Chinese Science Bulletin

**** ******* ** ***** *****

Springer-Verlag

Pollen-recorded climate changes between 13.0 and 7.0

14

C ka BP in southern Ningxia, China

SUN AiZhi1, MA YuZhen1,2, FENG ZhaoDong1,3, LI Fei1 & WU HuiNing1

1

MOE Key Laboratory of Western China s Environmental Systems & College of Earth & Environment Sciences, Lanzhou University,

Lanzhou 73000, China;

2

College of Resources and Technology, Beijing Normal University, Beijing 100875, China;

3

Department of Geology, Baylor University, Waco, TX 76798, USA

A pollen record from Haiyuan section in the southern part of Ningxia revealed a detailed history of

vegetation variation and associated climate changes during the period from ~13.0 to ~7.0 14C ka BP. A

steppe landscape under a moderately dry (and probably cool) condition (~12.7 ~12.1 14C ka BP) was

replaced by a coniferous forest dominating the landscape under a generally wet climate from ~12.1 to

~11.0 14C ka BP. This generally wet period, corresponding to the European B lling/All rod period, can

be divided into three stages: a cool and wet stage between ~12.1 and ~11.4 14C ka BP, a mild and rela-

tively dry stage between ~11.4 and ~11.2 14C ka BP, and a mild and wet stage between ~11.2 and ~11.0

14

C ka BP. The coniferous forest-dominated landscape was then deteriorated into steppe landscape

(~11.0 ~10.6 14C ka BP) and further into a desert steppe landscape from ~10.6 to ~9.8 14C ka BP, being

correspondent to the European Younger Dryas period. After a brief episode of a cool and wet climate

(~9.8 ~9.6 14C ka BP), a relatively mild and dry condition prevailed during the early Holocene (~9.6

~7.6 14C ka BP) and then a warm and humid climate started the mid-Holocene (~7.6 ~7.2 14C ka BP).

pollen record, vegetation history, climate changes, southern Ningxia

large-scale climate changes[5,7,8,11]. It should be noted

To understand the global climate changes and the con-

trolling mechanisms, high-resolution climate records are that general surveys of the Holocene strata including

pollen data from this region were reported[12,13], but un-

needed from different and climatically sensitive regions,

and often typical time periods (e.g., MIS 5a and MIS 2) fortunately the chronologic control was not established.

are targeted to explore the particular features of climatic Here, we present a high-resolution pollen record from

changes during particular periods. The transitional pe- the Haiyuan section to retrieve the histories of vegeta-

riod from the late glacial to the Holocene has been one tion variations and associated climate changes between

~13.0 and ~7.0 14C ka BP. It is expected that the regional

of such targeted periods. Our understanding toward this

period has been greatly improved by recent efforts[1 11].

reconstruction will be helpful in understanding large-

scale climate changes and thus in facilitating our under-

But disputes remain regarding the climatic stability and

standing of the controlling mechanisms during the tran-

the general characteristics[1 11]. To further improve our

sition from the late glacial to the Holocene.

understanding of the transitional period, we studied an

eolian-fluvial section (i.e. Haiyuan section) in the south-

Received July 1, 2006; accepted November 21, 2006

ern part of Ningxia situated in the semiarid northwestern doi: 10.1007/s11434-007-0163-7

Corresponding author (email: abpi3l@r.postjobfree.com)

part of Chinese Loess Plateau. The study area has been Supported by the National Natural Science Foundation of China (Grant Nos.

demonstrated to be one of the most sensitive areas to 40331012 and 40421101) and NSF (Grant Nos. EAR 04-02509 and BCS 06-23478)

www.scichina.com www.springerlink.com Chinese Science Bulletin April 2007 vol. 52 no. 8 1080-1088

ARTICLES

1 Stratigraphy and laboratory methods consists of eight stratigraphical units from the bottom to

the top (Figure 2). The first (850 750 cm) and third

The study area is situated in the loess-covered hilly re-

(580 380 cm) units are gray-yellowish fluvial loess

gion of the southern part of Ningxia (Figure 1), where

layers. The second unit (750 580 cm) is a dark-grayish

the annual mean temperature is 7 9 and the annual

clay layer containing laminated grayish silty sub-layers.

mean precipitation is 300 450 mm. Due to severe hu-

This unit is distinctively characterized by readily ob-

man disturbance, the natural vegetation is persevered servable powdery carbonate, half-decomposed organic

only on the tops of Liang (steeply-sloped loess ridges)

matter (plant litter) and snail shells. The fourth (380

and Mao (dome-like loess hills) and margins of

360 cm) and sixth units (250 220 cm) are dark-grayish

Yuan (flat-topped loess highlands) and dominated by

silty clay layers with low average medium sizes (16.1

Stipa bungeana, S. breviflora, S. grandis, S. kryocii,

and 13.8 m, respectively). The fifth unit (360 250 cm)

Thymus mongolicu, Artemisia gmelinii and Artemisia

is an eolian loess layer with a high average medium size

frigida. Moreover, sub-alpine meadows and secondary

(30.8 m) and the seventh unit (220 65 cm) is a

birch forests are observable in higher elevations in the

strongly pedogenically-altered eolian-fluvial complex

nearby Nanhua Mountain and Quwu Mountain.

with a low average medium size (14.2 m). The top

The Haiyuan section (36 26 N, 105 58 E, 1600 m

layer (65 0 cm) is fluvially-altered minor paleosols

a.s.l.) is a natural exposure on the second terrace of

(Entisol) with a low average medium size (13.2 m) and

Xianma River, a branch of Qingshui River (its drainage

area is about 14480 km2). The section is 8.5 m thick and abundant charcoals and root canals.

GEOLOGY

Figure 1 The location of Haiyuan section.

SUN AiZhi et al. Chinese Science Bulletin April 2007 vol. 52 no. 8-108*-****-****

Figure 2 Lithology, median size and age-depth relationship at Haiyuan section.

The chronology of the section studied (i.e. Haiyuan are no eolian strata of the mid-late Holocene, being quite

section) is based on 9 of 11 AMS 14C dates measured in unusual in the northwestern part of the Chinese Loess

Plateau. However, the stratigraphically-documented

the NSF AMS Facility at University of Arizona (Table

1). Two dates at the depths of 265 cm (10459 57 14C a missing of the mid-late Holocene eolian sequence at the

BP) and 335 334 cm (10987 59 14C a BP) seem too Haiyuan section seems to be well supported by the gen-

eral geological survey data of the southern Ningxia[13].

old and they probably result from contamination by re-

We obtained 210 samples with sampling intervals of

worked older organic matters. We construct an age-

depth model based on 9 AMS 14C dates (1 year = 0.15 2 5 cm. For pollen analysis, exotic Lycopodium tablets

were first added for calculation of pollen concentration

cm) and then extrapolated the ages of the bottom (~12.7

14

C ka BP at 850 cm) and the top (~7. 2 14C ka BP at 0 and the samples were then treated with HCl (5% 10%)

and HF (36%). The pollen residues were mounted on

cm). It means that this section spans the late glacial and

slides and identified using reference materials[14 16]. 34

the Early Holocene. Our field work in the Qingshui

River basin does not provide any clue as to why there families and 48 genera of fossil pollen were identified

Table 1 The AMS data of Haiyuan section

No AA56718 AA56716 AA56717 AA56719 AA56720 AA56721 AA56723 AA56724 AA56725 AA56726 AA56727

Depth(cm) 155-***-***-*** 734 752

25-30-138-***-*** 335-***-***-*** 531

14

C age (aBP) 737*-**-****-** 800*-**-*****-** 109**-**-*****-** 110**-**-*****-** 113**-**-*****-** 12216 64

1082 SUN AiZhi et al. Chinese Science Bulletin April 2007 vol. 52 no. 8 1080-1088

ARTICLES

from samples. More than 300 pollen grains (not includ- component (Pinus, Picea and Platycladus) is present

ing spores) were counted for each sample and the per- and the temperate deciduous component (e.g., Ulmus,

centage of each pollen type was calculated based on the Quercus and Betula) is traceable.

Zone 2 (754 582 cm, ~12.1 ~11.0 14C ka BP):

sum of all counted pollen grains. Particle size of bulk

sample was measured using the Malvern Co. Ltd. Mas- Pinus-Picea-Platycladus-Artemisia assemblage. This

tersizer 2000 laser diffraction particle size analyzer. All zone is approximately within the second stratigraphic

experiments were conducted at MOE Key Laboratory of unit (750 580 cm, a dark-grayish clay layer containing

Western China s Environmental Systems of Lanzhou laminated grayish silty sub-layers) and quite distin-

University. guishable by a fluctuating and rather high pollen con-

centration (up to 7.69 103 grains/g). It is quite notice-

2 Pollen records and results able that the percentage of coniferous tree pollen in-

creases significantly at the expenses of the percentage of

Pollen assemblage zones were identified based on varia-

herb pollen. This zone can be further divided into three

tions in the pollen concentration and the percentages and

subzones:

concentrations of coniferous trees, temperate deciduous

Subzone 2a (754 650 cm, ~12.1 ~11.4 14C ka BP)

trees, and herbs (including shrubs). Six pollen assem-

and subzone 2c (620 582 cm, ~11.2 ~11.0 14C ka BP)

blage zones were identified from the bottom to the top

are very high in the percentages of Pinus pollen (30.47%

(Figures 3 and 4).

Zone 1 (850 754 cm, ~12.7 ~12.1 14C ka BP): 73.95%) and Picea pollen (3.11% 22.19%) with an

extremely high pollen concentration (up to 7.69 103

Artemisia-Chenopodiaceae-Gramineae assemblage. This

grain/g). It is notable that the percentage of Platycladus

zone is within the first stratigraphic unit (850 750 cm,

(up to 17.63%) is also rather high in subzone 2c as well

a gray-yellowish fluvial loess layer) and characterized

by a moderate pollen concentration ((0.08 4.39) 103 as in subzone 2b. Subzone 2b (650 620 cm, ~11.4

~11.2 14C ka BP) is distinguishably lower in pollen con-

grain/g) and a rather high herb pollen percentage

centration, higher in the percentage of herb pollen (e.g.

(70.55% 97.49%). The herb component includes Ar-

Artemisia, Chenopodiaceae and Gramineae) and lower

temisia (up to 64.05%), Chenopodiaceae (up to 44.30%)

in the percentage of coniferous tree pollen (Pinus, Picea

and Gramineae (up to 7.83%), accompanied by Aster

and Tusga) than those in subzones 2a and 2c.

type, Polygonum and Liliaceae. The coniferous trees

GEOLOGY

Figure 3 Pollen percentage assemblage and pollen total concentration of Haiyuan section with chronology.

SUN AiZhi et al. Chinese Science Bulletin April 2007 vol. 52 no. 8-108*-****-****

Figure 4 Main pollen concentrations of Haiyuan section.

Zone 3 (582 395 cm, ~11.0 ~9.8 14C ka BP): dark-grayish silty clay layer), is marked by a shape rise

in the pollen concentration (up to 5.18 103 gain/g), and

Chenopodicene-Artemisia-Gramineae assemblage. This

also characterized by a dramatic increase in the per-

zone approximately within the stratigraphic unit 3 (580

centage of coniferous tree pollen (up to 81.62%) at the

380 cm, a gray-yellowish fluvial loess layer) can be

expenses of herb pollen and temperate deciduous tree

subdivided into two subzones. Subzone 3a (582 530

pollen.

cm, ~11.0 ~10.6 14C ka BP) has a moderate pollen

Zone 5 (365 60 cm, ~9.6 ~7.6 14C ka BP):

concentration ((0.39 3.96) 103 grain/g) and is marked

Gramineae-Artemisia-Chenopodiceae assemblage. This

by an abrupt increase in the percentage of herb pollen

zone approximately corresponds to three stratigraphic

including Artemisia (23.58% 54.30%), Chenopodicene

units (unit 5: eolian loess; unit 6: silty clay; unit 7: pe-

(11.91% 26.70%) and Gramineae (3.46% 5.33%) dogenically-altered eolian-fluvial complex) and is char-

and also by a drastic decrease in the percentage of co- acterized by a very low pollen concentration (~1000

niferous tree pollen consisting of Pinus (1.62% grains/g) and by the dominance of herb pollen (up to

26.42%) and Picea (1.32% 6.50%). Subzone 3b (530 95.58%) including Artemisia (up to 52.61%), Gramineae

395 cm, ~10.6 ~9.8 14C ka BP) is distinguishable (up to 33.44%) and Chenopodiceae (up to 40.74%). The

percentages of coniferous tree pollen (e.g., Pinus and

from subzone 3a by having a lower pollen concentration

Picea) and temperate deciduous tree pollen (e.g., Ulmus)

and a considerably higher percentage of Chenopodicene

are slightly higher in zone 5 than in zone 3. It should be

pollen (up to 72.34%) with a dramatic decrease in the

particularly mentioned that although the herb component

percentage of coniferous tree pollen. It is also noticeable

in zone 5 remains relatively constant, the percentages of

that the percentages of some dry-tolerant herd pollen

major herb types (i.e. Artemisia, Gramineae and

types (e.g. Plumbaginaceae, Ephedra and Tribilus) and

Chenopodiaceae) vary considerably. It is quite notice-

the percentages of temperate deciduous tree pollen types

able that the percentages of Gramineae (7.77%

(e.g., Ulmus and Salix) increase slightly in the subzone

3b. 33.44%) and Ulmus (0.63% 10.84%) reach their peaks

Zone 4 (395 365 cm, ~9.8 ~9.6 14C ka BP): at the depth of 260 215 cm (approximately corre-

Pinus-Picea assemblage. This zone, approximately cor- sponding to the stratigraphic unit 6: silty clay), being the

responding to the stratigraphic unit 4 (380 360 cm, a highest in the entire section (see shaded area in Figure 3).

1084 SUN AiZhi et al. Chinese Science Bulletin April 2007 vol. 52 no. 8 1080-1088

ARTICLES

loess), probably represents a steppe landscape under a

The peaks are accompanied with a relatively high per-

moderately dry condition, seemingly being consistent

centage of Platycladus (up to 9.28%) and also with a

with the data from Liushuwan[30] and Shizitan[31] sec-

relatively low percentage of Chenopodiceae. It is also

noticeable that the percentage of the Saussurea-type tions in north central China.

(2) ~12.1 ~11.0 14C ka BP (Pollen Zone 2). The

pollen reaches its peak (up to 62.12 %) at the depth of

90 60 cm (see shaded area in Figure 3). percentage of coniferous tree pollen increases signifi-

Zone 6 (60 00 cm, ~7.6 ~7.2 14C ka BP): Pinus- cantly at the expense of the percentage of herb pollen

with a fluctuating and rather high pollen concentration

Picea-Gramineae assemblage. This zone is within the

(up to 7.69 103 grain/g). This assemblage suggests that a

stratigraphic unit 8 (65 0 cm, a fluvially-altered minor

coniferous forest replaced a steppe to dominate the

paleosol) and characterized by an abrupt increase in the

landscape when the swamp-wetland facies were depos-

percentage (up to 85.63 %) and concentration (~1000

ited (stratigraphic unit 2: a dark-grayish clay layer), im-

grain/g) of coniferous pollen comprising Pinus (up to

plying that the climatic humidity or effective soil mois-

71.83%), Picea (up to 11.84 %) and Tsuga (up to 2.84%)

ture was greatly improved over the previous period.

and also by decreases in the percentage and concentra-

Within this coniferous forest-dominated period (~12.1

tion of herb pollen.

and ~11.0 14C ka BP), secondary changes are well ex-

pressed by the pollen subzones 2a, 2b and 2c. The sub-

3 Paleovegetation and paleoclimatic

zone 2a (~12.1 ~11.4 14C ka BP) represents a Pine and

evolution

Spruce co-dominated forest. The subzone 2b represents

To reconstruct the vegetation variations and associated an Arborvitae forest, being quite similar to the modern

climate changes between ~13.0 and ~7.0 14C ka BP, we vegetation stands in eastern Gansu and northern

need to understand relationships between modern pollen Shaanxi[32], and the subzone 2c represents an Arborvitae

and the associated vegetation. The following is a brief and Pine co-dominated forest.

summary of the published descriptions of modern pol- (3) ~11.0 ~9.8 14C ka BP. This assemblage ap-

len-vegetation relationships. Pinus is generally over- proximately corresponds to the stratigraphic unit 3 (580

represented in pollen spectrum and a relatively high

380 cm, a gray-yellowish fluvial loess layer) and is

percentage of Pinus pollen (25% 30%) may not be

somewhat similar to pollen zone 1 in that the concentra-

associated with the existence of Pinus forest[17 22]. Picea

tion and percentage of herb pollen increase drastically at

is another over-represented pollen type and the existence the expenses of the concentration and percentage of co-

of spruce forest may be indicated only if the percentage niferous pollen. The early part (~11.0 ~10.6 14C ka BP)

of Picea pollen is rather high (20% 80%)[18 20,23 25].

was a period of an Artemisia-dominated steppe land-

As for Ulmus, some studies indicate that Ulmus is a rep- scape. The later part (~10.6 ~9.8 14C ka BP) was a pe-

resentative type[26,27], while other studies show that it is riod of a Chenopodiaceae-dominated desert-steppe

under-represented[21,28]. Our recent work suggests that landscape, representing the driest period of the entire

GEOLOGY

Ulmus is an under-representative type[22]. Artemisia and study time frame (late glacial and early Holocene).

Chenopodiaceae are generally over-represented in sur- (4) ~9.8 ~9.614C ka BP. The high total pollen con-

face-soil pollen spectrum[17 19,21] and Gramineae is usu-

centration (up to 5.18 103 grains/g) and the high per-

ally under-represented[17 19,29].

centage (up to 81.6 %) of the coniferous tree pollen

According to the pollen assemblages with a consid- (primarily Pinus and Picea), in conjunction with the

eration of pollen concentrations (Figures 3 and 4), major swamp-wetland depositional facies (stratigraphic unit 4:

changes in vegetation and associated climate can be re- a dark-grayish silty clay layer), seem to suggest that a

constructed. coniferous forest re-occupied the landscape, indicating

(1) ~12.7 ~12.1 14C ka BP (Pollen Zone 1). The that the climatic humidity or effective soil moisture in-

creased dramatically at the beginning of the Holocene.

herb pollen (e.g. Artemisia, Chenopodiaceae and

(5) ~9.6 ~7.6 14C ka BP. This zone, approximately

Gramineae) overwhelmingly dominates the pollen spec-

trum with a moderate pollen concentration. The assem- corresponding to three stratigraphic units (unit 5: eolian

blage, corresponding to the stratigraphic unit 1 (fluvial loess; unit 6: silty clay; unit 7: pedogenically-altered

SUN AiZhi et al. Chinese Science Bulletin April 2007 vol. 52 no. 8-108*-****-****

eolian-fluvial complex), is characterized by a very low wet climate. Secondly, the Arborvitae-dominated forest

(subzone 2b: ~11.4 ~11.2 14C ka BP) indicates a rela-

pollen concentration (500 800 grain/g) and also by the

dominance of herb pollen. The relatively high percent- tively mild and dry climate. And thirdly, the Arborvitae

ages of coniferous tree pollen (e.g., Pinus and Picea) and pine co-dominated forest (subzone 2c: ~11.2 ~11.0

14

and temperate deciduous tree pollen (e.g., Ulmus) ap- C ka BP) implies a mild and wet climate.

pear to represent a steppe-forest landscape, but the ex-

4.2 Younger Dryas period

tremely low pollen concentration seems to suggest that

Dramatic increases in the percentage and concentration

the landscape was probably a sparsely-vegetated wood-

of herb pollen and the corresponding decrease of conif-

land dominated by Artemisia, Gramineae, Ulmus and

erous tree pollen suggest that a dry climate prevailed

Arborvitae. It should be noticed that although the herb

between ~11.0 and ~9.8 14C ka BP (calibrated age

percentage is relatively constant throughout this period

bracket: ~12.9 ~11.2 cal ka), being correspondent to

(~9.6 ~7.6 14C ka BP), the percentages of Gramineae,

the European Younger Dryas period (~12.7 ~11.5 cal

Artemisia and Chenopodiceae pollen fluctuate greatly,

ka) and also to Chinese stalagmites-recorded Younger

suggesting that the climatic humidity or effective soil

Dryas period (~12.9 ~11.6 ka BP, TIMS-U ages)[10,39].

moisture fluctuated considerably. The highest percent-

During the Younger Dryas (YD) period, the vegetation

ages of Gramineae and Ulmus pollen at the depth of 260

changed gradually from an Artemisia-dominated steppe

215 cm probably indicate that the vegetation density

to a Chenopodiaceae-dominated desert steppe, reflecting

was increased between ~8.9 and ~8.6 14C ka BP.

that the climate was gradually drying. This YD gradually

(6) ~7.6 ~7.2 14C ka BP: A relatively high percent-

drying in the northwestern part of the Chinese Loess

age (up to 85.63 %) and concentration (~1200 grain/g)

Plateau seems to be well corroborated by those well-

of coniferous pollen and a moderately high percentage

dated YD drying events in the Tibetan Plateau[4] and in

of herb pollen reflect a forest-steppe landscape when a

southeastern China[10]. However, our record from the

fluvially-altered minor paleosol (the stratigraphic unit 8)

Haiyuan section appears to be inconsistent with the re-

developed, being consistent with those reconstructed

cords from the nearby sites in the northern part of the

from the western Chinese Loess Plateau[11] and from

Chinese Loess Plateau (e.g., Yulin area) where a consid-

south central Inner Mongolia Plateau[33,34].

erably wet interval was reportedly bracketed by two dry

events during the Younger Dryas[30,36]. This inconsis-

4 Regional comparison and discussion

tence calls for more detailed studies in the northern part

of the Chinese Loess Plateau and its adjacent areas.

4.1 B lling/All rod period

4.3 Early Holocene

Our study indicates that the wettest climate in the

northwestern part of Chinese Loess Plateau occurred After a brief re-occupation of a coniferous forest proba-

from ~12.1 to ~11.0 14C ka BP (calibrated ages: 14.0 bly under a cool and wet climatic condition from ~9.8 to

~9.6 14C ka BP, a sparsely-vegetated woodland land-

12.9 cal ka) during the section-covered period (late gla-

cial and Early Holocene). This wet period seems to cor- scape dominated by Artemisia, Gramineae, Ulmus and

respond to the European B lling/All rod period and the Arborvitae replaced the coniferous forest landscape

from ~9.6 to ~7.6 14C ka BP probably under a relatively

abrupt beginning and the climatic instability docu-

mented by pollen data for this wet period appear to be mild and dry climate condition. The fluctuating per-

chronologically comparable with those recorded in centages of Gramineae, Artemisia and Chenopodiceae

Greenland ice core[3], deep-sea sediment[2], Guliya ice pollen indicate that the climate varied considerably dur-

core[4] and Chinese stalagmite[10,35], considering the un- ing the early Holocene (~9.6 ~7.6 14C ka BP), being in

certainties of the chronologies. Specifically, three major a reasonable agreement with those early Holocene re-

climate changes occurred during this wet period, being cords from the Tibetan Plateau[9,40] and from the north-

consistent with those records from the nearby northern ern margin of the Chinese Loess Plateau[41,42]. It should

part of Chinese Loess Plateau[36] and even beyond[37 39].

be particularly mentioned that our data from the Hai-

Firstly, the pine and spruce co-dominated forest (sub- yuan section that recorded a relatively mild and dry

zone 2a: ~12.1 ~11.4 14C ka BP) suggests a cool and early Holocene are inconsistent with the records from

1086 SUN AiZhi et al. Chinese Science Bulletin April 2007 vol. 52 no. 8 1080-1088

ARTICLES

returned very briefly (~9.8 ~9.6 14C ka BP) and was

the Liushuwan and Midiwan sections in the nearby Yulin

area where a warm and wet early Holocene was re- followed by a sparsely-vegetated woodland (~9.6 ~7.6

ported[30,36]. Again, this inconsistence calls for further 14

C ka BP). The mid-Holocene starts with the appear-

and more detailed studies to establish a regional model ance of a steppe-forest landscape (~7.6 ~7.2 14C ka

of the early Holocene climate changes.

BP).

(2) The pollen-indicated wettest period from ~12.1 to

5 Conclusions

~11.0 14C ka BP in the northwestern part of Chinese

Loess Plateau corresponds to the European B l-

(1) The vegetation changes inferred from the pollen

ling/All rod period and the climatic deteriorating period

assemblages approximately correspond to the changes in

from ~11.0 to ~9.8 14C ka BP corresponds to the Euro-

the stratigraphic characteristics considering certain lead

and lag responses. The pollen-indicated vegetation suc- pean Younger Dryas period. After a very brief episode of

a cool and wet climate (~9.8 ~9.6 14C ka BP), a mild

cession occurred during the transitional period from the

late glacial to the Holocene (~13.0 ~7.0 14C ka BP) in and dry condition prevailed during the Early Holocene

the following sequence. A steppe (~12.7 ~12.1 14C ka (~9.6 ~7.6 14C ka BP) and a warm and humid climate

BP) was followed by a coniferous forest (~12.1 ~11.0 started the mid-Holocene (~7.6 ~7.2 14C ka BP).

14

C ka BP). The steppe resumed in ~11.0 ~10.6 14C ka

BP and further deteriorated into a desert steppe after- Thanks go to Dr. WANG WeiGuo, Dr. AN ChengBang and Mr. ZHANG Jun

ward (~10.6 ~9.8 14C ka BP). The coniferous forest for their help in field work.

1 Broecker W S. Massive iceberg discharges as triggers for global cli- and the associated environmental changes in the western part of the

mate change. Nature, 1994, 374: 421 424 Chinese Loess Plateau. Palaeogeogr Palaeoclimatol Palaeoecol, 2006,

2 Bond G, Shower W, Cheseby M, et al. A pervasive millennial-scale 241(3-4): 440 456

cycle in North Atlantic Holocene and glacial climates. Science, 1997, 12 Li T B. Holocene strata of the Qingshui River valley at Tongxin

278: 1257 1266 county, Ningxia. Geol Bull China (in Chinese), 1997, 16(3):

3 O Brien S R, Mayevoski P A, Meeker L D, et al. Complexity of 253 258

Holocene climate as reconstructed from a Greenland ice core. Science, 13 Ningxia Bureau of Geology and Mineral Resources. Regional Geol-

1995, 270: 1962 1964 ogy of Ningxia Hui Autonomous Region (in Chinese). Beijing:

4 Yang Z H, Yao T D, Huang C L, et al. The record of Younger Dryas Geological Publishing House, 1990. 228 232

event in Guliya ice core. Chin Sci Bull (in Chinese), 1997, 42(18): 14 Mooer P D, Web J A. Palynology Analytical Manual (in Chinese)

1975 1978 (translated by Li W Y, Xiao X M, Liu G X). Nanning: Guangxi Peo-

5 An Z S, Stepen C P, Zhou W J, et al. Episode of Strengthened Summer ple s Publishing House, 1987

Monsoon Climate of Younger Dryas Age on the Loess Plateau of 15 Wang F X, Qian N F, Zhang Y L, et al. Pollen Flora of China, 2nd ed.

central China. Quat Res, 1993, 39: 45 54 (in Chinese). Beijing: Science Press, 1995

6 Wang P X, Bian Y H, Li B H, et al. The Younger Dryas event in 16 Wang K F, Wang X Z. Palynology Conspectus (in Chinese). Beijing:

west Pacific margin seas. Sci China Ser D-Earth Sci (in Chinese), Peking University Press, 1983

GEOLOGY

1996, 26(5): 452 460 17 Wu Y S, Sun X J. Numerical characteristics between surface pollen

7 Guo Z T, Petti-Maire N, Kropelin S. Holocene non-orbital climatic and vegetation in the forest of Xishan Mountain, Kunming. Acta

events in present-day arid areas of northern Africa and China. Global Botan Sin (in Chinese), 1987, 29: 204 211

and Planetary Change, 2000, 26(1-3): 97 103 18 Li W Y, Yan S. Research of Quaternary Polynology, Chaiwopu Basin

8 Zhou W J, Lu X F, Wu Z S, et al. Peat record reflecting Holocene (In Quaternary Climatic Change and Hydrologic Geological Events of

climatic change in the Zoige Plateau and AMS radiocarbon dating. Chaiwopu Basin, Xinjiang) (in Chinese). Beijing: Ocean Press, 1990.

Chin Sci Bull (in Chinese), 2001, 46(12): 104*-****-**-**

9 Liu X Q, Shen J, Wang S M. A 16000-year pollen record of Qinghai 19 Li W Y. Quaternary Vegetation and Environment of China (in Chi-

nese). Beijing: Science Press, 1998. 48 72

Lake and its paleoclimate and paleoenvi- ronment. Chin Sci Bull (in

20 Yan S. The discussion on the pollen of pine family in surface in Xin-

Chinese), 2002, 47(22): 1931 1936

jiang. Arid Land Geography (in Chinese), 1993, 16(3): 1 9

10 Qin J M, Yuan D X, Cheng H, et al. The Y. D. and climate abrupt

21 Zhang J H, Kong Z C, Du N Q. Pollen analysis of surface samples

events in the early and middle Holocene: Stalagmiter oxygen isotope

from Baihua and Dongling Mountains in Beijing. Mar Geol Quat Geol

record from Maolan, Guizhou, China. Sci China Ser D-Earth Sci,

(in Chinese), 1996, 16(3): 101 113

2005, 48(4): 530 537

22 Ma Y Z, Zhang H C, Pachui H J, et al. Modern pollen-inferred

11 Feng Z D, Tang L Y, Wang, H B, et al. Holocene vegetation variations

SUN AiZhi et al. Chinese Science Bulletin April 2007 vol. 52 no. 8-108*-****-****

Holocene climate change in the Tengger Desert, NW China. The 33 Wang B Y, Sun X J. The preliminary study on the environment of the

Holocene, 2004, 14(6): 841 850 Holocene in Chasuqi, Inner Mongolia. Chin Sci Bull (in Chinese),

23 Yan S, Xu Y Q. Spore-pollen association in surface-soil in Altay, 1997, 42(5): 514 518

Xinjiang. Arid Land Rea (in Chinese), 1989, 1: 26 33 34 Yang Z R, Shi P J, Fang X Q. Preliminary study on palaeovegetation

24 Weng C Y, Sun X J, Chen Y S. Numerical characteristics of pollen and environmental changes since 11 ka BP in Diaojiaohaizi lake area,

assemblages of surface samples from the west Kunlun mountains. Daqing Mountains, Inner Mongolia. Acta Phytoecol Sin (in Chinese),

Acta Botan Sin (in Chinese), 1993, 35: 69 79 1997, 21 (6): 551 563

25 Xu Q H, Wu C, Meng L Y, et al. Pollen assemblage feature of dif- 35 Yuan D X, Cheng H, Edwards R L, et al. Timing, duration and tran-

ferent geomorphologic units and different sendimentary facies, North sitions of the Last Interglacial Asian Monsoon. Science, 2004,

China Plain. Chin Sci Bull (in Chinese), 1994, 39(19): 179*-****-***(23): 575 578

26 Sun X J, Song C Q, Wang B Y, et al. Vegetation variations along 36 Zhou W J, Li X Q, Dong G R, et al. The turf record of high-resolution

southern margin of the Loess Plateau-a case study of pollen record at

in desert-loess transitional zone since Younger Drays. Sci China Ser

Weinan section. Chin Sci Bull (in Chinese), 1995, 40(13):

D-Earth Sci (in Chinese), 1996, 26(2): 118 124

1222 1224

37 Li S, Qiang M R, Li B S, et al. Rapid climate changes at northwestern

27 Li Y Y, Zhang X S, Zhou G S, et al. Quantitative relationships be-

margin of east Asian Monsoon region during the Last Deglaciation.

tween vegetation and several pollen taxa in surface soil from North

Geol Rev (in Chinese), 2004, 50(1): 106 112

China. Chin Sci Bull, 2000, 45(16): 1519 1523

38 Yao T D. The sudden change in climate of the Qinghai-Tibet Plateau

28 Liu H Y, Cui H T, Richard P, et al. The surface pollen of the wood-

at the Last Ice Age: contrast study of the Guliya ice core and the

land~steppe ecotone in southeastern Inner Mongolia. Rev Palaeobot

Greenland GRIP ice core. Sci China Ser D-Earth Sci (in Chinese),

Palynol, 1999, 105: 237 250

1999, 29(2): 175 184

29 Wang B Y, Song C Q, Sun X J. Study on surface pollen in middle

39 Wang Y J, Cheng H, Edwards R L, et al. A high-resolution abso-

Inner Mongolia, China. Acta Botan Sin (in Chinese), 1996, 38(11):

lute-dated Late Pleistocene Monsoon record from Hulu cave, China.

902 909

Science, 2001, 249(14): 2345 2348

30 Li X Q, Zhou W J, An Z S, et al. The palaeovegetation record of

40 Tang L Y, Shen C M, Liu K-B, et al. Changes in South Asian Mon-

monsoon evolution in the desert-loess transition zone for the last 13ka

soon: New high-resolution paleoclimatic records from Tibet, China.

BP. Acta Botan Sin (in Chinese), 2000, 42(8): 868 872

Chin Sci Bull (in Chinese), 1999, 44(18): 2004 2007

31 Xia Z K, Chen G, Zheng G W, et al. Climate background of the evo-

41 Shi Q, Chen F H, Zhu Y, et al. Lake evolution of the terminal area of

lution from Palaolithic to Neolithic cultural transition during the Last

Shiyang River drainage in arid China since the Last Glaciation. Quater

Deglaciation in the middle reaches of the Yellow River. Chin Sci Bull,

Intern, 2002, 93-94: 31 43

2002, 47(1): 71 75

42 Jin H L, Dong G R, Su Z Z, et al. Reconstruction of the spatial patterns

32 Hou X Y. Geography of Vegetation and Chemical Elements of Main

of desert-loess boundary belt in north China during the Holocene.

Dominant Plants in China (in Chinese). Beijing: Science Press, 1982.

Chin Sci Bull, 2001, 46(12): 969 973

79 83

1088 SUN AiZhi et al. Chinese Science Bulletin April 2007 vol. 52 no. 8 1080-1088

Contact this candidate