Palaeogeography, Palaeoclimatology, Palaeoecology 399 (2014) 21–30

Contents lists available at ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology

j ourna l homepage: www.e lsev ie r .com/ locate /pa laeo

Palynological evidence for the latest Oligocene−early Miocenepaleoelevation estimate in the Lunpola Basin, central Tibet

Jimin Sun a,⁎, Qinghai Xu b, Weiming Liu c, Zhenqing Zhang d, Lei Xue e, Ping Zhao f

a Key Laboratory of Cenozoic Geology and Environment, Institute of Geology and Geophysics, Chinese Academy of Sciences, P.O. Box 9825, Beijing 100029, Chinab College of Resources and Environment Science, Hebei Normal University, Shijiazhuang, Chinac Institute of Mountain Hazards and Environment, Chinese Academy of Sciences, Chengdu, Chinad Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences, Changchun 130012, Chinae Xiamen Seismological Survey and Research Center, Xiamen, Chinaf National Meteorological Information Center, Beijing 100081, China

⁎ Corresponding author. Tel.: +86 10 8299 8389.E-mail address: jmsun@mail.igcas.ac.cn (J. Sun).

http://dx.doi.org/10.1016/j.palaeo.2014.02.0040031-0182/© 2014 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 3 June 2013Received in revised form 16 January 2014Accepted 5 February 2014Available online 12 February 2014

Keywords:PaleoaltitudeLunpola BasinPollenCentral TibetLatest Oligocene–earliest Miocene

Uplift of the Tibetan Plateau is ultimately driven by the Cenozoic collision between the Indian and Eurasian platesand their continued convergence. One approach for studying the Tibetan Plateau uplift is to verify thepaleoelevation changes from collision to present day. This is important for understanding both the tectonicsand the climatic effects. The new high resolution palynological record of the uppermost Oligocene to the lowestMiocene strata from the Lunpola Basin indicates that the vegetation types during the latest Oligocene–earliestMiocenewere dominated bymixed coniferous–broadleaved forests being different from themodern steppe veg-etation. By using the CoexistenceApproach to the fossil pollen records, after calibration the effects of temperaturedifference and the lapse rate, amaximumpaleoelevation of 3190±100m aslwas estimated in the Lunpola Basinin the latest Oligocene–earliest Miocene, being 1500 to 2000 m lower compared with the previous oxygen iso-tope paleoelevation in the same region.

© 2014 Elsevier B.V. All rights reserved.

1. Introduction

The uplift of the Tibetan Plateau is the result of the India and Eurasiacollision at least since the early Cenozoic (e.g. Molnar and Tapponnier,1975; Besse et al., 1984; Patrait and Achache, 1984; Harrison et al.,1992; Yin and Harrison, 2000). The Tibetan Plateau is characterizednot only by the highest elevation but also by the largest area amongoro-genic belts on earth. By now,many important tectonic questions are stillbeing debated: what mechanisms can account for the uplift of thehighest plateau? Did the different blocks which constitute the TibetanPlateau have initial elevations before collision? How about the questionof diachronous uplift among different parts of the plateau? When didthe plateau reach the present elevation?

Although there have been many different methods for understand-ing the above questions, one important approach is to study thepaleoelevation of different blocks of Tibet before and after collision(e.g. Xu et al., 1973; Zhao and Morgan, 1985; Coleman and Hodges,1995; Harrison et al., 1995; Fielding, 1996; Chung et al., 1998;Garzione et al., 2000a,b; Tapponnier et al., 2001; Williams et al., 2001;Wang et al., 2006; Deng et al., 2012; Wang et al., 2013).

In recent years, oxygen isotope paleoaltimetry has beenwidely usedin estimating paleoelevation in Tibet (e.g., Garzione et al., 2000a; Cyret al., 2005; Rowley and Currie, 2006; DeCelles et al., 2007; Rowleyand Garzione, 2007; Xu et al., 2013). This technique relies on the sys-tematic trends in the isotopic composition of modern precipitation ofdifferent topography. However, such studies have great uncertaintiesdue to: (1) the commonly used authigenic mineral is carbonate, whichis an unstable mineral under surface geochemical environment, boththe diagenesis and the climate-induced chemical weathering can alterthe isotopic compositions of carbonate; and (2) the Tertiary hydrologi-cal conditions, temperature and water vapor sources must be differentcompared with the present scenario, therefore the present stable isoto-pic altimeter cannot be directly used to the geological past.

Because the Tibetan Plateau elevation will have a direct effect on cli-mate and flora, in this sense, paleobotany provides an alternative poten-tial proxy for estimating paleoelevation. Xu et al. (1973) were the first touse the leaf fossils preserved in an elevation of 5700 m in the late Ceno-zoic deposits in the northern Himalaya to estimate the uplift history ofTibet. However, without considering the effect of Cenozoic climatecooling onplant distributions and elevations, their result underestimatedthe paleoelevation of the Himalaya Mountains. Recently, Spicer et al.(2003) estimated the paleoelevation of southern Tibet and concludedthat the present elevation was reached 15 million years ago (Ma) byusing a new technique on leaf physiognomy. Apart from the studies ofleaf fossils, pollen assemblages have been also used to reconstruct

22 J. Sun et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 399 (2014) 21–30

paleoclimate in the Tibetan Plateau (e.g. Wang et al., 1975; Hoorn et al.,2000; Sun and Wang, 2005; Lu et al., 2007; Dupont-Nivet et al., 2008;Miao et al., 2013a,b).

The Co-existence Approach (CoA) is a method for quantitative ter-restrial paleoclimate reconstructions in the tertiary (Mosbrugger andUtescher, 1997). The basic idea of CoA is very simple and follows thenearest living relative philosophy, with the assumption that the climaticrequirements of a fossil taxon are similar to those of a very close livingrelative. As awhole, CoAhas proven to be very useful in quantitative ter-restrial paleoclimate reconstructions (e.g., Ivanov et al., 2002; Lianget al., 2003; Ivanov et al., 2007; Utescher et al., 2009; Jacques et al.,2011), although there are a few exceptions (e.g., Grimm and Denk,2012; Hao et al., 2012). In this study, the maximum overlap of the envi-ronmental tolerances of all the nearest living relatives is regarded asbeing indicative of the most likely paleoaltitude following the methodused by Song et al. (2010). In this study, the temperature differencebetween the geological past and the present, used for correctingpaleoelevation, is not deduced by the CoA method but by GCMmodel-ing resultswith considerations of the past boundary conditions (e.g., pa-leogeographic position, CO2 concentrations and CH4 values).

Song et al. (2010) used this technique and estimated the Eocene–Miocene elevation of central Tibet. However, their chronology is fromearlier reported microfossil age assignments with great uncertainties

Fig. 1. (a) Digital elevation map of the Tibetan Plateau and its neighboring region showing themapof the study area (revisedWang, 2012); red star indicates the location of the section. (For inversion of this article.)

(e.g. Wang et al., 1975; Xu, 1980; Xia, 1982; Xia, 1983; Bureau ofGeology and Mineral Resources Xizang Autonomous Region, 1993;Rowley and Currie, 2006), and the used pollen data are from previouspalynological records covering a variety of outcrops and lake cores.

The aim of this paper is to reconstruct paleo-vegetation and estimatethe latest Oligocene to early Miocene paleoelevation in central Tibet,based on the new age control and the high resolution pollen record ofthe Cenozoic sediments in the Lunpola Basin.

2. Geological settings, materials and methods

The studied Lunpola Basin is a Tertiary rift basin situated along thecentral part of the Bangong–Nujiang Suture Zone, which extends east–west for a distance of approximately 220 km and a width of 15–20 km(Fig. 1a). The present elevations of this basin vary between 4600 mand 5040 m above sea level (asl). The Cenozoic deposits in the basincan be up to 4000m thick (Xu, 1980). Such deposits are deformed asso-ciated with the Cenozoic thrusting and foldingwithin the basin. The Ce-nozoic sediments mainly consist of two stratigraphic units: the NiubaoFormation in the lower part, and the Dingqing Formation in the upperpart. The Niubao Formation, with a thickness of about 3000 m, ismarked by reddish clastic deposits, dominated by mudstone, sandstoneand gravel, mostly representing fluvial to marginal lake environment. It

main tectonic terranes of Tibet and location of the Lunpola Basin. (b) Schematic geologicalterpretation of the references to color in this figure legend, the reader is referred to theweb

Fig. 3. Different age assignments for the Dingqing Formation in the Lunpola Basin.

23J. Sun et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 399 (2014) 21–30

was assigned an age of Paleocene to Eocene based on previous micro-fossil studies (Bureau of Geology and Mineral Resources XizangAutonomous Region, 1993). The Dingqing Formation is about 1000 mthick and is characterized by lacustrine fine-grained gray mudstone,fine siltstone, and limestone.

In the present work, the studied section (32°04′19″N, 89°36′57″ E,4607 m asl) is cut by the Zagya Zangbo River, located in the northernLunpola basin (Fig. 1b), exposing the Dingqing Formation. However, be-cause the upper part of the Dingqing Formation is not accessible due tothe deep river water, samples were collected only from the middle tolow part of the Dingqing Formation along the northern bank of theriver (Fig. 2).

The ages of the Dingqing Formation vary greatly among different au-thors (Fig. 3). The previous age assignments were mainly based onmicro-fossils.Wang et al. (1975) suggested aMiocene age based on pol-len interpretations. Xu (1980) proposed a Miocene–Pliocene age byinterpreting their fossil ostracode and palynological records. This agewas also used by Rowley and Currie (2006), when they estimated thepaleoelevation of the Lunpola Basin. Xia (1982) attributed the DingqingFormation an age of Oligocene based on ostracodes. Xia (1983) sug-gested an age of late Oligocene to Pliocene based on both pollen andostracoda records.

Because of the great uncertainties of the chronology of the DingqingFormation, it is necessary to have new age control for both paleo-environment and paleoelevation reconstruction.

In this study, a bentonite layer, which is a potassium-rich illitic clayof volcanic ash,was identified from themiddle part of the Dingqing For-mation (Fig. 4). The age of this bentonite layer was dated by measuringthe U–Pb ages of zircon using the Cameca IMS-1280 ion microprobe atthe Institute of Geology and Geophysics, Chinese Academy of Sciences.

Additionally, 430 oriented paleomagnetic samples were collected inthe field in order to acquire high-resolution magnetostratigraphy. Allsamples were subjected to stepwise (averaging 18 steps) thermaldemagnetization by using aMagneticMeasurement Thermal Demagne-tizer (MMTDModel 80). The heating routine is as follows: (i) 50 °C stepsup to 500 °C; (ii) 25 °C steps up to 610 °C; and (iii) 10 °C or 5 °C steps upto 670 °C or 680 °C. Magnetic remanence was measured by using a2G-760 U-channel, three-axis, cryogenic magnetometer housed infield-free space (b300 nT), at the Paleomagnetism and GeochronologyLaboratory of the Institute of Geology and Geophysics, Chinese Acade-my of Sciences. The characteristic remanence directions were deter-mined using principal component analysis (Kirschvink, 1980), anddirections were analyzed using Fisher statistics (Fisher, 1953).

In this study, 111 bulk samples were used for palynological analysis.At least 100 g of sediment was used for each sample in order to acquireenough pollen grains. Samples were treated with HCl (35%) and HF(70%) to remove carbonates and silica. Separation of the palynomorphsfrom the residuewas performedby using ZnCl2 (density=2), followingthe method of Faegri and Iversen (1989). Slides were prepared bymounting the pollen grains in glycerin jelly, and then counted under a

Fig. 2. Photo shows the middle to lower part of th

microscope using ×400magnifications. Pollen identificationswere per-formed using the pollen flora of China (Wang et al., 1997).

3. Results

3.1. Chronology of the Dingqing formation

The bentonite layer within the section is conformably interbeddedin the silty mudstones (Fig. 4). Hence, the dating of the bentonite pro-vides a depositional age for the sedimentary layers that bracket thelayer. Zircons were separated for U–Pb zircon dating, most of themare euhedral, transparent and colorless, and have oscillatory zoningunder cathodoluminescence (CL). All the analyses were conducted onportions of zircon with oscillatory zoning, which is interpreted asreflecting growth in magma, therefore representing the eruption ageof the bentonite layer (He et al., 2012). A total of 30 analyses were per-formed on 30 zircon grains, and 26 yield reasonable U–Pb ages rangingfrom 22.6 ± 0.5 Ma to 24.4 ± 0.4 Ma, concordant data points giving aweighted mean 206Pb–238U age of 23.5 ± 0.2 Ma (2σ, MSWD = 1.1),

e Dingqing Formation in the Lunpola Basin.

Fig. 4. Strata of the middle to lower parts of the Dingqing Formation, noting that there is a bentonite layer (shown by photo) within the section, U–Pb zircon dating yields an ageof 23.6 ± 0.2 Ma.

24 J. Sun et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 399 (2014) 21–30

which is interpreted as the eruption age for the bentonite layer withinthe section. This is, for the first time, an absolute age constraint in theLunpola Basin. Our new chronology demonstrates that the middle tolower Dingqing Formation accumulated in the late Oligocene–earlyMioceneperiod being older than the age used by previous stable isotopepaleoelevation estimation (Rowley and Currie, 2006).

Except this absolute age dating, mammalian fossil within the basinalso provides useful biostratigrahic age control. Although this study fo-cused on the middle to lower part of the Dingqing Formation, mamma-lian fossil of Plesiaceratherium sp. from the upper middle part of theDingqing Formation in the nearby section suggests an age of ~18 Ma(Deng et al., 2012). This age limits the uppermost age of the studied sec-tion to be older than 18 Ma.

Fig. 5. (a)Magnetostratigraphy of the studied Dingqing Formation.Magnetic polarity is comparness based on magnetostratigraphic correlation.

Based on the above age controls, the magnetostratigraphy can becorrelated to the standard polarity time scale. The declination and incli-nation obtained for the characteristic remanent magnetization of thespecimens were used to calculate the virtual geomagnetic pole (VGP)latitude, yielding a magnetic polarity sequence (Fig. 5a). This polaritysequence was correlated to the geological time scale 2004 (GTS 2004)by Gradstein et al. (2004), covering an age range of 25.5–19.8 Ma(Fig. 5a).

Moreover, field observations indicate that the studied Dingqing For-mation is dominated bymudstone representing a stable lacustrine envi-ronment, therefore, there would be a relative constant sedimentationrate of the strata if our age assignment is correct. Based on the tie pointsof paleomagnetic polarity age model, very similar sedimentation rates

edwith the geological time scale of GTS2004 (Gradstein et al., 2004). (b) Age versus thick-

Table 1List of the taxa identified in this study.

Gymnospermous pollen Angiospermous pollen (continued)Podocarpidites CornaceoipollenitesAbietineaepollenites ArtemisiaepollenitesPinuspollenites PotamogetonaciditesParcisporites CompositoipollenitesTaxodiaceaepollenites ChenopodipollisAbiespollenites RanunculaciditesPiceaepollenites UmbelliferaepitesGinkgoretectina PersicarioipollisTsugaepollenites LiliaciditesKeteleeriaepollenites GraminiditesCedripites ScabiosapollisEphedripites Labitricolpites

Angiospermous pollen PalaeocoprosmaditesBetulaepollenites QinghaipollisCarpinipites CruciferaeipitesAlnipollenites PlantaginaceaeQuercoidites EvonymoipitesCupuliferoipollenites MalvacearumpollisJuglanspollenites NitrariaditesCastanopsispollenites LonicerapollisCaryapollenites CaryophylliditesUlmipollenites TyphaceaeRhoipites ImpatiensiditesCeltispollenites TricolpopollenitesFraxinoipollenites TricolporopollenitesMeliaceoidites RetitricolpitesAceripollenites MonocolpopollenitesLiquidambarpollenites SporopollisMoraceoipollenites SpheripollenitesFaguspollenites Pteridophyte pollenHamamelidaceae Trilete sporeMyrtaceidites Monolete sporeRutaceoipollis PterisisporitesIlexpollenites CrassoretitriletesTiliaepollenites LycopodiumsporitesSalixipollenites HymenophyllumsporitesOleoidearumpollenites OsmundaciditesMomipites SinopteridaceaeEuphorbiacites

Fig. 6. Pollen diagram showing percentage abundances of pollen taxa within

25J. Sun et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 399 (2014) 21–30

are yielded (Fig. 5b), further suggesting the reasonable chronology ofthe paleomagnetic polarity correlation.

3.2. Pollen results

Except eight samples which are poor in pollen, all the other sampleshave pollen counts between 177 and 386 (average 227). A total of 75pollen taxa were identified (Table 1), among them, tree, shrub, herband fern taxa occupy 70%, 12%, 11% and 6%, respectively. Major pollentaxa are shown in the pollen diagram (Fig. 6), and photographs ofselected pollen grains are in Fig. 7. The Constrained Incremental Sumof Squares (CONISS) cluster technique within TGView 2.02 was usedto define boundaries between the most distinguishable pollen zones(Grimm, 1987), and we recognize three pollen zones in the DingqingFormation, starting with the oldest (Fig. 6).

Zone I (25.5 to 23.2Ma): Tree pollen occupies 72%, while shrubs andherb only occupy 12% and 11%, respectively. Among the tree pollen, co-niferous species are prominent, dominated by Pinuspollenites (21–54%,averaged 30%), other coniferous species include Piceaepollenites(0–7.2%, averaged 2.6%), Keteleeriaepollenites (0–2.2%, averaged 0.6%),and Podocarpidites (0–4.9%, averaged 1.3%); broadleaved species occupy21% (average), dominated by Quercoidites (2.2–22.3%, averaged 11.7%)and Rhoipites (0–8.9%, averaged 3.5%). Shrub-steppe elements are dom-inated by Ephedripites (2.2–20.6%, averaged 8.9%) and Chenopodipollis(0.4–11.6%, averaged 4.4%). Zone I is characterized by the highest per-cents of Quercoidites, Piceaepollenites, and Liliacidites (Fig. 6). The pollencompositions of Zone I suggest the vegetation was dominated by amixed coniferous–broadleaved forest.

Zone II (23.2 to 21.1 Ma): Tree pollen occupies 74.8%, being thehighest of the section. Shrubs andherb occupy 7.0% and 11%, respectively.Coniferous species are still the major pollen types, dominated byPinuspollenites (31–69%, averaged 50%), Keteleeriaepollenites (0–7.1%, av-eraged 2.1%), and Podocarpidites (0–3.9%, averaged 1.6%); broadleavedspecies are dominated by Quercoidites (0.7–17.4%, averaged 7.1%),Rhoipites (0–9%, averaged 3.2%), and Betulaepollenites (0–17.9%, 2.4%).Shrub-steppe elements are still dominated by Ephedripites (0–12.8%,

the section, indicating CONISS dendrogram and associated pollen zones.

Fig. 7. Representative photographs of pollen grains from the section, the scale bar equals 10 μm.

26 J. Sun et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 399 (2014) 21–30

27J. Sun et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 399 (2014) 21–30

averaged 4.7%) and Chenopodipollis (0–23.3%, averaged 4.1%). Althoughthe vegetation type of Zone II also indicates a mixed coniferous–broadleaved forest, the warm–humid species of Keteleeriaepollenites andPodocarpidites have the highest percents of the whole section, whereasthe dry element of Ephedripites is the minimum (Fig. 6), suggesting theenvironment of Zone II is the most optimum during the studied timespan.

Zone III (21.1 to 19.8 Ma): Tree pollen occupies 64%, being the low-est of thewhole section. Shrubs and herb occupy 17% and 10.4%, respec-tively. Coniferous species are dominated by Pinuspollenites (9–66%,averaged 38.7%), Cedripites (0–5.1%, averaged 1.5%), Piceaepollenites(0–5.5%, averaged 1.1%), and Podocarpidites (0–4.3%, averaged 1.1%).Broadleaved species occupy 18% (average), dominated by Quercoidites(0–21.9%, averaged 8.8%), Rhoipites (0–9.4%, averaged 3.3%), andBetulaepollenites (0–7%, averaged 2.6%). Shrub vegetation is dominatedby Ephedripites (1–25%, averaged 14%), while the steppe vegetation isdominated by Chenopodipollis (0.5–12%, averaged 3.9%). Zone I is char-acterized by the highest percent of the drought-tolerant pollen specie ofEphedripites, suggesting that although the vegetationwas still dominat-ed by amixed coniferous–broadleaved forest, the climate was relativelydry.

3.3. Latest Oligocene to early Miocene elevation

The paleoecological analysis of the fossil pollen taxa, together withthe knowledge of the distributions of their present equivalents, provideinformation about the past vegetation types, paleoenvironment, andpast relief gradient. The pollen record indicates multiple vegetationtypes: the tropical to subtropical evergreen broad-leafed species ofCastanopsispollenites andMeliaceoidites; the abundant temperate decidu-ous broad-leaved species of Betulaepollenites,Quercoidites,Ulmipollenites,and Rhoipites; and the temperate mountain coniferous species ofPiceaepollenites, Abiespollenites, and Cedripites. It is worthy to stress thatthe high abundance of Pinuspollenites pollen, with its ease in air andwater transport (Suc and Drivaliari, 1991), may reflect long distancetransport.

The diversity property of pollen taxa suggests that the vegetationtype is dominated bymixed coniferous–broadleaved forests in the latestOligocene–earliest Miocene in the Lunpola Basin. Moreover, the co-existences of sub-tropical, temperate, and mountain coniferous species

Fig. 8. Plot of altitude ranges of presumed nearest living relatives of modern Tibetan vegetationiments in the Lunpola Basin.

indicate vertical zonal vegetation, implying that the existence of elevat-ed areas or neighboring high mountains.

More precise elevation estimation can be done by using the Co-existing Approach (Mosbrugger and Utescher, 1997). This method aimsto reconstruct ranges, which are tolerated by all of the nearest living rel-atives of the fossil plant taxa found in the samples. The maximum over-lap of the environmental tolerances of all the nearest living relatives isthen regarded as being indicative of the most likely paleoenvironment(Song et al., 2010).

Because each modern plant species has its own altitudinal range inTibet. Such data had been well compiled by Wu (1983, 1985, 1986a,1986b, 1987). In using the Co-existing Approach, we compile the altitu-dinal ranges of modern Tibetan vegetation equivalents of fossil pollenassemblages in the latest Oligocene–earliest Miocene (Fig. 8). Themax-imum overlap of all the nearest living relatives provides a Co-existingApproach altitude of 2400–2600 m asl, averaged at 2500 m asl (Fig. 8).

Nevertheless, this paleoelevation estimate is only based on the alti-tudinal ranges of modern vegetation equivalents in Tibet. However,due to the Cenozoic climatic cooling and the India–Asia collision in-duced changing paleogeographic positions, this estimated altitude re-quires correcting for secular climate change with consideration of thepast lapse rates and the paleogeographic positions (Meyer, 2007).

Previous work has estimated the paleotemperature and lapse ratesin the same region for the early Eocene and the middle Miocene (Songet al., 2010). Moreover, a deep-sea temperature record for the past50 Ma has been reconstructed from the magnesium/calcium ratio(Mg/Ca) of benthic foraminiferal calcite which shows a consistent longterm decrease over the past 50 Ma (Miller et al., 1987; Lear et al.,2000). According to this temperature curve, the temperature at the lat-est Oligocene–earliestMiocenewas between the Eocene and themiddleMiocene. Therefore, we can use the averaged lapse rate (5.79 °C/km)and the temperature difference (4 °C) during the two time slices(Eocene and middle Miocene) estimated by Song et al. (2010) to yielda corrected altitude of 690 m. The relatively higher temperatureand the lower lapse rate (being less than the present lapse rate of6.4 °C/km at Lhasa) during the latest Oligocene–earliest Miocene inthe Lunpola Basin allows the raw CoA altitude of 2400–2600m to be in-creased by 690 m to between 3090 m and 3290 m. The new estimatedelevations are lower than others (e.g. Williams et al., 2001; Rowleyand Currie, 2006; DeCelles et al., 2007) (Fig. 9).

equivalents of fossil pollen taxa recovered from the latest Oligocene–earliest Miocene sed-

Fig. 9. Estimates of the altitude of the surface of the Tibetan Plateaudeduced frompreviousdata sources (modified Harris, 2006) and the correlation with the result of this study.

28 J. Sun et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 399 (2014) 21–30

4. Discussion

The new pollen results from the Lunpola Basin, after Cenozoic globalcooling corrections, indicate that the paleoelevation of central Tibet wasaveraged at 3190± 100m (between 3090m and 3290m) asl in the lat-est Oligocene–earliest Miocene. This elevation is the maximum heightfor the Lunpola Basin. Firstly, the sedimentary type of the Dingqing For-mation is lacustrine sediment, the preserved pollen assemblages withinthe strata weremainly transported from the lake catchment transportedby river and/or wind, and the pollen assemblages preserved in theLunpola Basin mainly reflect the vegetation types of the nearby landsand the neighboring mountains within the catchment (Fig. 10). In thepresent, the largest river within the Lunpola Basin is the Zagya ZangbuRiver which originates from the Tanggula Mountains with a peak eleva-tion of 6621 m asl. In this sense, the paleoelevation of 3190 ± 100 m isthe maximum elevation of the Lunpola Basin in the latest Oligocene–earliest Miocene. This result is different from the previous views thatthe Tibetan Plateau reached the present elevation as early as the lateEocene or middle Oligocene (e.g. Rowley and Currie, 2006; DeCelleset al., 2007; Wang et al., 2008; Xu et al., 2013). Secondly, althoughpollen may be transported long distances by seasonal cyclonic winds(Cour et al., 1999; Rousseau et al., 2003, 2006, 2008), and the upslope

Fig. 10. Topography of the Lunpola Basin

southwesterly Indian monsoon winds might transport some lower ele-vation pollen grains to the Lunpola Basin (e.g. Song et al., 2010), suchamounts of the remote transportation cannot be significant comparingwith the local pollen sources.

Different from the stable oxygen isotope paleoaltimetry ofauthigenic andweathering pronemineral of carbonate, themost advan-tages of the CoA pollen altitude are: (1) a striking structural feature ofpollen grain is the tough outer coat which is resistant to decay (Brooksand Shaw, 1978), the well-preserved pollen in lake sediment is thekey to reveal past vegetation, landscape and environment; (2) whenthe estimated CoA pollen altitude was corrected by climatic changecan yield more convincible paleoelevation.

This elevation is comparable to some previous suggestions (Murphyet al., 1997; Song et al., 2010; Deng et al., 2012), but being at least 1500to 2000 m lower than the suggested elevations (reaching the presentaltitude) in the latest Eocene by using altitude estimation of carbonateoxygen isotopes in the same region (Rowley and Currie, 2006).

Actually, care must be taken when using quantitative stable isotopepaleoaltimetry reconstruction based on soil carbonates or lacustrinecarbonates.

Firstly, in order to provide independent estimates of paleoaltitude, asimplemodel that describes the systematic behavior of oxygen isotopesin precipitation as a function of elevation has been proposed by assum-ing that equilibrium isotope fractionation during Rayleigh distillation islinked to the thermodynamics of atmospheric ascent and water vaporcondensation (Rowley et al., 2001). However, this technique was criti-cized by Hou et al. (2003) on the grounds that Rayleigh distillationprovides an idealized model that assumes the immediate removal ofprecipitation, and do not account for the complexity of the physical pro-cesses involved in the formation of precipitation. Based on the observedδ18O versus elevation gradients, Hou et al. (2003) proposed that theused model by Rowley et al. (2001) does not necessarily validate inhigh elevation, and such model overestimates altitude. Moreover, thecommonly used δ18O values of soil carbonates or lacustrine carbonatesby this technique cannot be easily transferred to the oxygen isotopiccompositions of paleometeoric water (Wang et al., 2006) due to the un-certainty in the carbonate/water fractionation factor, additionally thelong-term digenesis and/or chemical weatheringmust have caused iso-tope fractionation of the authigenic carbonate.

Secondly, the equilibrium isotope fractionation versus elevation isbased on single water source. Nevertheless, the actual water sourcesof a specific site are quite complicated. In the Lunpola Basin, there are

and surrounding high mountains.

Fig. 11. 30-year (1971–2000) averaged water vapor transport vector fields at 500 hPa pressure level for winter (a, December to February) and summer (b, June to August). Vectorunits: kg m−1 s−1, the arrows indicate the direction of water vapor flux. The color background shows topographic information and the 3000 m contour line outlines the TibetanPlateau. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

29J. Sun et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 399 (2014) 21–30

at least two water sources: the southwest Indian summer monsoon;and the Atlantic water source transported by westerlies (Tian et al.,2001, 2007). According to the averaged 30-year (from 1971 to 2000)meteorological data of water vapor flux, during the winter season(from December to February), westerly winds dominate over most ofthe research area and bring little moisture (Fig. 11a); whereas duringthe summer season (June to August), both the southwest monsoon sys-tem and the mid-latitude westerlies bring moisture from the IndianOcean and the Atlantic to central Tibet (Fig. 11b). The isotope composi-tions of meteorite water of these two moisture sources after differentdistance transportations are variable (e.g. Craig and Gordon, 1965).Moreover, the different moisture sources and their seasonal variationsstrongly affect the stable isotopes in precipitation over Tibet (Tianet al., 2007). For example, modern observations indicate that in thesummer of the year 2000, δ18O values of precipitation decreased from3‰ to −22‰ in central Tibet in the middle of July (Tian et al., 2007).

Thirdly, just as argued by Wang et al. (2006) even if we assume thatδ18O values of soil carbonates record the δ18O of paleometeoric water,care must be taken when using the modern precipitation δ18O-elevationrelationships to the geologic past. Previous study demonstrates that thedeep-sea temperature has decreased 12 °C over the past 50 Ma (Learet al., 2000). On account of the fact that the sea water isotopic fraction istemperature-dependent, this will affect the oxygen isotope compositionsof rainwater originated from the ocean. In this context, uncertainties doexist in the isotopic lapse rate of precipitation and the temperate-dependent Rayleigh distillation. It is still not a straightforward way toapply the modern δ18O versus elevation relationship for paleoelevationreconstruction.

5. Conclusions

Based on the high resolution magnetostratigraphy, together withthe biostratigraphic evidence and bentonite layer dating, the studiedmiddle to lower part of the Dingqing Formation in the Lunpola Basin,central Tibet, is attributed a new age range from the latest Oligoceneto the earliest Miocene (25.5–19.8 Ma). This provides the basis forpaleoenvironment and paleoelevation reconstruction.

Pollen results indicate that the vegetation types during the latestOligocene–earliest Miocene was dominated by forestry, characterizedbymixed coniferous–broadleaved forests indicating awarm–temperateclimate, being quite different from the present steppe vegetation in cen-tral Tibet.

By using the Co-existing Approach, the new pollen results from theLunpola Basin, after accounting for Cenozoic global cooling, suggestthat the paleoelevation of this basin had a maximum altitude of3190 ± 100 m asl in the latest Oligocene–earliest Miocene.

Acknowledgments

This work was supported by the “Strategic Priority ResearchProgram” of the Chinese Academy of Sciences (XDB03020500), theNational Basic Research Program of China (2010CB833400) and theNational Nature Science Foundation of China (grants 41290251 and41272203). We thank Lin Ding, Tao Deng for helps in the field expedi-tion and Bihong Fu and Yingying Jia for assistance in drawing the digitalelevation map. We also thank the editor Finn Surlyk and the two anon-ymous reviewers for their helpful comments and careful reviews.

References

Besse, J., Courtillot, V., Pozzi, J.P., Westphal, M., Zhou, Y.X., 1984. Palaeomagnetic estimatesof crustal shortening in the Himalayan thrusts and Zangbo suture. Nature 311,621–626.

Brooks, J., Shaw, G., 1978. Sporopollenin: a review of its chemistry, palaeochemistry andgeochemistry. Grana 17, 91–97.

Bureau of Geology andMineral Resources Xizang Autonomous Region, 1993. Regional Ge-ology of Xizang (Tibet) Autonomous Region. Geological Publishing House, Beijing.

Chung, S.L., Lo, C.H., Lee, T.Y., Zhang, Y., Xie, Y., Li, X., Wang, K.L., Wang, P.L., 1998.Diachronous uplift of the Tibetan Plateau starting 40 Myr ago. Nature 394, 769–773.

Coleman, M., Hodges, K., 1995. Evidence for Tibetan Plateau uplift before 14Myr ago froma new minimum estimate for east–west extension. Nature 374, 49–52.

Cour, P., Zhang, Z., Duzer, D., Calleja, M., Yao, Z., 1999. Vegetational and climatic signifi-cance of modern pollen rain in northwestern Tibet. Rev. Palaeobot. Palynol. 104,183–204.

Craig, H., Gordon, L.I., 1965. Deuterium and oxygen 18 variations in the ocean and thema-rine atmosphere. In: Tongiorgi, E. (Ed.), Stable Isotopes in Oceanographic Studies andPaleotemperatures. Laboratorio di Geologia Nucleare, Pisa, Italy, pp. 9–130.

Cyr, A., Currie, B.S., Rowley, D.B., 2005. Geochemical and stable isotopic evaluation ofFenghuoshan Group lacustrine carbonates, north-central Tibet: implications for thepaleoaltimetry of Late Eocene Tibetan Plateau. J. Geol. 113, 517–533.

DeCelles, P.G., Quade, J., Kapp, P., Fan, M.J., Dettman, D.L., Ding, L., 2007. High and dry incentral Tibet during the Late Oligocene. Earth Planet. Sci. Lett. 253, 389–401.

Deng, T.,Wang, S.Q., Xie, G.P., Li, Q., Hou, S.K., Sun, B.Y., 2012. Amammalian fossil from theDingqing Formation in the Lunpola Basin, northern Tibet, and its relevance to age andpaleo-altimetry. Chin. Sci. Bull. 57, 261–269.

Dupont-Nivet, G., Hoorn, C., Konert, M., 2008. Tibetan uplift prior to the Eocene–Oligoceneclimate transition: evidence from pollen analysis of the Xining Basin. Geology 36,987–990.

Faegri, K., Iversen, J., 1989. Textbook of Pollen Analysis. John Wiley & Sons, New York.Fielding, E.J., 1996. Tibet uplift and erosion. Tectonophysics 260, 55–84.Fisher, R.A., 1953. Dispersion on a sphere. Proc. R. Soc. A 217, 295–305.

30 J. Sun et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 399 (2014) 21–30

Garzione, C.N., Dettman, D.L., Quade, J., DeCelles, P.G., Butler, R.F., 2000a. High times onthe Tibetan plateau: paleoelevation of the Thakkhola graben, Nepal. Geology 28,339–342.

Garzione, C.N., Quade, J., DeCelles, P.G., English, N.B., 2000b. Predicting paleoelevation ofTibet and the Himalaya from δ18O versus altitude gradients in meteoric water acrossthe Nepal Himalaya. Earth Planet. Sci. Lett. 183, 215–229.

Gradstein, F., Ogg, J., Smith, A., 2004. A Geologic Time Scale 2004. Cambridge UniversityPress, Cambridge.

Grimm, E.C., 1987. CONISS: a FORTRAN 77 program for stratigraphically constrained clus-ter analysis by themethod of incremental sum of squares. Comput. Geosci. 13, 13–35.

Grimm, G.W., Denk, T., 2012. Reliability and resolution of the coexistence approach — arevalidation using modern-day data. Rev. Palaeobot. Palynol. 172, 33–47.

Hao, H., Ferguson, D.K., Chang, H., Li, C.S., 2012. Vegetation and climate of the Lop Nurarea, China, during the past 7 million years. Clim. Chang. 113, 323–338.

Harris, N., 2006. The elevation history of the Tibetan Plateau and its implications for theAsian monsoon. Palaeogeogr. Palaeoclimatol. Palaeoecol. 241, 4–15.

Harrison, T.M., Copeland, P., Kidd, W.S.F., Yin, A., 1992. Raising Tibet. Science 255,1663–1670.

Harrison, T.M., Copeland, P., Kidd, W.S.F., Lovera, O.M., 1995. Activation of theNyainquentanghla Shear Zone: implications for uplift of the southern Tibet plateau.Tectonics 14, 658–676.

He, H.Y., Sun, J.M., Li, Q.L., Zhu, R.X., 2012. New age determination of the Cenozoic LunpolaBasin, central Tibet. Geol. Mag. 149, 141–145.

Hoorn, C., Ohja, T., Quade, J., 2000. Palynological evidence for vegetation development andclimatic change in the Sub-Himalayan Zone (Neogene, Central Nepal). Palaeogeogr.Palaeoclimatol. Palaeoecol. 163, 133–161.

Hou, S.G., Masson-Delmotte, V., Qin, D., Jouzel, J., 2003. Modern precipitation stable iso-tope vs. elevation gradients in the High Himalaya, Comment on “A new approachto stable isotope based paleoaltimetry: implications for paleoaltimetry andpaleohypsometry of the High Himalaya since the Late Miocene”. Earth Planet. Sci.Lett. 209, 395–399.

Ivanov, D., Ashraf, A.R., Mosbrugger, V., Palamarev, E., 2002. Palynological evidence forMiocene climate change in the Forecarpathian Basin (Central Paratethys, NWBulgaria). Palaeogeogr. Palaeoclimatol. Palaeoecol. 178, 19–37.

Ivanov, D.A., Ashraf, A.R., Mosbrugger, V., 2007. Late Oligocene and Miocene climate andvegetation in the Eastern Paratethys area (northeast Bulgaria), based on pollendata. Palaeogeogr. Palaeoclimatol. Palaeoecol. 255, 342–360.

Jacques, F.M.B., Guo, S.X., Su, T., Xing, Y.W., Huang, Y.J., Liu, Y.S., Ferguson, D.K., Zhou, Z.K.,2011. Quantitative reconstruction of the Late Miocene monsoon climates of south-west China: a case study of the Lincang flora from Yunnan Province. Palaeogeogr.Palaeoclimatol. Palaeoecol. 304, 318–327.

Kirschvink, J., 1980. The least square line and the analysis of paleomagnetic data. Geophys.J. R. Astron. Soc. 62, 699–718.

Lear, C.H., Elderfield, H., Wilson, P.A., 2000. Cenozoic deep-sea temperatures and globalice volumes from Mg/Ca in benthic foraminiferal calcite. Science 287, 269–272.

Liang, M.M., Bruch, A., Collinson, M., Mosbrugger, V., Li, C.S., Sun, Q.G., Hilton, J., 2003.Testing the climatic estimates from different palaeobotanical methods: an examplefrom the Middle Miocene Shanwang flora of China. Palaeogeogr. Palaeoclimatol.Palaeoecol. 198, 279–301.

Lu, H., Wu, N., Yang, X., Shen, C., Zhu, L., Wang, L., Li, Q., Xu, D., Tong, G., Sun, X., 2007. Spa-tial pattern of Abies and Picea surface pollen distribution along the elevation gradientin the Qinghai–Tibetan Plateau and Xinjiang, China. Boreas 37, 254–262.

Meyer, H.W., 2007. A review of paleotemperature–lapse rate methods for estimatingpaleoelevation from fossil floras. Rev. Mineral. Geochem. 66, 155–171.

Miao, Y.F., Fang, X.M., Wu, F.L., Cai, M.T., Song, C.H., Meng, Q.Q., Xu, L., 2013a. LateCenozoic continuous aridification in the western Qaidam Basin: evidence fromsporopollen records. Clim. Past 9, 1485–1508.

Miao, Y.F., Wu, F.L., Herrmann, M., Yan, X.L., Meng, Q.Q., 2013b. Late early Oligocene EastAsian summer monsoon in the NE Tibetan Plateau: evidence from a palynological re-cord from the Lanzhou Basin, China. J. Asian Earth Sci. 75, 46–57.

Miller, K.G., Fairbanks, R.G., Mountain, G.S., 1987. Tertiary oxygen isotope synthesis, sealevel history, and continental margin erosion. Paleoceanography 2, 1–19.

Molnar, P., Tapponnier, P., 1975. Cenozoic tectonics of Asia: effects of a continental colli-sion. Science 189, 419–426.

Mosbrugger, V., Utescher, T., 1997. The coexistence approach—a method for quantitativereconstructions of Tertiary terrestrial palaeoclimate data using plant fossils.Palaeogeogr. Palaeoclimatol. Palaeoecol. 134, 61–86.

Murphy, M.A., Yin, A., Harrison, T.M., Dürr, S.B., Chen, Z., Ryerson, F.J., Kidd, W.S.F., Wang,X., Zhou, X., 1997. Did the Indo-Asian collision alone create the Tibetan Plateau?Geology 25, 719–722.

Patrait, P., Achache, J., 1984. India–Eurasia collision chronology has implications for crust-al shortening and driving mechanism of plates. Nature 311, 615–621.

Rousseau, D.D., Duzer, D., Cambon, G., Jolly, D., Poulsen, U., Ferrier, J., Schevin, P., Gros, R.,2003. Long distance transport of pollen to Greenland. Geophys. Res. Lett. 30, 1765.http://dx.doi.org/10.1029/2003GL017539.

Rousseau, D.D., Schevin, P., Duzer, D., Cambon, G., Ferrier, J., Jolly, D., Poulsen, U., 2006.New evidence of long distance pollen transport to southern Greenland in Late Spring.Rev. Palaeobot. Palynol. 141, 277–286.

Rousseau, D.D., Schevin, P., Ferrier, J., Jolly, D., Andreasen, T., Ascanius, S.E., Hendriksen, S.,Poulsen, U., 2008. Long-distance pollen transport from North America to Greenlandin spring. J. Geophys. Res. 113, G02013. http://dx.doi.org/10.1029/2007JG000456.

Rowley, D.B., Currie, B.S., 2006. Palaeo-altimetry of the late Eocene to Miocene Lunpolabasin, central Tibet. Nature 439, 677–681.

Rowley, D.B., Garzione, C.N., 2007. Stable isotope-based paleoaltimetry. Annu. Rev. EarthPlanet. Sci. 35, 463–508.

Rowley, D.B., Pierrehumbert, R.T., Currie, B.S., 2001. A new approach to stable isotope-based paleoaltimetry: implications for paleoaltimetry and paleohypsometry of theHigh Himalaya since the Late Miocene. Earth Planet. Sci. Lett. 188, 253–268.

Song, X.Y., Spicer, R.A., Yang, J., Yao, Y.F., Li, C.S., 2010. Pollen evidence for an Eocene toMiocene elevation of central southern Tibet predating the rise of the High Himalaya.Palaeogeogr. Palaeoclimatol. Palaeoecol. 297, 159–168.

Spicer, R.A., Harris, N.B.W.,Widdowson,M., Herman, A.B., Guo, S.X., Valdes, P.J., Wolfe, J.A.,Kelley, S.P., 2003. Constant elevation of southern Tibet over the past 15 million years.Nature 421, 622–624.

Suc, J.P., Drivaliari, A., 1991. Transport of bisaccate coniferous fossil pollen grains to coastalsediments: an example from the earliest Pliocene Orb Ria (Languedoc, SouthernFrance). Rev. Palaeobot. Palynol. 70, 247–253.

Sun, X.J., Wang, P.X., 2005. How old is the Asia monsoon system? Palaeobotanical recordsfrom China. Palaeogeogr. Palaeoclimatol. Palaeoecol. 222, 181–222.

Tapponnier, P., Xu, Z.Q., Roger, F., Meyer, B., Arnaud, N., Wittlinger, G., Yang, J.S., 2001.Oblique stepwise rise and growth of the Tibet Plateau. Science 294, 1671–1677.

Tian, L.D., Masson-Delmotte, V., Stievenard, M., Yao, T.D., Jouzel, J., 2001. Tibetan Plateausummer monsoon northward extent revealed by measurements of water stable iso-topes. J. Geophys. Res. 106, 28081–28088.

Tian, L.D., Yao, T.D., MacClune, K., White, J.W.C., Schilla, A., Vaughn, B., Vachon, R.,Ichiyanagi, K., 2007. Stable isotopic variations in west China: a consideration of mois-ture sources. J. Geophys. Res. 112, D10112. http://dx.doi.org/10.1029/2006JD007718.

Utescher, T., Ivanov, D., Harzhauser, M., Bozukov, V., Ashraf, A.R., Rolf, C., Urbat, M.,Mosbrugger, V., 2009. Cyclic climate and vegetation change in the late Miocene ofWestern Bulgaria. Palaeogeogr. Palaeoclimatol. Palaeoecol. 272, 99–114.

Wang, Y.S., 2012. Regional Geological Survey of the People's Republic of China (AngdarCo, I45C004004). China University of Geosciences Press, Beijing.

Wang, K.F., Yang, J.W., Li, Z., Li, Z.R., 1975. On the tertiary sporo-pollen assemblages fromLunpola Basin of Xizang, China and their palaeogeographic significance. Sci. Geol. Sin.4, 366–374.

Wang, F.X., Qian, N.F., Zhang, Y.L., Yang, H.Q., 1997. Pollen Flora of China. Science Press,Beijing.

Wang, Y., Deng, T., Biasatti, D., 2006. Ancient diets indicate significant uplift of southernTibet after ca. 7 Ma. Geology 34, 309–312.

Wang, C.S., Zhao, X.X., Liu, Z.F., Lippert, P.C., Graham, S.A., Coe, R.S., Yi, H.S., Zhu, L.D., Liu, S.,Li, Y.L., 2008. Constraints on the early uplift history of the Tibetan Plateau. Proc. Natl.Acad. Sci. U. S. A. 105, 4987–4992.

Wang, Y., Xu, Y.F., Khawajia, S., Passey, B.H., Zhang, C.F., Wang, X.M., Li, Q., Tseng, Z.J.,Takeuchi, G.T., Deng, T., Xie, G.P., 2013. Diet and environment of a mid-Pliocenefauna from southwestern Himalaya: paleo-elevation implications. Earth Planet. Sci.Lett. 376, 43–53.

Williams, H.M., Turner, S., Kelley, S., Harris, N., 2001. Age and composition of dikes inSouthern Tibet: new constraints on the timing of east–west extension and its rela-tionship to postcollisional volcanism. Geology 29, 339–342.

Wu, C.Y., 1983. Flora Xizangica, vol. One. Science Press, Beijing.Wu, C.Y., 1985. Flora Xizangica, vol. Two. Science Press, Beijing.Wu, C.Y., 1986a. Flora Xizangica, vol. Three. Science Press, Beijing.Wu, C.Y., 1986b. Flora Xizangica, vol. Four. Science Press, Beijing.Wu, C.Y., 1987. Flora Xizangica, vol. Five. Science Press, Beijing.Xia, W.G., 1982. Ostracode fauna from Lunpola Group in Xizang (Tibet) and its geological

age. In: Chinese Geology Bureau Tibetan Plateau Proceeding Editorial Committee(Ed.), Contribution to the Geology of the Qinghai–Xizang (Tibet) Plateau (10). Geo-logical Publishing House, Beijing, pp. 149–159.

Xia, J.B., 1983. Cenozoic of Baingoin and its borders, Xizang (Tibet). In: Chinese GeologyBureau Tibetan Plateau Proceeding Editorial Committee (Ed.), Contribution to the Ge-ology of the Qinghai–Xizang (Tibet) Plateau (3). Geological Publishing House, Beijing,pp. 243–254.

Xu, Z.Y., 1980. The Tertiary and its petroleum potential in the Lunpola Basin, Tibet. Oil GasGeol. 1, 153–158.

Xu, R., Tao, J.R., Sun, X.J., 1973. On the discovery of a Quercus semicarpifolia bed in MountShisha Pangma and its significance in botany and geology. J. Integr. Plant Biol. 15,103–119.

Xu, Q., Ding, L., Zhang, L.Y., Cai, F.L., Lai, Q.Z., Yang, D., Liu-Zeng, J., 2013. Paleogene highelevations in the Qiangtang Terrane, central Tibetan Plateau. Earth Planet. Sci. Lett.362, 31–42.

Yin, A., Harrison, T.M., 2000. Geologic evolution of the Himalayan–Tibetan orogen. Annu.Rev. Earth Planet. Sci. 28, 211–280.

Zhao, W.L., Morgan, W.J., 1985. Uplift of Tibetan Plateau. Tectonics 4, 359–369.