal changes in - ferred from Lake Chen Co, South ern Ti bet. Jour nal of Paleolimnology, 42 (3): 343–358.

Zhu L., Peng P., Xie M., Wang J., Frenzel P., Wrozyna C., Schwalb A. (2010) Ostracod-based en vi ron men tal re con struc - tion over the last 8,400 years of Nam Co Lake on the Ti betan pla - teau. Hydrobiologia, 648 (1): 157–174.

Zhu L.P., Li Y.F., Li B.Y. (2002) The ostracod as sem blages and their en vi ron men tal sig nif i cance in the Chen Co area, south ern Ti bet in re cent 1400 years. Jour nal of Geo graph ical Sci ences, 12 (4):

451–459.

Cytaty

Powiązane dokumenty

The re la tion ship be tween the chem i cal com - po si tion and li thol ogy of Late Gla cial and Ho lo cene biogenic de pos its of the ¯abieniec mire (Cen tral Po land).. Geo

The application of malacological analysis in the study of slope deposits: late Pleis- tocene and Holocene of the Podhale Basin (Carpathians, Poland).. Acta Geologica Polonica, 65

Qua - ter nary plank tonic foraminiferal as sem blage zones of the north - east Gulf of Mex ico, Co lom bia Ba sin (Ca rib bean Sea), and trop - i cal At lan tic Ocean: Graphic cor

OCCURRENCE: Middle Triassic; late Ladinian, Muel- leritortis firma and Muelleritortis cochleata radiolarian zones. In the study area it occurs in the Günlük Formation...

Stable isotope composition (δ 18 O and δ 13 C) of biogenic carbonates derived from the Lake Lednica sediments at Rybitwy, western Poland, was applied to obtain data on climatic

The Imio∏ki site was drilled next to the northwest shore of the Lake Lednica, about 300 metres south of the village of Imio∏ki (Text-fig. 2), where an elongated ridge of sandy

Five lithofacies types (L1–L5) have been dis tin guished in the sandy de pos its (Figs. 2 and 3) based on changes in grain size, tex tural pa ram e ters and their re la

The Late Gla cial and Early Ho lo cene shal low lake his tory in Es to nia is doc u mented from the fresh wa ter ostracod subfossil re cord. Three cores stud ied con sist of Late