Assembly of the Pamirs: Age and origin of magmatic belts from the southern Tien Shan to the southern Pamirs and their relation to Tibet

Assembly of the Pamirs: Age and origin of magmatic belts

from the southern Tien Shan to the southern Pamirs

and their relation to Tibet

Martina Schwab,1,2 Lothar Ratschbacher,2 Wolfgang Siebel,1 Michael McWilliams,3

Vladislav Minaev,4 Valery Lutkov,4 Fokun Chen,1 Klaus Stanek,2 Bruce Nelson,5

Wolfgang Frisch,1 and Joseph L. Wooden6

Received 14 September 2003; revised 3 March 2004; accepted 27 April 2004; published 2 July 2004.

[1] Magmatic rocks and depositional setting ofassociated volcaniclastic strata along a north-southtraverse spanning the southern Tien Shan and easternPamirs of Kyrgyzstan and Tajikistan constrain thetectonics of the Pamirs and Tibet. The northern Pamirsand northwestern Tibet contain the north facing Kunlunsuture, the south facing Jinsha suture, and theintervening Carboniferous to Triassic Karakul–Mazarsubduction accretion system; the latter is correlatedwith the Songpan-Garze–Hoh Xi system of Tibet. TheKunlun arc is a composite early Paleozoic to latePaleozoic-Triassic arc. Arc formation in the Pamirs ischaracterized by �370–320 Ma volcanism thatprobably continued until the Triassic. The crypticTanymas suture of the southern northern Pamirs is partof the Jinsha suture. A massive ��227 Ma batholithstitches the Karakul–Mazar complex in the Pamirs.There are striking similarities between the Qiangtangblock in the Pamirs and Tibet. Like Tibet, the regionalstructure of the Pamirs is an anticlinorium that includesthe Muskol and Sares domes. Like Tibet, themetamorphic rocks in these domes are equivalentsto the Karakul –Mazar–Songpan-Garze system.Granitoids intruding the Qiangtang block yield�200–230 Ma ages in the Pamirs and in centralTibet. The stratigraphy of the eastern Pshart area in thePamirs is similar to the Bangong-Nujiang suture zonein the Amdo region of eastern central Tibet, but aTriassic ocean basin sequence is preserved in thePamirs. Arc-type granitoids that intruded into the

eastern Pshart oceanic-basin–arc sequence (�190–160 Ma) and granitoids that cut the southern Qiangtangblock (�170–160 Ma) constitute the Rushan-Pshartarc. Cretaceous plutons that intruded the central andsouthern Pamirs record a long-lasting magmatichistory. Their zircons and those from late Miocenexenoliths show that the most distinct magmatic eventswere Cambro-Ordovician (�410–575 Ma), Triassic(�210–250 Ma; likely due to subduction along theJinsha suture), Middle Jurassic (�147–195 Ma;subduction along Rushan-Pshart suture), and mainlyCretaceous. Middle and Late Cretaceous magmatismmay reflect arc activity in Asia prior to the accretion ofthe Karakoram block and flat-slab subduction along theShyok suture north of the Kohistan-Ladakh arc,respectively. Before India and Asia collided, thePamir region from the Indus-Yarlung to the Jinshasuture was an Andean-style plate margin. Our analysissuggests a relatively simple crustal structure for thePamirs and Tibet. From the Kunlun arc in the northto the southern Qiangtang block in the souththe Pamirs and Tibet likely have a dominantlysedimentary crust, characterized by Karakul–Mazar–Songpan-Garze accretionary wedge rocks. The crustsouth of the southern Qiangtang block is likelyof granodioritic composition, reflecting long-livedsubduction, arc formation, and Cretaceous-Cenozoicunderthrusting. INDEX TERMS: 1020 Geochemistry:

Composition of the crust; 1035 Geochemistry: Geochronology;

8102 Tectonophysics: Continental contractional orogenic belts;

8110 Tectonophysics: Continental tectonics—general (0905);

KEYWORDS: Pamirs, Tibet, continental tectonics, magmatism,

geochronology, geochemistry. Citation: Schwab, M., et al.

(2004), Assembly of the Pamirs: Age and origin of magmatic

belts from the southern Tien Shan to the southern Pamirs and their

relation to Tibet, Tectonics, 23, TC4002, doi:10.1029/

2003TC001583.

1. Introduction

[2] Like much of central Asia, the crust of the Pamirs(Figure 1b) amalgamated over several orogenic cyclesduring Paleozoic and Mesozoic times [e.g., Yin andHarrison, 2000]. While first-order attempts to correlate

TECTONICS, VOL. 23, TC4002, doi:10.1029/2003TC001583, 2004

1Institut fur Geowissenschaften, Universitat Tubingen, Tubingen,Germany.

2Institut fur Geowissenschaften, Technische Universitat BergakademieFreiberg, Freiberg, Germany.

3Geological and Environmental Sciences, Stanford University, Stanford,California, USA.

4Geological Institute of the Tajik Academy of Science, Dushanbe,Tajikistan.

5Department of Geological Sciences, University of Washington, Seattle,Washington, USA.

6U.S. Geological Survey, Menlo Park, California, USA.

Copyright 2004 by the American Geophysical Union.0278-7407/04/2003TC001583$12.00

TC4002 1 of 31

sutures in the Pamirs and the Tien Shan with those in Tibetand Afghanistan have been made [e.g., Burtman andMolnar, 1993], mapping of ophiolites, magmatic arcs, andrelated rocks has not yet illuminated the sutures unequivo-cally, leading to variable interpretations of the westwardcontinuation of the Paleozoic and Mesozoic sutures fromTibet into the Pamirs (Figure 1b) [e.g., Burtman and Molnar,1993; Yin and Harrison, 2000]. Important questions remainunanswered. For example, is the Bangong-Nujiang sutureequivalent to the Shyok suture, or does it continue into theRushan-Pshart zone, a suture assigned Mesozoic age byBurtman and Molnar [1993]? How are the latest Paleozoic/early Mesozoic Jinsha and Kunlun sutures manifested in thePamirs? Does the Paleozoic suture within the northernPamirs correspond to the early Paleozoic Kudi or to the latePaleozoic Kunlun suture along the northern rim of westernTibet?[3] This paper documents the evolution of magmatic

rocks along a north-south traverse across the southernmostTien Shan and eastern Pamirs of Kyrgyzstan and Tajiki-stan. Our study was guided by two concepts. The first isthat magmatic arcs and postcollisional magmatic rocksdefine the nature and limits of continental margins andlinear collisional orogens [e.g., Sengor et al., 1993].

Second, field and laboratory studies in the Pamirs andTibet highlight a history of accretion, collision, and igne-ous activity and allow us to constrain better the geo-dynamic processes common to the Pamirs and Tibet. Webegin by redefining the age zonation of magmatic rocks(principally plutonic rocks) in the Pamirs and Tibet on thebasis of published data from Tibet and our new geochro-nology from the Pamirs. We next characterize the mag-matic belts geochemically, evaluating their first-ordertectonic setting. We then use existing stratigraphic records,supplemented by our field observations, to derive tectonicfacies interpretations for (volcano-) sedimentary depositsassociated with the magmatic belts. Finally, we discuss theamalgamation history of the Pamirs and construct correla-tions with Tibet. We argue that the northern Pamirs containthe late Paleozoic to early Mesozoic opposite facingKunlun and Jinsha sutures and related arcs and an inter-vening subduction accretion system (the Karakul-Mazarcomplex equivalent to the Songpan-Garze–Hoh Xi sys-tems in Tibet), stitched together by a voluminous postcolli-sional batholith. This subduction accretion system emergesin the central Pamirs as the core of the Qiangtang block,equivalent to the Gongma–Shuang Hu antiforms in centralTibet. The Bangong-Nujiang suture of Tibet is represented

Figure 1. Regional tectonic map of the Pamirs and Tibet, showing (a) major (micro)continents and(b) major suture zones and Cenozoic fault systems. Modified after the compilations of Burtman andMolnar [1993] for the Pamirs and Yin and Harrison [2000] for Tibet. Figure 1a shows localities foreastern Tibetan reference samples F218 (Jinsha suture) and TR257 (Songpan-Garze basement); for detailssee Table 1.

TC4002 SCHWAB ET AL.: ASSEMBLY OF THE PAMIRS AND TIBET

2 of 31

TC4002

by the Rushan-Pshart zone in the southern central Pamirs.The southern Pamirs comprise the superposed magmaticarcs related to the (?)Tirich Mir, Shyok, and Indus-Yarlungsutures.

2. Magmatic Units and Related Sutures of

Western Tibet and the Pamirs

[4] The distribution of magmatic rocks in Tibet and thePamirs is closely linked to the tectonic evolution of centralAsia and guides the trace of successively accreted conti-nental fragments and magmatic arcs. In the Pamirs, mag-matic belts are bent northward and/or are offset along majorstrike-slip faults as a result of the India-Asia collision(Figure 1). Figure 2 shows the age and location of magmaticrocks in central and western Tibet and the Pamirs based onreliable and well-located zircon and monazite U/Pb ages(Table A1, in auxiliary material1, data from Tibet). Figure 3presents our new geochronology from the Pamirs and Tibet(Table 1 for sample description and location) and relates itto the published ages from western Tibet. On the basis ofthe magmatic zonation and suture locations suggested inFigures 2 and 3, we next discuss our analytical data,followed by a discussion of the blocks and sutures of thePamirs and their geodynamic evolution.

2.1. Geochronology

2.1.1. Analytical Methods[5] We conducted conventional isotope dilution analyses

(ID-TIMS) of 50 zircon fractions from 15 samples atTubingen and Seattle. At Stanford, ion microprobe datingwas used to investigate five samples with complex conven-tional U/Pb systematics and multifaceted zircon internalstructures. The U/Pb geochronological data of these samplesare listed in Table A2 of auxiliary material. Pb evaporationages were measured from two zircons at Tubingen. Figure 4summarizes the conventional U/Pb isotopic data on con-cordia diagrams, the evaporation method data in histograms,and the sensitive high-resolution ion microprobe (SHRIMP)results in Terra-Wasserburg concordia plots and weightedmean diagrams for selected age groups. We used 15 kVcathodoluminenscence (CL) to image the internal structureof the zircons prior to analysis and focused on composi-tional variations between core and rim, the nature of crystalgrowth zones, and irregular boundaries that may recordinterrupted zircon growth. In addition, we employed zircontypology classification (Pupin diagrams [Pupin, 1980]) toqualitatively assess the alkalinity of the melts of the Pamiranmagmatic belts. For conventional U-Pb analyses, zirconfractions consisting of a few morphological identical grainswere washed shortly in 6N HCl and hot 7N HNO3 prior todissolution to remove surface contamination; analytical anddata processing details are similar to Chen et al. [2000].Zircon evaporation ages were determined using the proce-dures of Kober [1986, 1987] and Kroner and Hegner

[1998]. Ages were calculated only from measurements with206Pb/204Pb ratios higher than 5000. Common Pb correctionwas performed according to Cocherie et al. [1992] using theStacey and Kramers [1975] two-stage Pb evolution model.Age uncertainties are reported as 2s from the mean207Pb/206Pb ages. The Stanford-USGS SHRIMP-RG meth-odology is that of DeGraaff-Surpless et al. [2002]; ages arebased on 206Pb/238U ratios referenced to zircon standardAS57 (1099 ± 1 Ma [Paces and Miller, 1993]). Uncertain-ties are quoted at 2s, and the age precision is typicallywithin 2–3%.[6] Table 2 summarizes the results of 40Ar/39Ar, Rb/Sr,

and K/Ar dating of hornblende and mica. Table A3(in auxiliary material) lists the 40Ar/39Ar results from sizefractions ranging from 63 to 355 mm. Analytical proceduresare similar to those ofHacker et al. [1996] andGrimmer et al.[2003]; gas was released by furnace step-heating and single-grain laser fusion in one case (muscovite sample 96P4e). Ageuncertainties are quoted at ±1s. One K/Ar sericite age wasobtained from a <2 mm clay fraction at the Russian Academyof Sciences, Saint Petersburg. Radiogenic 40Arwasmeasuredby isotope dilution and potassium was determined in dupli-cate by flame photometry. Rb/Sr mineral ages were measuredby isotope dilution at Tubingen using whole rock and whitemica following the analytical procedures of Hegner et al.[1995]. Figure 5 shows the 40Ar/39Ar age spectra. Isotopecorrelation diagrams are presented for those samples thatappear to contain excess 40Ar, as indicated by systematicallydiscordant (e.g., saddle-shaped) age spectra and/or nonatmo-spheric 40Ar/36Ar intercepts. Rb/Sr whole rock–white mica‘‘isochrons’’ are plotted in Figure 5 adjacent to the 40Ar/39Arspectra of the same samples for comparison.2.1.2. Results2.1.2.1. Garm––Turkestan-Alai Zone, SouthernTien Shan[7] The Garm–Turkestan-Alai zone (Figures 3 and 6a:

Garm–Turkestan-Alai sections) is interpreted as Proterozoiccontinental basement with early to late Paleozoic passivemargin and accretionary complex rocks south of the SouthFergana suture and north of the South Gissar suture; thezone is intruded by granodiorite and biotite/K-feldspargranite of Early Permian, and locally probably also LateDevonian-(?)Carboniferous age [Vlasov et al., 1991]. Wedated two of these rocks by 40Ar/39Ar (samples TS12b andTS20a, Figure 5) and obtained identical 277 ± 2 Ma biotiteages, even though the sampling localities are separated bymore than 100 km. The Early Permian plutons intrudemostly Devonian and Carboniferous unmetamorphosed tolow-grade cover, suggesting shallow emplacement and littlepost-Paleozoic denudation.2.1.2.2. Kunlun Arc, Northern Pamirs[8] In the Trans-Alai range of the northernmost Pamirs,

we collected mafic to acid, deformed, low-grade volcanicrocks that are intruded locally by plagiogranite (Altyn-Daravalley, Figure 3). These are part of a Carboniferous toTriassic igneous-sedimentary sequence deposited in anoceanic basin/arc environment (Figure 6b, southern Altyn-Dara and Altyn-Dara sections), and have been assignedEarly Carboniferous to Permian ages; the sequence may rest

1Auxiliary material is available at ftp://ftp.agu.org/apend/tc/2003TC001583.


3 of 31

TC4002

Figure

2


4 of 31

TC4002

locally on Proterozoic basement (Figure 6b, Kurgovatsection) [e.g., Vlasov et al., 1991; Ruzhentsev andShol’man, 1982]. It was later thrust over Cretaceous-Tertiary rocks in the Trans-Alai fold and thrust belt.Metarhyolite AD2a yielded three zircon fractions withelongate magmatic habit and long (fraction F1) and smallprisms (F3 and F5). The zircons are <200 mm in size, oftenbroken, with occasional rounded corners. CL zonationindicates strong compositional differences in most grains.Rare metamict to diffuse cores were observed. Thesezircons plot in the calc-alkaline and subalkaline fields ofthe Pupin classification. ID-TIMS ages from two fractionsare concordant at 329 ± 5 Ma, which we interpret as thecrystallization age; one fraction is weakly reversely discor-dant. We extracted small (80–150 mm) zircons from anandesite-dacite body (AD6c) along the northern edge of theCarboniferous metavolcanic unit. Data from transparentzircon fraction F2 are plotted together with those fromAD2a in Figure 4. We interpret the ages from AD2a andAD6c as indistinguishable. Hornblende from two differentmetaandesites yielded complicated 40Ar/39Ar apparent agespectra (Figure 5). AD2e contains excess 40Ar, but an age of350 ± 20 Ma can be calculated from the isotope correlationdiagram. AD6b yields a total gas age of �320 Ma and a�355 Ma age for the highest temperature step, similar toAD2e. Both samples exhibit Ar loss in the low temperaturepart of the spectrum, suggestive of a younger thermaldisturbance. Collectively, the U/Pb and 40Ar/39Ar agessuggest Early Carboniferous magmatism in the Kunlun arc.2.1.2.3. Karakul––Mazar Granitoid Belt,Northern Pamirs[9] The southern rim of the northern Pamirs in the

Karakul lake area is characterized by an east-west trendingbelt of mostly undeformed granodiorite to monzograniteplutons (Figure 3). These intrude Carboniferous to Permianand ?Triassic metaclastic and metavolcanic rocks, locallyaccompanied by intrusions of gabbro (Figure 6b, Sarykol-Karakul section) [Vlasov et al., 1991]. Three zircon frac-tions from sample P26, the northernmost biotite-graniteoutcrop of the belt, were analyzed by ID-TIMS (Figure 4).The results are neither concordant nor colinear. SHRIMPanalysis was performed on spots in 10 zircon grains. TheCL images of these zircons show multiple magmatic growthand resorption facets; inner grain portions are sometimesmetamict (Figure 7, P26 images). Convincing evidence forthe existence of cores has not been detected. The rims andcenters of several grains were analyzed, yielding indistin-guishable ages. Excluding one old grain (�402 Ma) and oneyoung grain (�189 Ma), the weighted mean of 14 spot agesis 227.0 ± 4.1 Ma (Figure 4). Granite sample P22 from thesouthern tip of Karakul Lake yielded two clear, prismaticfractions for ID-TIMS analysis: F1, 150–250 mm andcolorless, and F2, 80–200 mm and brownish in color. CL

images show a fine magmatic zoning with some zonedcores. F1 plots on the concordia, indicating a minimum ageof 215.0 ± 1.5 Ma. F2 lies close to, but above the concordiaat about 222 Ma (Figure 4). The Karakul Lake granitoids(four biotite granites and one two-mica granite) yieldedmuscovite and biotite 40Ar/39Ar cooling ages between 191and 207 Ma (Figure 5). The similar U/Pb and 40Ar/39Arages of these samples suggest that the granitoids of theKarakul area may be a large batholith emplaced and cooledto upper crustal temperatures in the Late Triassic. In contrastto the northern Pamiran Kunlun arc and the central Pamirs(see below), the Karakul-Mazar granitoid belt experiencedlittle Cenozoic tectonothermal reactivation. As no basementis known to the volcaniclastic Sarykol-Karakul rocks alongour studied section and their Songpan-Garze equivalentsfurther east (see below), we dated a basement granite of theSongpan-Garze system in the westernmost Longmen Shanof eastern Tibet (Figure 1b and Table 1); there biotite cooledat �752 Ma through �300�C (40Ar/39Ar age; Figure 5).2.1.2.4. Qiangtang Block, Central Pamirs[10] The central Pamirs contain deformed Paleozoic and

Triassic-Jurassic metasiliciclastic and carbonate platform-type rocks (Figure 6c, Akbai–northern Qiangtang andKalaktash–southern Qiangtang sections) that mantle highlydeformed, low- to high-grade, Barrovian metamorphicrocks [Vlasov et al., 1991]. They were interpreted toconstitute a Precambrian to Paleozoic continental fragment(Figure 6c, Sares-Muskol–central Qiangtang section) thatcollided with Asia, probably in the Permian [Burtman andMolnar, 1993]. This basement crops out in the Muskol andSares antiforms (Figure 3). It is intruded by small bodiesranging in composition from metagabbro to diorite.Its hanging contact is a normal fault zone, multiplyoverprinted by Cenozoic contractional and extensionalstructures reflecting high strain. U/Pb, Rb/Sr, and 40Ar/39Arages define an important Tertiary tectonothermal event(L. Ratschbacher et al., manuscript in preparation, 2004).Two zircon fractions were separated from an undeformedaplite dike (P17) that cuts high-grade, locally migmatiticbiotite-gneiss with early Miocene 40Ar/39Ar hornblendeand mica cooling ages, and Early Cretaceous red beds;the 40Ar/39Ar biotite age of P17 is also early Miocene(L. Ratschbacher et al., manuscript in preparation, 2004).Fraction D2 has 63 to 80 mm zircons, while those of DE2range from 80 to 125 mm. Applying a crystallization/recentPb loss model for these two fractions, the crystallization ageis 531 ± 30 Ma. Allowing Pb loss at 44 ± 22 Ma yields anupper intercept at 539 ± 5 Ma (Figure 4). SHRIMP analysisof 14 grains was done to search for possible Tertiary rims.The zircons are complex internally and rich in inclusions(Figure 7, P17 images). Spot ages range from 561 to424 Ma, but no Tertiary rims were detected; they may bemicron size and not detectable with the spot size used. The

Figure 2. Regional tectonic map, showing age and location of magmatic rocks in central and western Tibet and theeastern Hindu Kush, Karakoram, Kohistan, and Ladakh. Ages are mostly based on U/Pb geochronology on zircon andmonazite; see Table A1 in auxiliary material for data and references. No complete data coverage for the Kohistan arc isshown due to space problems. Suture-zone location interpretation as developed in this paper is based on this age data set,that of the Pamirs (Figure 3), and the geological evidence discussed in the text and Figure 6.


5 of 31

TC4002

Figure

3.

Ageandlocationofmagmaticrocksin

northwestern

TibetandthePam

irsbased

onU/Pb,40Ar/39Ar,andRb/Sr

geochronology.Datafrom

northwestern

Tibetarefrom

published

sources

(summarized

inTableA1in

auxiliary

material),

anddatafrom

thePam

irsarefrom

thisstudy(Table2andTablesA2andA3in

auxiliary

material;seeFigures4,5andtext

fordiscussion);

only

well-documentedandlocateddataareshown.FaultsareofCenozoic

ageandtogether

withthe

lithotectonicunitsandtheageassignmentofundated

magmaticrocksareinterpretedfrom

Vlasovetal.[1991],Bureauof

GeologyandMineralResources

ofXinjiangUygurAutonomousRegion[1992],Matteet

al.[1996],Chang[1996],and

ourownobservations.


6 of 31

TC4002

weighted mean of 14 spot ages, excluding a young�425 Maoutlier, is 536 ± 18 Ma (Figure 4). The zircons in this dikethus appear to be inherited and probably reflect the age ofthe pre-late Paleozoic Qiangtang basement. We use the fieldevidence of the dike cutting Early Cretaceous red beds (seeabove) to suggest that the possible �44 Ma Pb loss eventmay constrain the dyke emplacement as Tertiary.[11] Sample 96M9a is a metaleucogabbro that intrudes

biotite-gneiss of the eastern-most Muskol antiform and isrepresentative of the numerous small metagabbros found inthe eastern Muskol and Sares domes. It yielded a Tertiary40Ar/39Ar hornblende age (L. Ratschbacher et al., manu-script in preparation, 2004). Zircons from 96M9a are�200 mm and euhedral. Fraction F1 comprises elongate,brownish-pink zircons, F2 somewhat smaller, clear andpinkish zircons, F4 also pinkish, medium-sized zircons,and F5 the largest, yellow-brownish to pinkish, but shortprismatic zircons. Many zircons have opaque inclusions. CL

images show sector zoning along the prismatic long axesand large, well-rounded cores. The cores exhibit oscillatoryzoning and have embayments probably caused by resorp-tion of inclusions in the melt. Many grains are fractured butnot broken. All fractions yielded discordant U/Pb data.Fraction F2 has very low 206Pb/204Pb ratio (<100), andtherefore it was excluded from further calculations. Thediscordia through F1, F3 and F5 has a lower intercept at169 ± 7 Ma, which we interpret as the minimum age ofcrystallization. The upper intercept is at 828 ± 95 Ma(Figure 4). Pb evaporation on two large zircons withoutcores yielded indistinguishable ages of 231 ± 10 Ma(226 ratios, two step evaporation) and 228 ± 10 Ma(260 ratios, single step evaporation); the 206Pb/208Pb ratioevolution, likely a U/Th proxy, is typical for magmaticzircons in a fractionating granitic melt (Figure 4).[12] A series of biotite granites intrude Triassic-(?) Juras-

sic rocks, interpreted as Qiangtang cover (Figure 6c, Kalak-

Table 1. Sample Description and Location

Sample Stop Rock N Latitude E Longitude

Tien ShanTS1 TS1 weakly deformed granite 40�06.3160 73�31.6920

TS2 TS2 weakly deformed granite 40�08.3680 73�30.3150

TS12b TS12 granodiorite 39�46.080 73�36.720

TS18a TS18 diorite 39�32.1420 72�06.1900

TS20a TS20 diorite 39�33.770 72�04.640

PamirsAD1a AD1 meta-andesite, Altyn Dara valley 39�19.2290 72�15.7410

AD2a AD2 meta-rhyolite, Altyn Dara valley 39�12.5060 72�14.3840

AD2e AD2 hornblende-rich meta-volcanic rock, Altyn Dara valley 39�13.640 72�15.180

AD2f AD2 deformed meta-basalt 39�15.0440 72�15.5690

AD5a AD5 meta-andesite 39�16.80 72�160

M3 AD6 meta-andesite 39�17.20 72�16.050

AD6b AD6 hornblende-rich meta-andesite, Altyn Dara valley 39�17.680 72�16.080

AD6c AD6 meta-andesite, Altyn Dara valley 39�18.1770 72�16.1850

AD6d AD6 meta-basalt, Altyn Dara valley 39�18.180 72�16.190

AD6e AD6 meta-andesite 39�18.2880 72�16.0700

AD6f AD6 meta-dacite 39�18.5030 72�15.9600

AD7c AD7 meta-andesite 39�20.7530 72�18.3420

P2 P21 granodiorite, western Murgab valley 38�07.670 73�51.130

P5 P25 meta-granodiorite, western Murgab valley 38�10.20 73�34.10

P7 P32 white-mica bearing granite, eastern Murgab valley 38�11.750 74�15.330

P20 P57 two-mica granite, southwestern Karakul lake 38�48.10 73�17.030

P22 P59 granite, Karakul lake 38�55.930 73�22.870




P17 P46 aplite dike in high-grade biotitegneiss, Akbai valley 38�32.730 73�31.270

A96S1b A96-S-1 subvolcanic diorite, Akbai valley 38�31.50 73�42.20

96M9a 96-M-9 leucogabbro, south of Rankul lake 38�20.840 74�02.280

96M18a 96-M-18 Silurian phyllite, Rankul lake 38�17.10 74�04.50

L96M25a 96-M-25 biotitegranite, Pshart valley 38�12.480 74�01.7240

A96M18h A96-M-18 chlorite-muscovite-staurolite-garnet schist, Pshart valley 38�15.050 74�02.270

96P4e 96-P-4 granite pebble in Lower Miocene red beds, Pshart valley 38�16.460 73�44.530

L96A9 L96-A-9 biotite granite, Aksu valley 38�11.720 74�14.590

M96A7 M96-A-7 biotite granite intruding Jurassic sedimentary rocks, Aksu valley 38�09.140 74�24.5996A10b 96-A-10 granite, Aksu valley 38�08.100 74�28.110

Eastern Tibet (Jinsha Suture)F218 68 biotite-gneiss; exotic horse within serpentinite and Permian melange 29�58.40 99�04.70

Eastern Tibet (Longmen Shan–Songpan-Garze)TR257 87 granite 30�040 102�10.30


7 of 31

TC4002

Figure 4. U/Pb and Pb/Pb zircon data and age interpretation. See Table 1 for sample location and seeTable A2 in auxiliary material for data listings. ID-TIMS results are shown with 2s error ellipses anddiscordias are shown as lines. We show a variety of possible discordia fits, several of them clearlyinsignificant, to illustrate the interpretation margins; see text for discussion. The 206Pb/208Pb ratioevolution (96M9a) is likely a U/Th proxy and indicates a magmatic origin for these zircons. SHRIMPresults are plotted in Terra-Wasserburg concordia diagrams uncorrected for common Pb with 2s errorellipses. The errors in SHRIMP analyses are dominated by the uncertainty in the common Pb correctionproducing subvertical dispersion. Selected age groups are displayed as weighted average diagrams. Dataevaluation and plotting was supported by the program package of Ludwig [1999]. MSWD, mean squareweighted deviation.


8 of 31

TC4002

tash–southern Qiangtang section). We studied samples westof the Sares dome along the border to the southerly abuttingPshart zone (Figure 3), now a Cenozoic fault. Two-micamonzogranite sample M96A7 yielded zircons with large andsometimes nested cores that are either metamict or displaymagmatic zoning. The inner cores are well rounded, theouter cores are only abraded at the corners, and nearly allhave rims. ID-TIMS data from the five zircon fractions lackconcordance and colinearity. Connecting F2, F3, F4, and F5yields a lower intercept age at 131.1 ± 1.6 Ma (Figure 4).SHRIMP analysis of 24 grains was done to probe cores andrims (Figure 7, M96A7 images). We detected three agegroups in the cores. One spot yielded 2187 ± 13 Ma. Threespots in rounded metamict cores ranged from 232 to 250 Mawith one weakly zoned inner grain spot at 214 Ma thatprobably also belongs to this group. Four rounded, metamict(one weakly zoned) cores have 156 to 182 Ma ages. Themajority of spots probed spectacularly zoned idiomorphicgrains with ages ranging from 104 to 130 Ma (the weighted

mean of 17 spots is 113.0 ± 3.4 Ma), indicating a Cretaceouscrystallization age. Three spots of idiomorphic, well-zonedgrains and outermost idiomorphic rims of grains with coresyielded even younger ages ranging from 91 to 94 Ma. TheRb/Sr and 40Ar/39Ar muscovite ages of this granite are�104and �101 Ma, respectively, likely reflecting postemplace-ment cooling (Figure 5). Monzogranite sample 96A10balso intrudes probable Triassic rocks and yielded threezircon fractions, of which F1 and F2 plot on the concordia.Fraction F3 plots below the concordia suggesting Pb loss;this might also apply to F1. As no older component seemsto be inherited, we interpret the upper intercept as acrystallization age of 126 ± 50 Ma (Figure 4). SampleA96S1b, a small subvolcanic diorite of just north of theMuskol dome basement intrudes probable mid-Cretaceousred beds [Vlasov et al., 1991]. All six fractions containsmall zircons (�63–80 mm); half are long prismatic,transparent and without inclusions, the other half showstransparent to yellowish tabular grains. Many grains are

Figure 4. (continued)


9 of 31

TC4002



10 of 31

TC4002

Table 2. The 40Ar/39Ar, Rb/Sr, K/Ar Age Dataa

Sample Mineral JWeight,mg

GrainSize, mm

Total FusionAge, Ma

WeightedMean Age, Ma

IsochronAge, Ma MSWD 40Ar/36Ar %39Ar Used

Tien ShanGarm–Turkestan-Alai

TS12b bio 0.0007562 5 125–250 260.2 ± 0.8 277 ± 0.8b 277.2 ± 1.8 5.0/1.7 279 ± 2 92TS20a bio 0.0007592 4.7 125–250 277 ± 0.8 277 ± 0.8 277.9 ± 1.5 102/1.9 288 ± 7 98

PamirsKunlun

AD2e hbl 0.0007632 63 125–250 371.6 ± 1.1 357.2 ± 1.0c 357.1 ± 2.5 13/3.8 503 ± 8 40AD6b hbl 0.0007678 5 125–250 319.9 ± 4.6d na na na na na

Karakul-MazarP20 bio 0.0014516 2.2 63–200 202.6 ± 1.1 203.1 ± 1.1 203.6 ± 1.2 3.7/1.8 274 ± 10 98P20 mus 0.0014668 4.4 63–355 206.9 ± 1.1 206.9 ± 1.0 207 ± 1.1 1.1/1.9 294 ± 25 97P22 bio 0.0014649 0.4 63–355 206.2 ± 1.3 206.8 ± 1.3 205.1 ± 2.5 5.2/1.8 338 ± 34 97P24 bio 0.0014638 2.7 63–355 203.6 ± 1.1 204.1 ± 1.1 203.1 ± 1.1 1.4/1.9 309 ± 5 67P25 bio 0.0014555 0.4 125–250 190.1 ± 1.1 191 ± 1.1 192.4 ± 2.9 7.0/1.7 261 ± 62 96P26 bio 0.0014722 3.9 63–250 195.7 ± 1.0 196.2 ± 1.0 196.5 ± 1.1 3.7/1.7 288 ± 9 99

QiangtangM96A7 mus 0.0007469 4.8 125–250 101 ± 0.3 101.2 ± 0.3 101.9 ± 0.3 9.9/1.6 288 ± 1 100

Rushan-PshartA96M18h mus 0.0006961 3.4 125–250 162.5 ± 1 166.2 ± 0.5 166.5 ± 1.5 56/2.3 293 ± 11 8596P4e mus-la 0.000707 4.9 125–250 164.2 ± 1.8 151.4 ± 1.7e 151.4 ± 5.1 1.3/1.8 436 ± 73 100P7 mus 0.001511 0.2 125–250 105.3 ± 1.2 110 ± 0.9 100.9 ± 6.0 4.4/1.8 266 ± 15 100

Tirich Mir–Shyok–Kohistan-LadakhP2 bio 0.0015132 4.6 63–355 98.5 ± 0.6 98.2 ± 0.6 98.3 ± 0.7 4.0/1.9 295 ± 5 99P5 bio 0.0015132 1.5 200–355 70 ± 0.7 71.6 ± 0.5 73 ± 1.1 5.0/2.4 284 ± 5 82

Eastern TibetJinsha Suture

F218 hbl 0.004557 11 125–250 233.9 ± 2.2 219.9 ± 2.1f 219.9 ± 6.1 34/2.4 1260 ± 354 92F218 bio 0.004552 1.4 200–350 198.4 ± 1.9 198.6 ± 1.9 199.2 ± 2.0 76/1.8 290 ± 4 88

Songpan-Garze (Longmen Shan)TR257 bio 0.003346 0.3 125–250 748.2 ± 3.1 751.9 ± 3.1g 751.8 ± 3.9 9.1/1.9 931 ± 143 74

Sample Mineral Rb, ppm Sr, ppm 87Rb/86Srh 87Sr/86Sr 87Sr/86Sr*i Age, Ma

QiangtangM96A7 whole rock 209 212 2.8547 0.716054 ± 0.000010

muscovite 458.3 19.97 67.058 0.810614 ± 0.000012 0.711850 ± 0.000063 103.6 ± 1.1Rushan-Pshart

96P4e whole rock 250 121 5.9928 0.733269 ± 0.000010muscovite 937.4 13.47 212.77 1.287214 ± 0.000015 0.717215 ± 0.000229 188.4 ± 1.9

P7 whole rock 318 91 10.129 0.726648 ± 0.000011muscovite 1001 12.94 232.08 1.077421 ± 0.000011 0.710640 ± 0.000233 111.2 ± 1.1

Sample Mineral K, % 40Ar, ng/g Age, Ma

96M18a sericite 5.57 ± 0.04 35.3 ± 0.4 89 ± 2

aJ is the irradiation parameter; MSWD is the mean square weighted deviation [Wendt and Carl, 1991], which expresses the goodness of fit of theisochron [Roddick, 1978]; isochron and weighted mean ages are based on fraction of 39Ar listed in the last column. Abbreviations are as follows: bio,

biotite; mus, K-white mica; hbl, hornblende; na, not analyzed; la, laser total-fusion age. Complete tabulated data are available in Table A5 (auxiliary

material). Weighted mean plateau age values in italics; preferred age interpretation in bold.bWeighted mean age recalculated with 40Ar/36Ar of 279.cWeighted mean age recalculated with 40Ar/36Ar of 503; youngest steps down to �40 Ma, possibly homogeneously distributed excess 40Ar, probable age

350 ± 20 Ma.dHighest temperature step at �354 Ma.eWeighted mean age recalculated with 40Ar/36Ar of 436.fWeighted mean age recalculated with 40Ar/36Ar of 1260; formation at �220 Ma, loss profile to 25 Ma.gWeighted mean age recalculated with 40Ar/36Ar of 931; formation at 752 ± 5 Ma, loss profile to <386 Ma.hThe maximum error is ±1%.iNormalized for 86Sr/88Sr = 0.1194; maximum error 0.05%.


11 of 31

TC4002

broken. In the Pupin diagram, these zircons occupy thecalc-alkaline and subalkaline fields. Fractions F6 and F7plot near but above the concordia and F2 and F3 near butbelow the concordia. Fraction F8 probably includes olderinherited zircons. F5 might have suffered Pb loss. A lowerintercept age at 74.2 ± 0.7 Ma, calculated using F2, F3,F6, F7, and F8, is taken as the crystallization age; theupper intercept is �1900 Ma (Figure 4). Sample 96M18a,a fine-grained clastic layer within massive dolomite oflikely Silurian age [Vlasov et al., 1991], dates low-grademetamorphism of the Paleozoic-early Mesozoic strata ofthe Qiangtang cover sequence. The <2 mm sericite closedat �89 Ma (conventional K/Ar, Table 2), in agreementwith the widespread Cretaceous thermal overprint of theQiangtang block in the eastern Pamirs.[13] Because of the lack of datable material from the

boundary between the Sarykol-Karakul volcaniclastic rocks(Karakul-Mazar system, see below) and the Qiangtangblock (the Jinsha suture) along our studied section, wedated a basement paragneiss along the west-northwesternedge of the Qiangtang block in eastern Tibet (sample F218;Figure 1b and Table 1). The hornblende-biotite gneissoccurs as an exotic horse associated with amphibolite andserpentinite and floating in likely Permian melange rocks[Map Compiler Group, 1986]; this zone was mapped as theJinsha suture [Ratschbacher et al., 1996]. The hornblendeand biotite closed at �220 and �199 Ma, respectively(40Ar/39Ar age; Figure 5).2.1.2.5. Rushan-Pshart Zone, Central Pamirs[14] The Rushan-Pshart zone is an upper Paleozoic

to Jurassic volcano-sedimentary succession, probablyrepresenting a rift-margin and/or an oceanic basin–(?)arcsequence; the basin was closed by Early Cretaceoustimes [Leven, 1995] (Figure 6d, eastern Pshart section).Monzogranite L96M25a intrudes a Permo-Triassic meta-volcaniclastic sequence at the southeastern end of thePshart valley north of a major Cenozoic thrust. CL imagesshow that the long prismatic zircons are zoned, with thezonations interrupted in many grains by round resorptionrims. Large, mainly metamict cores abound. In the Pupindiagram, these zircons plot into the calc-alkaline tosubalkaline fields. The four fractions plot below theconcordia, probably owing to insufficient leaching by ayounger overgrowth or substantial Pb loss. We suggest thatthe crystallization age is between 167 and 235 Ma, usingthe lower intercepts of the discordias from fractions F2 +F4 and F4 + F5, respectively (Figure 4).[15] Monzogranite L96A9 also intrudes the upper Paleo-

zoic–Triassic volcano-sedimentary cover sequence; itssouthern border is a Cenozoic fault. In the Pupin diagram,the zircons represent alkaline and peralkaline mantle differ-entiates and a few zircons point to hybrid calc-alkalineorigins. CL images reveal two zircon types. One has no

cores and displays zoning, occasionally with resorptionfeatures; the other has large, mainly metamict cores. Frac-tions F1, F3 and F5 are clear pinkish zircons of variablesize, and F2 and F4 are huge euhedral prismatic grains.These fractions plot close to but below the Concordia but nosingle discordia can be drawn through all of them. Weillustrate a variety of fits in Figure 4 that likely straddle thepossible age of this rock; these lower intercepts are at �174to �205 Ma. SHRIMP analysis of 12 grains sampled coreand rim regions and zoning patterns (Figure 7, L96A9images). Three cores range from 194 to 202 Ma and thecorresponding rims yield ages between 154 and 171 Ma.Assuming that the older group of spot ages is representativeof the cores, the weighted mean of 13 measurements is201.3 ± 4.4 Ma. A younger group, possibly typifying rimages, yields a mean age of 170 ± 6 Ma. The youngest rim is�146 Ma, the oldest �217 Ma.[16] Sample A96M18h, a chlorite-muscovite-staurolite-

garnet schist at the base of the Permo-Triassic sequencenorth of L96M25a, was mapped as possibly Carboniferous[Vlasov et al., 1991]. Its 166.2 ± 0.5 Ma 40Ar/39Ar musco-vite age (Figure 5) reflects cooling after metamorphism orpluton emplacement in the Rushan-Pshart zone. Our inter-pretation of this age as postemplacement cooling fits the�167 Ma age suggested for L96M25a and may mean thatthe unusual high-grade metamorphism in the sedimentaryrocks north of L96M25a was produced by contact meta-morphism. We mapped this zone for a few kilometers alongstrike in the lower Pshart valley, where its exposure isconfined to the hanging wall of a major Cenozoic thrustthat emplaces these rocks onto Tertiary [Vlasov et al., 1991]red beds. Monzogranite pebble 96P4e from a conglomerateof a probable Miocene [Vlasov et al., 1991] intramontanebasin within the Rushan-Pshart zone (uppermost Pshartvalley) yielded Rb/Sr and 40Ar/39Ar muscovite ages of188 Ma and 151 Ma, respectively. These ages indicateJurassic magmatism and cooling in the Rushan-Psharthinterland. Two-mica monzogranite P7, from the sameintrusion complex as sample L96A9, yielded a distributedwhite mica 40Ar/39Ar spectrum with a total fusion age of�105 Ma and a Rb/Sr age of �111 Ma.2.1.2.6. Tirich Mir––Karakoram––Gangdese Arc onHindu Kush––Lhasa Block, Southern Pamirs[17] In the northernmost part of the southern Pamirs, a

series of granites and granodiorites intrude late Paleozoicto Jurassic rocks in blocks rimmed by Cenozoic dextraltranspressional faults; these blocks were thrust northwardonto the Rushan-Pshart zone northwest of Murgab [Vlasovet al., 1991; L. Ratschbacher et al., manuscript in prepa-ration, 2004]. Granodiorite P2 yielded one discordantzircon fraction. SHRIMP analysis of 15 zircons revealedages from well-rounded cores (Figure 7, P2 images) of�1624 Ma, 902 Ma, 417 Ma, 355 Ma, and 196 Ma. One

Figure 5. New 40Ar/39Ar spectra and Rb/Sr whole rock–white mica ‘‘isochrones’’. See Table 1 for sample locations,Table 2 for age data and interpretations, and Table A3 in auxiliary material for data listing. Weighted mean ages (WMA)and weighted mean plateau ages (WMPA) were calculated using shaded steps. TFA, total fusion age, IA, isochron age.Uncertainties are ±1s. ‘‘Atm.’’ in the isochron diagrams is the 36Ar/40Ar ratio of the atmosphere (1/295.5). MSWD, meansquare weighted deviation.


12 of 31

TC4002

Figure 5


13 of 31

TC4002



14 of 31

TC4002

158 Ma spot is from a nodular inner grain portion. Fifteenspots from rims and multifaced, idiomorphic grains rangefrom 126 to 102 Ma with a weighted mean age of 118.6 ±3.6 Ma (Figure 4). Sample P5 from a monzogranite furtherwest yielded two-zircon fractions that plot close to thelower discordia intercept. A simple model assuming minorinheritance places the crystallization age at 74.5 ± 8.8 Maand the inherited component at �710 Ma (Figure 4). Thebiotite 40Ar/39Ar ages of P2 and P5 are 98 and 72 Ma,respectively, recording postemplacement cooling. The P5biotite data are suggestive of a thermal disturbance youn-ger than 45 Ma, possibly due to Cenozoic thrusting andwrenching along the block-bounding faults.

2.2. Geochemistry

[18] We next evaluate the first-order tectonic settingof magmatic rocks of the southern Tien Shan and easternPamirs by major and trace element geochemistry (Table A4in auxiliary material) and Sr and Nd isotope data (Table 3and Figure 8). QAP classification identifies the Garm–Turkestan-Alai plutons as monzogranite and granodiorite,nearly all Kunlun metavolcanic rocks as basaltic andesite

and dacite, the Karakul–Mazar batholith rocks as monzog-ranite-granodiorite and all other plutons as monzogranite.Major and trace element discrimination classifies thesemagmatic rocks as volcanic arc related. That said, there area few trends and exceptions (Figures 8b–8g): (1) TheKunlunrocks are low-K, metaluminous, and indicative of an imma-ture arc. (2) In contrast, the Cretaceous plutonic rocks arehigh-K, peraluminous, S-type, and likely represent a mature-arc to late and postcollisional setting; the Karakul–Mazarbatholith rocks and, less clearly, the Garm–Turkestan-Alairocks occupy intermediate positions between these extremesin all classification diagrams. (3) The Triassic leucogabbroand the Cretaceous diorite, both of which penetrate theMuskol-dome basement (the Karakul–Mazar metasedimen-tary/volcanic rocks, see below) of the central Qiangtang, areexceptional. The leucogabbro is of within-plate origin. TheCretaceous subvolcanic, metaluminous diorite has a low-K,I-type signature and is transitional between subalkaline andalkaline. Chondrite-normalized REE patterns from all plu-tonic rocks (Figure 8h) are characterized by concave-upwardshapes and variably negative Eu anomalies. The immaturemetavolcanic Kunlun rocks lack an Eu anomaly; only morefractionated samples (e.g., rhyolite AD2a) show an anomaly.



15 of 31

TC4002

The Cretaceous granitoids show enrichment in LREE withrespect to the HREE, but offer different groups of LaN/LuNratios: the ratios of the southern Pamirs samples are 26–28,the Rushan-Pshart sample (P7) ratio is 22, and the CretaceousQiangtang samples have ratios of 13–16 with the exceptionof 96A10b, which has a ratio of 40. They also have aconsistent pattern of fractionation within the LREE andHREE groups (LaN/SmN= 4–5.5) and a negative Eu anomaly(Eu/Eu* = 0.5–1.0). Our grouping of samples is intended tohighlight common features (Figure 8h): The granitoids in-truding the Rushan-Pshart volcano-sedimentary successionand southern Qiangtang block are similar despite theirTriassic-Cretaceous age range, but they all have a Jurassicage component and probably typify the Rushan-Pshart arc. Inmajor and trace element geochemistry, the Triassic leucogab-bro and the Cretaceous diorite of the Qiangtang block aredifferent from the rest of the samples. They lack a Eu anomalybut have a similar REE pattern, likely reflecting an influenceby a similar lithospheric structure, as discussed later.[19] Initial Sr ratios are lowest for the Kunlun arc rocks

and for the central Qiangtang Triassic leucogabbro.Accordingly, their (143Nd/144Nd)i values are the highest.Their source region was close to the chondritic uniformreservoir (Figure 8i). The Sr initial ratios increase and theNd initial ratios decrease from the Karakul–Mazar andRushan-Pshart plutons (which have a nearly identicalpattern) to the Cretaceous granitoids, probably indicating anorth-south difference in the composition of the lithospheremantle and/or the mixing of crustal and mantle melts (seebelow). Nd model ages for the Kunlun arc and Karakul–Mazar magmatic rocks appear to be younger than for theQiangtang, Rushan-Pshart, and southern Pamiran rocks, thusagain indicating a north-south difference across the Pamirs;the Cretaceous-Cenozoic magmatites of the Lhasa block inTibet have the youngest model ages (Figure 8k). In thePamirs, a characteristic outliner is given by the 0.78 Gamodel age of the Qiangtang leucogabbro, emphasizing anorthern (Kunlun and Karakul–Mazar) affinity.

3. Stratigraphy

[20] To place our samples in a stratigraphic context, tosubstantiate our interpretation of the magmatic belts of thePamirs, and to support our lateral correlations, we presentidealized and simplified stratigraphic sections, condensedfrom detailed mapping reports and supplemented by ourown observations (Figure 6). The often detailed compila-tions on the lithostratigraphy of the Tien Shan and Pamirsare difficult to access and evaluate in their chronostrati-graphic detail, and for this reason they await checking bygeochronology and testing in modern tectonic models;pertinent to this paper are the summaries by Kravchenko[1979], Bazhenov and Burtman [1982], Burtman andMolnar [1993], Leven [1995], Brookfield [2000], and, inparticular, Vlasov et al. [1991]. For purposes of discussion,

Figure 9 integrates this stratigraphic information with thenew geochronology and geochemistry in this paper andprovides abbreviated interpretations for the geologic units ofthe southern Tien Shan and eastern Pamirs, their age,tectonic setting and amalgamation history.

3.1. Garm––Turkestan-Alai Zone

[21] The Garm block contains Precambrian basement thathas yielded a zircon age of �2.6 Ga (Figure 3) andamphibolite-grade, possibly Paleozoic metaclastic and rareultramafic rocks. The Turkestan-Alai complex comprisesearly Paleozoic clastics and late early Paleozoic carbonates.The Carboniferous is clastic and in part volcanogenic.The northern Turkestan unit is overthrust with southvergence by the Silurian to mid-Carboniferous oceanicand accretionary wedge nappes of the Turkestan ocean.Chiefly Early Permian, but also Late Devonian to LatePermian granitoids intrude these units (Figures 6a and 9a)[Vlasov et al., 1991; Brookfield [2000].

3.2. Kunlun Arc and Karakul––Mazar System

[22] This area (Figure 6b) covers the central northernPamirs and is not well understood (compilation after Vlasovet al. [1991] and our own mapping; see also Bazhenov andBurtmann [1982] (in particular their Figure 3) and Burtmanand Molnar [1993]). Kurgovat is probably a microcontinent,the Viskharv and Vanch are volcanic arcs, and oceanic crustis present. The base of the probable Lower Carboniferousoceanic sequence is dominated by mafic lavas, the UpperCarboniferous section is heterogeneous; the magmatic arcsequence is capped by limestones as young as Permian thatare replaced laterally by thick greywacke, in turn overlainby molasse. This sequence is partly replaced in the Karakullake area by thick metasandstones and phyllite of Silurian-Devonian to Permian, but probably also Triassic age. A largepart of it may be turbidite. The northern rim of thisvolcaniclastic sequence consists of a belt of ?Triassic ande-sites and coarse clastic rocks. Two zones of locally meta-morphosed siliciclastic rocks, rich in metavolcanics andcontaining numerous ultramafic and mafic lenses (serpenti-nized peridotite, metagabbro and metabasalt), occur discon-tinuously along the likely traces of the Kunlun and Jinshasutures (Figure 3). These are called the Markansu zone southof the Kunlun arc and the Kurgovat basement and theAkzhilga zone [Burtman and Molnar, 1993] within thesouthern part of the Karakul–Mazar system (Figure 3).

3.3. Qiangtang Block and Muskol and Sares Domes

[23] The stratigraphy of Qiangtang block in the centralPamrs (Figure 6c) is poorly known (see Vlasov et al.[1991] and our own mapping). Upper Carboniferous andLower Permian sandstone, limestone, and marl unconform-ably cover Lower Paleozoic shale and carbonates. TheTriassic to Lower Jurassic clastic sequence becomes

Figure 6. Idealized stratigraphic columns of the major lithotectonic units of the Pamirs simplified and modified afterVlasov et al. [1991], Kravchenko [1979], Bazhenov and Burtman [1982], Burtman and Molnar [1993], Leven [1995], andBrookfield [2000], complemented and arranged by our observations. The interpretations of the tectonic settings are our own.


16 of 31

TC4002

Figure

6


17 of 31

TC4002

Figure 7. Selected cathodoluminescence images of zircons from the Pamirs and location of SHRIMPspot ages. Light areas are low U cores or growth rims. Ages in million years.


18 of 31

TC4002

Table

3.Isotopic

Ratiosa

Sam

ple

Age,

Ma

Sr

Rb

Sm

Nd

87Rb/86Sr

87Sr/86Sr

87Sr/86Sr(i)

147Sm/144Nd

143Nd/144Nd

143Nd/144Nd(i)

eNd(T)

eNd(0)

TDM

beSr(0)

eSr(T)

Kunlun

AD2a

329±5c

74.32

1.148

1796

6.478

0.045

0.705596±08

0.705384

0.1676

0.512869±10

0.5125080

5.73

4.51

0.60

16

18.16

AD6b

350±20d

533.5

108.7

1.830

6.534

0.590

0.708277±09

0.705264

0.1728

0.512646±10

0.5122443

1.24

0.16

0.99

54

16.89

AD6c

329±5c

427.0

41.45

2.680

12.70

0.281

0.706972±09

0.705642

0.1276

0.512466±09

0.5121912

�0.45

�3.36

1.11

35

21.81

Karakul-Mazar

P22

216±2c

216.4

148.6

5.174

27.88

1.987

0.713857±08

0.707623

0.1122

0.512308±09

0.5121479

�4.09

�6.44

1.32

133

48.05

P26

227±5c

198.4

175.2

4.530

34.84

2.557

0.715323±09

0.706974

0.0786

0.512276±11

0.5121592

�3.64

�7.06

1.29

154

38.99

Qiangtang

96M9a

�170c

495.5

15

8.679

55.37

0.089

0.705601±10

0.705311

0.0967

0.512623±07

0.5124774

2.64

�0.29

0.78

16

15.43

M96A7

�90c

212

209

4.575

24.38

2.887

0.716052±10

0.710717

0.1158

0.512228±08

0.5119825

�9.53

�10.87

1.70

164

90.48

96A10b

126±50c

296

120

11.78

88.01

1.187

0.711870±10

0.709745

0.0826

0.512222±10

0.5119709

�9.85

�11.68

1.72

105

76.60

Rushan-Pshart

L96M25a

167–235c

217.5

155

5.151

27.22

2.087

0.714926±10

0.707951

0.1168

0.512070±09

0.5120484

�5.60

�8.00

1.46

148

52.99

L96A9

�170c

161.5

183

5.147

25.29

3.319

0.718987±10

0.709310

0.1256

0.512081±10

0.5120535

�6.62

�8.11

1.49

206

71.78

P7

111±2e

90

317.5

8.168

53.32

10.34

0.726548±10

0.710234

0.0945

0.512039±10

0.5120014

�9.64

13.00

1.69

313

83.29

TirichMir–Shyok–Kohistan-Ladakh

P2

119±4c

324

139

19.46

46.99

1.256

0.713211

±10

0.711105

0.2555

0.512017±09

0.5118197

�13.00

13.00

1.97

124

95.77

aErrors

are2standarddeviations(2s).

bTDM,depletedmantlemodel

ages.

cU/Pbzircon.

dAr/Arhornblende.

eRb/Srmuscovite.


19 of 31

TC4002

Figure 8


20 of 31

TC4002

increasingly rich in carbonate up-section. Upper Jurassicand Lower Cretaceous strata are usually absent. The entirepackage is interpreted as a passive margin or platformsequence by Burtman and Molnar [1993]. Our mappingtypified the basement of the Muskol and Sares domes asmonotonous biotite gneiss, kyanite-bearing garnet-micaschist, quartzite, quartz-actinolite schists with relictturbidite sequences (thin, graded conglomeratic layers,sand-siltstone, black shale), marble with subordinate am-phibolite and greenschist, locally with preserved pillowlava structures, ophicalcite, and intermediate to ?acidmetavolcanics (plagioclase-rich gneiss). Small gabbro bod-ies (cf. sample 96M9a) intrude the clastic rocks and

dolomite. Metamorphism is greenschist to upper amphib-olite facies with local migmatization of Tertiary age(L. Ratschbacher et al., manuscript in preparation, 2004).

3.4. Rushan-Pshart Zone and Southeastern Pamirs

[24] This locally well-studied unit [Leven, 1995](Figure 6d, eastern Pshart section) is principally Permianand Triassic limestone and radiolarian chert associated withpillow basalt, andesite, tuff and small ultramafic bodies. Theoverlying Jurassic and Cretaceous units comprise thickclastic sequences. Unconformable and probably Cretaceousbeds are interpreted as molasse deposits postdating closure


Figure 8. Major, trace, and isotope geochemistry of the magmatic rocks of the Pamirs. Eastern Kunlun and easternGangdese and Nyainqentangla data shown as insets in Figure 8h and 8k are for comparison and are from Harris et al.[1988a, 1988b]. Songpan-Garze data in Figures 8h–8k are from Roger et al. [2003], the data of the North Tibet and SouthTibet Cenozoic volcanics are from the compilation of Maheo et al. [2002], and those of the Gangdese magmatic rocks(Figure 8j) fromMiller et al. [2000]. Mixing curve between depleted mantle and High Himalayan Crystalline rocks and thatbetween depleted mantle and lower crust are based on data given by Ding et al. [2003] and Miller et al. [2000],respectively; all literature data are in italics.


21 of 31

TC4002

of an ocean basin in the latest Jurassic or Early Cretaceous.An ophiolite-bearing thrust system (the stippled outline inFigures 2 and 3), possibly rooting in the Rushan-Pshart zone,is exposed in the southwestern Pamirs (Figure 6d, Chatyr-tash–SE-Pamir section), and was likely emplaced in the LateTriassic to earliest Jurassic [Pashkov and Budanov, 1990;Dronov, 1986, 1988]. The southernmost thrust system in thesoutheastern Pamirs contains Jurassic reef limestone andLower Cretaceous conglomerate, limestone, and andesite-dacite (Figure 6d, Gurumda–SE-Pamir section) (see Vlasovet al. [1991] and our own mapping). It was probablyemplaced in the Early Cretaceous. The relationship of theseallochthonous volcaniclastic sections to the Rushan-Pshartzone is speculative [see, e.g., Leven, 1995; Pashkov and

Budanov, 1990]. An affiliation of the Lower Cretaceousvolcaniclastic section with the Tirich Mir–Karakoram–Gangdese arc (see below) is another viable interpretation.Other sections in the southeastern Pamirs display Paleozoicclastic and carbonate platform sedimentation, followed byrifting and passive-margin formation in the Carboniferous toTriassic.

3.5. Tirich Mir––Karakoram––Gangdese Arc onHindu Kush––Lhasa Block

[25] The southwestern Pamirs are most poorly studied.They contain Precambrian basement intruded by mostlyCretaceous and rarely Cenozoic plutons [Vlasov et al.,

Figure 9. Tectonic scenarios, including stratigraphic and magmatic evolution, for the southern TienShan and Pamirs based on the new data and the literature discussed in text. The arrangement of Figure 9reflects the Cenozoic thrusting of the northern Pamirs onto the Tien Shan.


22 of 31

TC4002

1991; Kravchenko, 1979]. Sedimentary cover is locallypreserved as Mesozoic platform sediments. The basementhas yielded 2130 ± 160 Ma and 1970 ± 170 Ma Rb/Srisochrons and a �1870 Ma upper intercept zircon age(data summarized by Hubbard et al. [1999]). The entiresouthern Pamirs was affected by Cretaceous and Cenozoicmagmatism and metamorphism. Further south in the HinduKush and Karakoram regions, metamorphism and magma-tism are better documented [Fraser et al., 2001; Hildebrandet al., 2001] and may be related to an Andean-type marginalong the southern margin of Asia that was active beforeand during its collision with the Kohistan arc (Shyok suture)and India (Indus-Yarlung suture).

4. Discussion

[26] In the following we discuss the amalgamation his-tory of the southern Tien Shan and the Pamirs and constructcorrelations with Tibet. In the Tien Shan panel of Figure 9awe suggest a tectonic scenario for the southern TienShan based on data compiled from Vlasov et al. [1991],Zonenshain et al. [1990], and Brookfield [2000], usingprincipally the nomenclature of Brookfield [2000]. TheEarly Permian Garm–Turkestan-Alai granitoids, typifiedby our southern Tien Shan samples, are part of a plutonicassociation dominating the Gissar unit, but also occurringnorth of it in the Zeravshan and Turkestan-Alai units.This association corresponds to calc-alkaline complexes ofmature arcs (Figure 8f) [Solov’ev, 1998]. Our tectonicinterpretation suggests that the Tien Shan formed north ofthe Tarim block and that the Baysunta basement massif(Figure 9a) is part of the Tajik continent that connects withthe Tarim block underneath the northern Pamirs. Possiblewestward extensions of early Paleozoic sutures in northernTibet such as Kudi (Figures 2 and 3) must be located southof the Tajik continent. Zonenshain et al. [1990] comparedthe Tarim block with the Karakum massif of Uzbekistan andconsidered the Baysunta massif as a microcontinent locatedto the south of a Carboniferous, south directed (in presentcoordinates) subduction zone south of the Tarim block. Toour knowledge, evidence for late Paleozoic southwardsubduction is absent in the southwestern Tien Shan as wellas northern Tibet (Figures 2, 3, and 9a) [Xiao et al., 2002a].[27] The northern Pamirs and northwestern Tibet contain

the north facing Kunlun and the south facing Jinsha suturesas well as an intervening, likely mostly Carboniferous-Triassic, volcaniclastic subduction accretion system we callKarakul–Mazar (Figure 9b). We correlate Karakul–Mazarwith the Songpan-Garze–Hoh Xi system of central andeastern Tibet, where it comprises thick, predominantlyTriassic, deep marine turbidite and forearc basin deposits[e.g., Yin and Harrison, 2000]. The Kunlun arc has beeninterpreted as a juxtaposition of arcs and microcontinents[e.g., Xiao et al., 2002a, 2002b]. In Figure 9b we suggest atectonic scenario for the northern Pamirs, and in Figure 10we attempt an integration of the northern Pamirs withnorthwestern Tibet. Xiao et al. [2002a] showed that the earlyPaleozoic Kunlun (the ‘‘north Kunlun’’) arc of northwesternTibet was built by north verging Early to Middle Ordovician

collision of a Late Cambrian-Early Ordovician intraoceanicarc (the Yixieke arc) with the Tarim block; north dippingsubduction beneath this collage led to the Middle Ordovi-cian-Silurian accretion of the Kudi microcontinent, andnorth dipping subduction and accretion along this amalga-mated complexmight have continued uninterrupted, althoughwith diminished magmatic and sedimentary expression, intothe Early Carboniferous (Kudi, Figure 2) or Permian (Oytag,Figure 2). Xiao et al. [2002b] reinterpreted the late Paleozoic-earlyMesozoic stratigraphy and facies of northwestern Tibet,defining a Carboniferous-Triassic subduction accretionscenario that included two north facing subduction com-plexes and a forearc-basin succession; these units werestitched together by 215–190 Ma plutons (the ‘‘southKunlun arc’’). The available geochronology (Figures 2, 3,and 10; Table A1 in auxiliary material) [see also Cowgill etal., 2003] suggests that the major stage of Kunlun arcformation in northwestern Tibet is early Paleozoic, reflectedin zircon ages of gneisses and granitoids from 491 to 452Ma;�405 Ma lamprophyres and a 384 ± 2 Ma pluton areposttectonic. Plutons in the eastern Kunlun of north centralTibet have a similar 518 to 389 Ma age range (Figure 2),while volcanic rocks likely cover the 370–320 Ma interval[Dewey et al., 1988]. Molasse deposited on Tarim basementis poorly constrained as Devonian in northwestern Tibet[Matte et al., 1996] and as Upper Devonian in the easternKunlun [Dewey et al., 1988]. The younger stage of Kunlunarc formation in northwestern Tibet is bracketed by the 277 ±6Ma Qytag granite gneiss and the 180 ± 10 Ma red porphyrysouth of Kudi; intervening are a few U/Pb crystallizationand Rb/Sr and 40Ar/39Ar cooling ages. This stage is probablybest recorded by the approximately Early Carboniferous to(?)Middle Triassic Mazar accretionary prism in northwesternTibet (Figure 10) [Xiao et al., 2002b]. In the eastern Kunlun,the younger stage is represented by the north Kunlunbatholith, dated at about 260–240 Ma [Harris et al.,1988a]. The chemistry of these plutons suggests an activecontinental margin [Harris et al., 1988b].[28] Given the remoteness of the northern Pamirs and the

locally intense Cenozoic tectonometamorphic overprint,reliable interpretation of the lithostratigraphy in terms oftectonic setting (Figures 6b and 9b) is difficult. We suggestthat the Early Carboniferous to Triassic arc complexesdocument the Kunlun arc in the Pamirs (Altyn-Dara belt;Figure 6b). Volcanism began between �370 and 320 Ma(Figure 3) and most likely extended into the Triassic.The geochemistry of the volcanic rocks suggests an arcsetting (Figure 8), probably evolving from an ocean basinsuccession (e.g., Kurgovat and Sarykol-Karakul sections,Figure 6b). We speculate that Kurgovat may be a correlativeof the early Paleozoic Kunlun arc of Tibet. The �400 Mazircon age from posttectonic granitoid P26 intruding thenorthern Sarykol-Karakul complex (Figures 3 and 6b) mayindicate recycling of Devonian arc rocks into the Carbon-iferous-Triassic subduction accretion complexes of thenorthern Pamirs. We interpret the Sarykol-Karakul complexas an accretion system that may be correlated with thesiliciclastic rock-dominated Mazar–Songpan-Garze–HohXi system of Tibet. The Markansu zone (Figure 3), rich in


23 of 31

TC4002

volcanics and containing ultramafic and mafic lenses southof the Kunlun arc and the Kurgovat ‘‘basement,’’ mayconstitute the subduction complex related to the Kunlunsuture (Figures 9b and 10). Similarly, the Akzhilga zone(Figure 3) within the southern part of the Karakul–Mazarunit may denote a subduction complex related to the Jinshasuture (Figures 9b and 10; see discussion below). Wetherefore interpret the cryptic Tanymas suture [Burtmanand Molnar, 1993] in the southern northern Pamirs as theJinsha suture. Similar to Tibet, there is a distinct agedifference between northward subduction beneath Kunlun,Ordovician-Triassic in northwestern Tibet and ?Devonian-Triassic in the Pamirs, and Permian-Late Triassic, probablyJurassic southward subduction underneath the Qiangtang[Matte et al., 1996; Yin and Harrison, 2000; Kapp et al.,2003a; Roger et al., 2003] (Figures 9b, 10, and below).[29] Some 2500–3000 km to the east, the Qinling of

eastern central China offer an analogue to the subduction-arc scenario of the Kunlun subduction arc complexes of the

Pamirs and northwestern Tibet in a comparable plate setting[Ratschbacher et al., 2003]. There, intraocean arc formationat �470–490 Ma (Erlangping unit–possibly equivalent tothe Yixieke arc) was followed by accretion of this arc to theNorth China craton (equivalent to the Tarim Block), closingthe intervening basin (possible Kudi ophiolite emplacementin northwest Tibet). Northward subduction resulted in theaccretion of the Qinling microcontinent (equivalent to theKudi ‘‘terrane’’) and built a �440–390 Ma Andean-type arcon the North China craton and the previously assembledcollage (equivalent to the ‘‘north’’ Kunlun arc). Devonian–Permian closure of the Paleotethys built an accretionarywedge (the Liuling unit; equivalent to the Carboniferous-Triassic subduction accretion complexes of the Pamirs andTibet), locally incorporating Carboniferous eclogite andproducing andesitic magmatism on the North China craton(equivalent to the ‘‘south’’ Kunlun arc).[30] Posttectonic granitoids, stitching the subduction ac-

cretion systems of the Mazar–Songpan-Garze–Hoh Xi

Figure 10. Integration of the sedimentary and magmatic evolution of the northern Pamirs withnorthwestern Tibet. (right) Reinterpretation of Xiao et al.’s [2002a, 2002b] subduction accretion and arcevolution scenario for northwestern Tibet. (left) Corresponding evolution in the northern Pamirs. Insetlower left shows available U/Pb zircon crystallization ages defining the Kunlun arc magmatism; for dataand references see Table A1 in auxiliary material.


24 of 31

TC4002

complexes, are apparently rare in central Tibet. Smallintrusions are increasingly recognized in detailed mapping(Figure 2) [e.g., Roger et al., 2003] and are commonlyattributed to intervening arcs (e.g., Triassic Litang arcsystem of eastern Tibet [Yin and Harrison, 2000], orsouthward migration of Kunlun arc magmatism [Xiao etal., 2002b; Roger et al., 2003]). In contrast, Late Triassic–Early Jurassic granite plutonism (�200 Ma) is welldeveloped in northwestern Tibet and a massive batholith(�227 Ma) intrudes the Karakul–Mazar complex rocks inthe Pamirs (Figure 3); we suggest that the granitoids (217–206 Ma) dated by Roger et al. [2003] in central and easternTibet, which outcrop in the sedimentary system between theKunlun and Jinsha sutures, belong to this posttectonicgroup. The cooling of the batholith covers the same 220–180 Ma age range in the Pamirs and northwestern Tibet and,again, similar cooling ages have been reported from the fewplutons dated so far in central and eastern Tibet (Figure 11a)[Roger et al., 2003]. The S-type plutons of the Pamirscontain two-mica granites and have an arc to late/postcolli-sional geochemical signature (Figure 8). We suggest thatthey are distinct from the Kunlun arc rocks, being overallyounger than the ‘‘south’’ Kunlun arc rocks, and having astronger continental signature; we speculate that the lattermay derive from incorporation of the thick accretionarywedge sediments of the Karakul–Mazar complex in themelts. The larger surface area of exposed granitoids in thePamirs and northwestern Tibet than in the rest of Tibet canbe explained by the deeper exhumation level, resulting froma concentration of Cenozoic shortening over a shorter N-Sdistance in the west than the east of the orogen.[31] The Qiangtang block is best studied in central Tibet.

Its regional map pattern is a >500-km-long, east plunginganticlinorium that initiated to growth during the Cretaceousbut formed principally during the Cenozoic (Figure 2) [Yinand Harrison, 2000; Kapp et al., 2000, 2003a, 2003b]. Theanticlinorium is characterized by upper Paleozoic strata andmetamorphic rocks in its core and Mesozoic strata alongits limbs. The upper Paleozoic strata consist of Carbonifer-ous-Permian shallow marine siliciclastics and limestonesinterbedded with rift-related mafic volcanic and volcani-clastic rocks. The Triassic-Jurassic strata in the interior ofQiangtang block are shallow marine limestones, volcanicand volcaniclastic rocks, and nonmarine strata. Triassicvolcanic rocks are widespread along the northern marginof the block and are arc-related, attributable to southwardsubduction along the Jinsha suture [Kapp et al., 2003a, andreferences therein; Pearce and Mei, 1988]. Kapp et al.[2000, 2003a] demonstrated that within the central meta-morphic belt epidote-bearing blueschist melange, whosemetamorphism is at about 220 Ma, and Carboniferous-Triassic clastic rocks equivalent to Songpan-Garze strataoccur in tectonic windows framed by Late Triassic-EarlyJurassic normal faults. They suggested Late Triassic-EarlyJurassic flat-slab southward subduction of Paleotethyanoceanic lithosphere along the Jinsha suture. Early Mesozoicnormal faulting exposed the Songpan-Garze melange rocksin extensional core complexes. Southward subduction of theSongpan-Garze system beneath the Qiangtang block along

the Jinsha suture is thus well documented in central Tibetand the purely north vergent subduction, accretion and arcformation model for northwestern Tibet [Matte et al., 1996;Xiao et al., 2002b] should be modified to allow for arcformation in and large-scale underthrusting of Songpan-Garze rocks beneath the Qiangtang block. Our �220 Mahornblende and �200 Ma biotite cooling ages from thebasement silver within the Jinsha suture of eastern Tibetand similar ages further northwest along the Jinsha suture(�244 Ma) [Roger et al., 2003] corroborate Triassicsubduction all across Tibet.[32] There are striking similarities between the Qiangtang

block in the Pamirs and central Tibet: (1) As in Tibet, theregional structure of the Pamirs is a spectacular Cenozoicanticlinorium (the Muskol and Sares domes, Figure 3). Incontrast to Tibet, where the core of the anticlinoriumexposes rocks devoid of Cenozoic metamorphism, Cenozoicmetamorphic rocks, including migmatite and anatexite,occur in the Pamirs. The Qiangtang anticlinorium thusmay be traced from the central Pamirs, offset �150 kmby the Karakoram fault to western Tibet to central Tibet,with metamorphosed, mostly late Paleozoic rocks in its core(Figure 2). (2) As in Tibet, the metamorphic rocks in thePamirs are bound by normal faults. (3) The basement rocksexposed in the central Pamiran windows (Sares-Muskol–central Qiangtang section, Figure 6b) are interpreted asKarakul–Mazar–Songpan-Garze rocks that were metamor-phosed and strongly deformed during the Tertiary. Monot-onous siliciclastic and volcaniclastic rocks and maficvolcanics are the characteristic rocks and are interpretedas a subduction accretion association, similarly to the‘‘melange’’ rocks of central Tibet. The Triassic-Jurassicgabbros that intrude the biotite gneiss of the Sares-Muskoldomes (compare sample 96M9a) are similar to gabbros thatintrude siliciclastic rocks of the Akzhilga subduction com-plex, and probably correspond to diorites in central Tibet(Gangma Co diorite) [Kapp et al., 2003a]. These maficrocks suggest oceanic subduction during the Late Triassic-Early Jurassic. The �830 Ma upper intercept age of gabbrosample 96M9a matches the most common age mode indetrital zircons in the Qiangtang melange sedimentary rocksof central Tibet (800–1200 Ma, 65% of all zircons) [Kappet al., 2003a]. Similar age granitoids constitute the basementof the Songpan-Garze system in eastern Tibet and areinterpreted as part of the South China craton (see Rogerand Calassou [1997] and our �752 Ma biotite cooling agefrom the eastern Tibet Songpan-Garze basement granite).(4) The Qiangtang block stratigraphy is similar along strikefrom the Paleozoic to the Jurassic, but volcanic rocks seemto be the exception and Triassic-Jurassic rocks are moresiliciclastic in the Pamirs than in Tibet. (5) The granitoidsintruding the Qiangtang block have either similar �200–230 Ma ages or ion probe age components in the Pamirs andcentral Tibet. In contrast to the more northern blocks, theQiangtang contains Cretaceous granitoids (Figures 2 and11b). (6) Pre-late Paleozoic basement is rarely exposed inthe Qiangtang block. Zircons from a sliver of garnet-amphibole gneiss in central Tibet (Gangma Co gneiss)[Kapp et al., 2000] and the inherited zircons from Pamiran


25 of 31

TC4002

granitoid P17 are interpreted as samples of this pre-latePaleozoic basement in the Pamirs (age range 425–575 Ma,Figure 11c). The Amdo gneiss of eastern central Tibet [Xu etal., 1984], along the Qiangtang-Lhasa block boundary, has asimilar 531 ± 14 Ma age (Figure 2). This age group ischaracteristic of Gondwana crust north of the Lesser Hima-layas, for example the �508 and �562 Ma Kangmar

gneisses of southern Tibet [Lee et al., 2000; Scharer etal., 1986; see also DeCelles et al. [2000], Kapp et al.[2003a]).[33] The Shiquanhe-Gaize-Amdo thrust belt of Tibet [Yin

and Harrison, 2000] straddles the southern margin of theQiangtang block and the Bangong-Nujiang suture zone. Itinvades the northern and central Lhasa block and was

Figure 11. (a) Comparison of the cooling ages of the Karakul–Mazar batholith in the Pamirs andnorthwestern Tibet displayed as cumulative probability plots; for data see Table 2, Table A2 in auxiliarymaterial, and Matte et al. [1996], and references therein; the four reference data (shown as boxes withwidths corresponding to errors) from central and eastern Tibet are from Roger et al. [2003].(b) Comparison of ion probe single-grain zircon ages of granitoids intruding the Qiangtang block in thePamirs and central Tibet; data from this study and Kapp et al. [2000, 2003a]. (c) Comparison of ion probesingle-grain zircon ages from a sliver of basement gneiss (Gangma Co gneiss) [Kapp et al., 2000] fromcentral Tibet, the inherited zircons from Pamiran granitoid P17 [this study], the Amdo gneiss of easterncentral Tibet [Xu et al., 1984], and the Kangmar gneisses of southern Tibet [Lee et al., 2000].(d) Distribution of Cretaceous U/Pb zircon (lower intercept and rim ages) and Rb/Sr and 40Ar/39Armineral ages in the southern and central Pamirs. See text for discussion.


26 of 31

TC4002

developed during latest Jurassic ophiolite-accretionarywedge emplacement [Girardeau et al., 1985] and EarlyCretaceous-Tertiary thin-skinned thrusting, possibly rootingin the Bangong-Nujiang suture [e.g., Murphy et al., 1997;Kapp et al., 2003b]. The Bangong-Nujiang suture zone isoutlined by a more than 1000 km long belt of scatteredophiolitic fragments embedded in Jurassic flysch, melange,and volcanic rocks. Between Lhasa and Amdo, the belt isunusually broad, forming a system of ophiolite nappes(stippled outline in Figure 2) stretching �150 km fromthe suture to the south. A similar broad belt was mappedrecently in western Tibet [Kapp et al., 2003b]. The ophiolitefragments formed during the Jurassic, prior to the depositionof unconformably overlying uppermost Jurassic-lowermostCretaceous nonmarine to shallow marine strata [Zhou et al.,1997; Girardeau et al., 1984]. The stratigraphy of the easternPshart area conforms well to that of the Amdo region, but acomplete Triassic ocean basin sequence is preserved in thePamirs (Figure 6d, eastern Pshart section). The ophiolite-bearing thrust systems (Figure 6d, Chatyrtash–SE-Pamirand Gurumda–SE-Pamir sections) of the southeasternPamirs may root in the Rushan-Pshart zone (stippled outlinein Figure 3) and are probable equivalents of the broadLhasa-Amdo thrust system, although the emplacement agesmay differ. Arc formation is better documented for theRushan-Pshart than the Bangong-Nujiang suture. Intrusionof arc granitoids (Figure 8) into the Rushan-Pshartoceanic basin–arc sequence probably occurred between200 and 160 Ma. All the granitoids that intruded thesouthern Qiangtang block that we dated by ion probecontain a Middle Jurassic ages (170 ± 6 Ma rim ages inL96A9, �156–182 Ma core ages in Cretaceous granitoidM96A7). The �169 Ma lower intercept of sample 96M9amay indicate a mid-Jurassic disturbance in �230 Mazircons (Figure 4). We interpret this Middle Jurassic thermalevent as related to Rushan-Pshart arc formation along thesouthern margin of the Qiangtang block. The only knowncoeval event along the Bangong-Nujiang suture is regionalmetamorphism affecting the Amdo gneiss at �171 Ma(Figure 2) [Xu et al., 1985]. Intrusion of Jurassic Rushan-Pshart granitoids into the suture zone itself and the narrowwidth of the Jurassic magmatic belt north of the suture(Figure 3) [Vlasov et al., 1991] suggest either steep subduc-tion or considerable postcollisional shortening. Evidencefor the latter includes Cenozoic thrusts (equivalent to theShiquanhe-Gaize-Amdo thrust belt of Tibet) and transpres-sive wrench structures recording high strain. We concludethat the Rushan-Pshart suture in the Pamirs connects withthe Bangong-Nujiang suture of Tibet and is related tothe amalgamation of the Lhasa block with the rest of Asia.[34] The southern Pamirs of Tajikistan, the Wakhan

corridor of Afghanistan, and northern Pakistan remainone of the least studied areas in central Asia. The regionbetween the Indus-Yarlung and the Rushan-Pshart suturesis �300 km wide and contains Jurassic to Late Cretaceousplutonic rocks locally overprinted by Cenozoic leucogran-ite (Figures 2 and 3). In many aspects, the amalgamationhistory is speculative [see also Hildebrand et al., 2001;Fraser et al., 2001, and references therein]. The oceanic

Kohistan-Ladakh arc has mostly Late Cretaceous plutonemplacement ages (Figure 2) and was accreted to Asiaalong the north dipping Shyok Zone at about 80 Ma.Rolland et al. [2000] connected the Kohistan-Ladakh arcwith the Andean-type Gangdese arc of southern Tibet andinterpreted the Shyok suture as closing a back arc basin.North of the Shyok suture the Karakoram comprisesbelts of metamorphic and sedimentary rocks intruded bysubduction-related granitoids with predominantly EarlyCretaceous and locally Jurassic ages (Figure 2). In theHindu Kush region of northwestern Pakistan–easternAfghanistan the Karakoram is bound in the north bythe mafic and ultramafic rock-bearing Tirich Mir fault(?suture) zone; its eastward continuation is little known,but was proposed along the Kilik fault [Gaetani, 1997].Emplacement (suturing?) of the mafic-ultramafic rocksprior to 115 Ma was suggested [Hildebrand et al.,2001]. In the southern Pamirs, Cretaceous plutons intrudedEarly Proterozoic continental basement and a Cambrian toJurassic passive margin section (Figure 6d, Gurumdasection) [Vlasov et al., 1991]. As in the Qiangtang, someof the basement may turn out to comprise Phanerozoicrocks, affected by high-grade Cretaceous to Cenozoicmetamorphism [cf. Hubbard et al., 1999]. Plutonism isconcentrated in the central southern Pamirs, where up to50% of the surface exposure may be plutonic rocks. Wetraced Cretaceous plutonism and metamorphism as faras the central Pamirs. The currently furthest northernCretaceous intrusion is the subvolcanic diorite of thenorthern Qiangtang block. Characteristic for its positionabove the Karakul–Mazar complex that subducted under-neath the Qiangtang block, its inherited component(�1.9 Ga) traces a major age component of detrital zirconages in the Songpan–Garze system of eastern Tibet[Bruguier et al., 1997], and a major intrusion period inthe Tarim [Kapp et al., 2003a; Gehrels et al., 2003] andNorth China blocks [Grimmer et al., 2003]. Also itsgeochemical signature, being similar to the central Qiang-tang leucogabbro, is characteristic for its position abovethe Karakul–Mazar subduction accretion complex andthus for a northern Pamiran/Tibetan lithosphere (see alsobelow), and different from the other Cretaceous granitoids(Figure 8h). The latter are distinct in their geochemicalsignature, having Sr and Nd isotope patterns reflectingrecycling of continental crust (Figure 8). Exceptionally,Cretaceous plutons have also been described from theQiangtang block of Tibet (Figure 2).[35] The zircon geochronology (Figure 4) of the Creta-

ceous and Cenozoic Pamiran granitoids (see above) and ofcrustal xenoliths erupted from 90–100 km depth in thesoutheastern Pamirs during the Miocene (locality seeFigure 3) [Ducea et al., 2003; Hacker et al., 2004] chron-icles multiple Phanerozoic magmatic additions to a Protero-zoic crust derived from Gondwana. The most distinctmagmatic events occurred during the Cambro-Ordovician(410–575 Ma, Qiangtang and southern Pamirs), Triassic(210–250 Ma; Qiangtang, possibly by southward subduc-tion along the Jinsha suture), Middle Jurassic (147–195 Ma;Qiangtang and Rushan-Pshart; possibly by northward


27 of 31

TC4002

subduction along Rushan-Pshart–Bangong-Nujiang suture),and mainly Cretaceous. Judging from the distribution ofages from our U/Pb zircon (lower intercept and rim ages)and Rb/Sr and 40Ar/39Ar mineral ages in the southern andcentral Pamirs, the Cretaceous activity may have beenepisodic (Figure 11d), with peaks at �120 Ma and�80 Ma. Speculatively, we suggest that the mid-Cretaceousevent may reflect arc activity on Asia prior to the accretionof the Karakoram block along the Tirich Mir-Kilik ?suture(likely accretion at �115 Ma, see above). The LateCretaceous magmatism may portray flat-slab subductionof young (mid-Cretaceous) back arc crust north of theKohistan-Ladakh arc [Rolland et al., 2000], permittingmagmatism in the far interior of Asia (Shyok closure�80 Ma, see above). Prior to the India-Asia collision thePamiran region, stretching from the Indus-Yarlung sutureto the Jinsha suture, must have corresponded to a magmati-cally [this paper; Lutkov [2003]) and structurally [e.g.,Murphy et al., 1997; Kapp et al., 2003b] thickenedAndean-type plate margin.[36] Our study suggests a relatively simple first-order

crustal structure for the Pamirs and Tibet (Figure 12). Fromthe Kunlun arc in the north to the southern Qiangtang blockin the south, the Pamirs and Tibet likely have a dominantlysedimentary middle and lower crust, characterized byKarakul–Mazar–Songpan-Garze accretionary wedge rocks.The crust south of the southern Qiangtang block is likelygranodioritic, reflecting long-lived subduction and arc for-mation and Cretaceous-Cenozoic underthrusting of the arc

segments underneath the Qiangtang block [see also Murphyet al., 1997; Kapp et al., 2003b; Ducea et al., 2003]. Thisscenario is supported by the following data: (1) Karakul–Mazar–Songpan-Garze accretionary wedge rocks have beenmapped from the Kunlun arc to the central Qiangtang block[Kapp et al., 2003a; this study]; (2) Cretaceous magmatismis found as far north as the northern Qiangtang block [thisstudy]; (3) xenoliths from north central Tibet are mostlysedimentary [Hacker et al., 2000; Jolivet et al., 2003], butmostly granodioritic in the southeastern Pamirs [Ducea etal., 2003; Hacker et al., 2004]; (4) zircon geochronology inPamiran granitoids traces the sources of the Karakul–Mazar–Songpan-Garze rocks, i.e., Tarim and North andSouth China, as far south as the southern Qiangtang block,and Gondwana crust as far north as the northern Qiangtangblock [this study]; (5) isotope geochemical patterns ofgranitoid samples indicate a north-south difference in theirmagma source. The eNd values are lower, eSr values arehigher, and the degree of incompatible element enrichmentof the source region was higher in the south (Jurassic-Cretaceous granitoids of the Qiangtang, excluding the leu-cogabbro intruding Karakul–Mazar rocks in the Muskoldome, Rushan-Pshart, southern Pamirs) than in the north(Kunlun and Karakul–Mazar). Although limited by the lackof data, our study indicates that the Kunlun rocks and theKarakul–Mazar leucogabbro are essentially mantle-derived.The Qiangtang, Rushan-Pshart, and southern Pamiran gran-itoids display a wide range in the eNd versus 87Sr/86Srdiagram; these rocks roughly follow mixing lines between

Figure 12. North-south difference in the first-order lithosphere structure of the Pamirs and Tibet andsupporting arguments. From the Kunlun arc in the north to the southern Qiangtang block in the south, thePamirs and Tibet likely have a dominantly sedimentary middle and lower crust, characterized byKarakul–Mazar–Songpan-Garze accretionary wedge rocks. The crust south of the southern Qiangtangblock is likely granodioritic, reflecting long-lived subduction and arc formation and Cretaceous-Cenozoicunderthrusting of these arc segments underneath the Qiangtang block.


28 of 31

TC4002

depleted mantle and crust with isotopic characteristics of theHigh Himalayan crystalline rocks (Figure 8j). This trendmay suggest that the lithospheric mantle source of theQiangtang to southern Pamiran granitoids were mixed witha continental crustal component of Gondwana origin (com-pare Maheo et al. [2002] for Cretaceous-Cenozoic Kara-koram magmatites). The suggested mixing trend allows for alaterally heterogeneous mantle lithosphere, variable metaso-matism, and magma mixing. It is noteworthy that ourpattern, reflecting pre-India-Asia collision magmatism, isalso reflected by the Cenozoic volcanics of Tibet (Figure 8i).Characteristically, the Cretaceous magmatites of the south-ern Pamirs overlap the array of Sr and Nd initials of theCenozoic volcanics of southern Tibet, and the granitoids,penetrating the Karakul–Mazar–Songpan-Garze rocks ofthe northern Pamirs occupy the array of the Cenozoicvolcanics of northern Tibet.

5. Conclusions

[37] The evolution of magmatic rocks along a north-south traverse across the southern Tien Shan and easternPamirs (Kyrgyzstan, Tajikistan) is studied by geochronol-ogy, geochemistry, a review of the stratigraphy, and aninterpretation of the depositional setting. In the southernTien Shan, the Early Permian Garm–Turkestan-Alai zonegranitoids are part of a mature arc formed north of thesouth Gissar suture. The northern Pamirs and northwesternTibet contain the north facing Kunlun and the south facingJinsha sutures and an intervening, Carboniferous-Triassicsubduction accretion complex, called Karakul–Mazar. It iscorrelated with the Songpan-Garze–Hoh Xi system ofTibet. The Kunlun arc is the juxtaposition of an earlyPaleozoic and a late Paleozoic-Triassic arc. In the Pamirs,the volcanic rocks are �370–320 Ma and most likelyextend into the Triassic. The Markansu zone, rich inultramafic and mafic rocks south of the Kunlun arc, mayconstitute the subduction complex related to the Kunlunsuture. Similarly, the Akzhilga zone within the southernpart of the Karakul–Mazar unit may mark the subductioncomplex related to the Jinsha suture. The cryptic Tanymassuture of the southern northern Pamirs is the Jinsha suture.Posttectonic granitoids, stitching the Karakul–Mazarsubduction accretion system are Late Triassic–EarlyJurassic (�200 Ma) in northwestern Tibet and a massivebatholith (�227 Ma) occurs in the Pamirs.[38] There are striking similarities between the Qiangtang

block in the Pamirs and central Tibet: As in Tibet, theregional structure in the Pamirs is a spectacular Cenozoicanticlinorium (the Muskol and Sares domes). As in Tibet,the fault-bounded basement rocks exposed in the Sares-Muskol domes are interpreted as Karakul–Mazar–Song-pan-Garze rocks. The granitoids intruding the Qiangtangblock have similar ages or ion probe age components of�200–230 Ma in the Pamirs and Tibet. Zircons from agneiss in central Tibet and those from a Pamiran granitoidoverlap closely (age range 425–575 Ma) and represent thepre-late Paleozoic basement of the Qiangtang block;this age group is characteristic for Gondwana crust. The

stratigraphy of the eastern Pshart area (Rushan-Pshartsuture zone) of the Pamirs conforms well to that of theBangong-Nujiang suture zone in the Amdo region ofcentral Tibet, with the addition that a Triassic ocean basinsequence is preserved in the Pamirs. Ophiolite-bearingthrust systems, possibly rooting in the Rushan-Pshartsuture, are likely equivalents to the Lhasa-Amdo thrustsystem in Tibet. Intrusion of arc granitoids into theRushan-Pshart oceanic-basin-arc sequence occurred be-tween 200–160 Ma and all granitoids intruding the south-ern Qiangtang block revealed a Middle Jurassic age group(�170–160 Ma); these intrusions constitute the Rushan-Pshart arc. The Rushan-Pshart suture of the Pamirsconnects to the Bangong-Nujiang suture of Tibet andamalgamated the Lhasa block with the rest of Asia.[39] The area between the Indus-Yarlung suture and

the northern Qiangtang block contains widespread andvoluminous Cretaceous and scattered Cenozoic granitoids.The Cretaceous intrusions record Proterozoic crust thatexperienced multiple Phanerozoic magmatic additions.Their zircons and those from late Miocene xenolithsshow that the most distinct magmatic events wereCambro-Ordovician (�410–575 Ma), Triassic (�210–250 Ma; likely due to subduction along the Jinshasuture), Middle Jurassic (�147–195 Ma; subductionalong Rushan-Pshart –Bangong-Nujiang suture), andmainly Cretaceous. The Cretaceous activity may havebeen episodic, with peaks at �120 Ma and �80 Ma.The mid-Cretaceous event may reflect arc activity in Asiaprior to the accretion of the Karakoram block. TheLate Cretaceous magmatism may be a sign of flat-slabsubduction along the Shyok suture north of the Kohistan-Ladakh arc. Prior to the India-Asia collision, the Pamiranregion between the Indus-Yarlung and Jinsha suturesmust have been a magmatically and structurally thickenedAndean-type margin.[40] Our interpretations suggest a relatively simple crustal

structure for the Pamirs and Tibet. From the Kunlun arc inthe north to the southern Qiangtang block in the south thePamirs and Tibet likely have a dominantly sedimentarymiddle and lower crust, characterized by Karakul –Mazar–Songpan-Garze accretionary wedge rocks. The crustsouth of the southern Qiangtang block is likely of grano-dioritic composition, reflecting long-lived subduction, arcformation and Cretaceous-Cenozoic underthrusting of arcsegments beneath the Qiangtang block.

[41] Acknowledgments. We dedicate this paper to the Soviet geolo-gists who did the primary field mapping in the Pamir Mountains, for whomone reward was the privilege to see these mountains in peaceful times.Without their work, it would be impossible to construct more detailedstudies such as ours. We thank V. I. Dronov and E. J. Leven, whom we hadthe pleasure to meet. U. Hermann did some of the initial U/Pb analysis atSeattle. S. Semiltektin, M. Kornilov, and S. Zamoruyev aided us in the fieldin 1993 and 1996, as did J. Kuhlemann in 1996. P. Kapp provided moststimulating ‘‘in press’’ papers about the central Qiangtang and the Lhasablocks in Tibet. We thank him and an anonymous colleague for stimulatingreviews. Y. Pushkarev contributed the K/Ar analysis. E. Sobel is thankedfor enlightening discussions. L. R. thanks Stanford University colleaguesfor hospitality during visits to make 40Ar/39Ar and SHRIMP U/Pb measure-ments. Deutsche Forschungsgemeinschaft grants Fr 610/11 and Ra 442/12,23 and a ‘‘Jubilaumsstiftung’’ grant from the University of Wurzburg toL. R. funded this work.


29 of 31

TC4002

References

Bazhenov, M. L., and V. S. Burtman (1982), The kine-matics of the Pamir arc, Geotectonics, 16, 288 –301.

Brookfield, M. E. (2000), Geological developmentand Phanerozoic crustal accretion in the westernsegment of the southern Tien Shan (Kyrgyzstan,Uzbekistan, Tajikistan), Tectonophysics, 328, 1 –14.

Bruguier, O., J. R. Lancelot, and J. Malavieille (1997),U-Pb dating on single detrital zircon grains from theTriassic Sonpan-Garze flysch (central China):Provenance and tectonic correlations, Earth Planet.Sci. Lett., 152, 217–231.

Bureau of Geology and Mineral Resources of XinjiangUygur Autonomous Region (1992), Regional geo-logical map of Xinjiang Uygur Autonomous Re-gion, 1:1,500,000, Geol. Publ. House, Beijing.

Burtman, V. S., and P. Molnar (1993), Geologicaland geophysical evidence for deep subduction ofcontinental crust beneath the Pamir, Spec. Pap.Geol. Soc. Am., 281, 1 – 76.

Chang, C. (1996), Geology and tectonics of the Qinghai-Xizang plateau, in Solid Earth Sciences Research inChina, pp. 1 – 153, Science Press, Beijing.

Chen, F., E. Hegner, and W. Todt (2000), Zircon agesand Nd isotopic and chemical compositionsof orthogneisses from the Black forest, Germany:Evidence for a Cambrian magmatic arc, Int. J. EarthSci., 88, 791–802.

Cocherie, A., C. Guerrot, and P. Rossi (1992), Single-zircon dating by step-wise Pb evaporation: Compar-ison with other geochronological techniques appliedto the Hercynian granites of Corsica, France, Chem.Geol., 101, 131–141.

Cowgill, E., A. Yin, T. M. Harrison, and W. Xiao-Feng(2003), Reconstruction of the Altyn Tagh faultbased on U-Pb geochronology: Role of backthrusts, mantle sutures, and heterogeneous crustalstrength in forming the Tibetan Plateau, J. Geophys.Res., 108(B7), 2346, doi:10.1029/2002JB002080.

DeCelles, P. G., G. E. Gehrels, J. Quade, B. LaReau,and M. Spurlin (2000), Tectonic implications ofU/Pb zircon ages of the Himalayan orogenic beltin Nepal, Science, 288, 497–499.

DeGraaff-Surpless, K., S. A. Graham, J. L. Wooden,and M. McWilliams (2002), Detrital zircon prove-nance analysis of the Great Valley Group, Califor-nia: Evolution of an arc-forearc system, Geol. Soc.Am. Bull., 114, 1564–1580.

Dewey, J. F., R. M. Shackleton, C. Chang, and Y. Sun(1988), The tectonic evolution of the Tibetanplateau, Philos. Trans. R. Soc. London, Ser. A,327, 379–413.

Ding, L., P. Kapp, D. Zhong, and W. Deng (2003),Cenozoic volcanism in Tibet: Evidence for a trans-ition from oceanic to continental subduction,J. Petrol., 44, 1833–1865.

Dronov, V. I. (1986), Stratigraphy of the Jurassic andolder volcanogenic sedimentary sequences in theBashgumbez, Irikyak, Sedek, and Tashdzhilga riverbasins (in Russian), Dokl. Akad. Nauk Tadjik SSR,29, 549–553.

Dronov, V. I. (1988), The volcanogenic type of Triassicdeposits in southeastern Pamir (in Russian), Dokl.Akad. Nauk USSR, 303, 437–440.

Ducea, M. N., et al. (2003), Building the Pamirs: Theview from the underside, Geology, 31, 849–852.

Fraser, J. E., M. P. Searle, R. R. Parrish, and S. R.Noble (2001), Chronology of deformation, meta-morphism, and magmatism in the southern Kara-koram Mountains, Geol. Soc. Am. Bull., 113,1443–1455.

Gaetani, M. (1997), The Karakorum block in centralAsia, from Ordovician to Cretaceous, Sediment.Geol., 109, 339–359.

Gehrels, G. E., A. Yin, and X.-F. Wang (2003), Detritalzircon geochronology of the northeastern Tibetanplateau, Geol. Soc. Am. Bull, 115, 881–896.

Girardeau, J., J. C. Marcoux, J. C. Allegre, J. P.Bassoulett, T. Youking, X. Xuchang, Z. Yougon,

and W. Xibin (1984), Tectonic environment andgeodynamic significance of the Neo-CimmerianDonqiao ophiolite, Bangong-Nujiang Zone, Tibet,Nature, 307, 27–31.

Girardeau, J., J. C. Marcoux, E. Fourcade, J. P.Bassoullet, and T. Youking (1985), Xainxa ultrama-fic rocks, central Tibet, China: Tectonic environmentand geodynamic significance, Geology, 13, 330 –333.

Grimmer, J. C., L. Ratschbacher, M. McWilliams,L. Franz, I. Gaitzsch, M. Tichomirowa, B. R.Hacker, and Y. Zhang (2003), When did the ultra-high-pressure rocks reach the surface?:A207Pb/206Pbzircon, 40Ar/39Ar white mica, Si-in-white mica,single-grain provenance study of Dabie Shan syn-orogenic foreland sediments, Chem. Geol., 197,87– 110.

Hacker, B. R., J. L. Mosenfelder, and E. Gnos (1996),Rapid emplacement of the Oman ophiolite: Ther-mal and geochronological constraints, Tectonics,15, 1230–1247.

Hacker, B. R., E. Gnos, L. Ratschbacher, M. Grove,M. McWilliams, W. Jiang, and Z. Wu (2000), Hotand dry xenoliths from the lower crust of Tibet,Science, 287, 2463–2466.

Hacker, B. R., P. Luffi, V. Lutkov, V. Minaev,L. Ratschbacher, M. Ducea, M. McWilliams, andJ. Metcalf (2004), High-pressure processing of asubducted continent beneath the Pamirs, J. Petrol.,in press.

Harris, N. B. W., R. Xu, C. L. Lewis, C. J.Hawkesworth, and Y. Zhang (1988a), Isotopegeochemistry of the 1985 Tibet geotraverse, Lhasato Golmud, Philos. Trans. R. Soc. London, Ser. A,327, 263–285.

Harris, N. B. W., R. Xu, C. L. Lewis, and C. Jin(1988b), Plutonic rocks of the 1985 Tibet geotra-verse, Lhasa to Golmud, Philos. Trans. R. Soc. Lon-don, A327, 145–168.

Hegner, E., H. J. Walter, and M. Satir (1995), Pb-Sr-Nd isotopic compositions and trace element geochemistry of megacrysts and melilitites fromthe Tertiary Urach volcanic field: Source composi-tion of small volume melts under SW Germany,Contrib. Mineral. Petrol., 122, 322–335.

Hildebrand, P. R., S. R. Noble, M. P. Searle, D. J.Waters, and R. R. Parrish (2001), Old origin foran active mountain range: Geology and geochronol-ogy of the eastern Hindu Kush, Pakistan, Geol. Soc.Am. Bull., 113, 625–639.

Hubbard, M. S., E. S. Grew, K. V. Hodges, M. G. Yates,and N. N. Pertsev (1999), Neogene cooling andexhumation of upper-amphibolite facies ‘‘whites-chists’’ in the southwest Pamir Mountains, Tajiki-stan, Tectonophysics, 305, 325–337.

Jolivet, M., M. Brunel, D. Seward, Z. Xu, J. Yang,J. Malavieille, F. Roger, A. Leyreloup, N. Arnaud,andC.Wu (2003), Neogene extension and volcanismin the Kunlun Fault Zone, northern Tibet: Newconstraints on the age of the Kunlun Fault, Tectonics,22(5), 1052, doi:10.1029/2002TC001428.

Kapp, P., A. Yin, C. E. Manning, M. Murphy, T. M.Harrison, M. Spurlin, L. Ding, X.-G. Deng, andC.-M. Wu (2000), Blueschist-bearing metamorphiccore complexes in the Qiangtang block reveal deepcrustal structure of northern Tibet, Geology, 28,19– 22.

Kapp, P., A. Yin, C. E. Manning, T. M. Harrison, M. H.Taylor, and L. Ding (2003a), Tectonic evolution ofthe early Mesozoic blueschist-bearing Qiangtangmetamorphic belt, central Tibet, Tectonics, 22(4),1043, doi:10.1029/2002TC001383.

Kapp, P., M. A. Murphy, A. Yin, T. M. Harrison,L. Ding, and J. Guo (2003b), Mesozoic andCenozoic tectonic evolution of the Shiquanhearea of western Tibet, Tectonics, 22(4), 1029,doi:10.1029/2001TC001332.

Kober, B. (1986), Whole grain evaporation for207Pb/206Pb-age investigations on single zircons

using a double-filament ion source, Contrib.Mineral. Petrol., 93, 482–490.

Kober, B. (1987), Single-zircon evaporation combinedwith Pb+ emitter bedding for 207Pb/206Pb-ageinvestigations using thermal ion mass spectrometry,and implications to zirconology, Contrib. Mineral.

Petrol., 96, 63–71.Kravchenko, K. N. (1979), Tectonic evolution of the

Tien Shan, Pamir and Karakoram, in Geodynamicsof Pakistan, edited by A. Farah and K. A. DeJong,pp. 25 – 40, Geol. Surv. of Pakistan, Peshawar,Pakistan.

Kroner, A., and E. Hegner (1998), Geochemistry, singlezircon ages and Sm-Nd systematics of granitoidrocks from the Gory Sowie (Owl Mts), Polish WestSudetes: Evidence for early Palaeozoic arc-relatedplutonism, J. Geol. Soc. London, 155, 711 –724.

Lee, J., B. R. Hacker, W. S. Dinklage, Y. Wang,P. Gans, A. Calvert, J. Wan, W. Chen, A. E. Blythe,and W. McClelland (2000), Evolution of the Kang-mar dome, southern Tibet: Structural, petrologic,and thermochronologic constraints, Tectonics, 19,872 –895.

Leven, E. J. (1995), Permian and Triassic of theRushan-Pshart zone (Pamir), Riv. Ital. Paleontol.Stratigr., 101, 3 –16.

Ludwig, K. R. (1999), User’s Manual for Isoplot/Exversion 2.01, Spec. Publ. 1a, 47 pp., BerkeleyGeochron. Cent., Berkeley, Calif.

Lutkov, V. S. (2003), Petrochemical evolution andgenesis of potassium pyroxenite-eclogite-granuliteassociations in mantle and crustal xenoliths in Neo-gene fergusites of the southern Pamir, Tajikistan(in Russian), Geochimia, 2003/3, 254–265.

Maheo, G., S. Guillot, J. Blichert-Toft, Y. Rolland, andA. Pecher (2002), A slab breakoff model for theNeogene thermal evolution of South Karakoramand South Tibet, Earth Planet. Sci. Lett., 195,45 –58.

Map Compiler Group (1986), Geological map of theNujiang, Lancang, and Jinsha River area,1:1,000,000, Geol. Publ. House, Beijing.

Matte, P., P. Tapponnier, N. Arnault, L. Bourjot, J. P.Avouac, P. Vidal, Q. Liu, Y. Pan, and Y. Wang(1996), Tectonics of Western Tibet, between theTarim and the Indus, Earth Planet. Sci. Lett., 142,311 –330.

Miller, C., R. Schuster, U. Klotzli, W. Frank, andB. Grasemann (2000), Late Cretaceous-Tertiarymagmatic and tectonic events in the Transhimalayabatholith (Kailas area, SW Tibet), Schweiz. Mineral.Petrogr. Mitt., 80, 1 – 20.

Murphy, M. A., A. Yin, T. M. Garrison, S. B. Durr,Z. Chen, F. Ryerson, W. Kidd, X. Wang, andX. Zhou (1997), Significant crustal shortening insouth-central Tibet prior to the Indo-Asian collision,Geology, 25, 719–722.

Paces, J. B., and J. D. Miller (1993), Precise U-Pb agesof Duluth complex and related mafic intrusions,northeastern Minnesota: Geochronological insightsto physical, petrologic, paleomagnetic, and tectono-magnetic processes associated with the 1.1 Gamidcontinent rift system, J. Geophys. Res., 98,13,997–14,013.

Pashkov, B. R., and V. I. Budanov (1990), The tectonicsof the intersection between the southeastern andsouthwestern Pamirs (in Russian), Geotektonika,3, 70 –79.

Pearce, J. A., and H. Mei (1988), Volcanic rocks of the1985 Tibet Geotraverse: Lhasa to Golmud, Philos.Trans. R. Soc. London, Ser. A, 327, 169–201.

Pupin, J. P. (1980), Zircon and granite petrology,Contrib. Mineral. Petrol., 73, 207–220.

Ratschbacher, L., W. Frisch, C. C. Cheng, and G. Pan(1996), Cenozoic deformation, rotation, and stressin Eastern Tibet and Western Sichuan, in The

Tectonic Evolution of Asia, edited by A. Yin andT. M. Harrison, pp. 227 – 249, Cambridge Univ.Press, New York.


30 of 31

TC4002

Ratschbacher, L., B. R. Hacker, A. Calvert, L. E. Webb,J. C. Grimmer, M. McWilliams, T. Ireland,S. Dong, and J. Hu (2003), Tectonics of the QinlingBelt (central China): Tectonostratigraphy, geochro-nology, and deformation history, Tectonophysics,366, 1 –53.

Roddick, J. C. (1978), The application of isochrondiagrams in 40Ar-39Ar dating: A discussion, EarthPlanet. Sci. Lett., 41, 233 –244.

Roger, F., and S. Calassou (1997), Gechronologie U-Pbsur zircons et geochimie (Pb, Sr et Nd) du socle dela chaıbe de Songpan-Garze (Chine), C. R. Acad.Sci. Paris, Ser. IIa, 324, 819 –826.

Roger, F., N. Arnaud, S. Gilder, P. Tapponnier,M. Jolivet, M. Brunel, J. Malavieille, Z. Xu,and J. Yang (2003), Geochronological and geochem-ical constraints on Mesozoic suturing in eastcentral Tibet, Tectonics, 22(4), 1037, doi:10.1029/2002TC001466.

Rolland, Y., A. Pecher, and C. Picard (2000), Mid-Cretaceous back-arc formation and arc evolutionalong the Asian margin: The Shyok Zone zone innorthern Ladakh (NW Himalaya), Tectonophysics,325, 145–173.

Ruzhentsev, S. V., and V. A. Shol’’man (1982),The Pamirs, in Alpine Structural Elements:Carpathian-Balkan-Caucasus-Pamir OrogenicZone, edited by M. Mahel, pp. 115–130, VEDA,Bratislava.

Scharer, U., R. Xu, and C. J. Allegre (1986), U-(Th)-Pbsystematics and ages of Himalayan leucogra-nites, south Tibet, Earth Planet. Sci. Lett., 77,35– 48.

Sengor, A. M. C., B. A. Natal’in, and V. S. Burtman(1993), Evolution of the Altaids tectonic collageand Paleozoic crustal growth in Eurasia, Nature,364, 299–307.

Solov’ev, S. G. (1998), Rare-earth and other trace ele-ments in rocks from the W-bearing magmatic com-plexes of the southern Tien Shan, Geochem. Int.,36, 1252–1268.

Stacey, J. S., and J. D. Kramers (1975), Approximationof terrestrial lead isotope evolution by a two-stagemodel, Earth Planet. Sci. Lett., 26, 207–221.

Vlasov, N. G., Yu. A. Dyakov, and E. S. Cherev (Eds.)(1991), Geological map of the Tajik SSR and adja-cent territories, 1:500,000, Vsesojuznoi Geol. Inst.Leningrad, Saint Petersburg.

Wendt, J. I., and C. Carl (1991), The statistical distribu-tion of the mean squared weighted deviation, Chem.Geol., 86, 275 –285.

Xiao, W., B. F. Windley, J. Hao, and J. Li (2002a),Arc-ophiolite obduction in the Western KunlunRange (China): Implications for the Palaeozoicevolution of central Asia, J. Geol. Soc. London,159, 517–528.

Xiao, W. J., B. F. Windley, H. L. Chen, G. Ch. Zhang,and J. L. Li (2002b), Carboniferous-Triassic subduc-tion and accretion in the western Kunlun, China:Implications for the collisional and accretionary tec-tonics of the northern Tibet plateau, Geology, 30,295–298.

Xu, R., U. Scharer, and C. J. Allegre (1984), Mag-matism and metamorphism in the Lhasa block(Tibet): A geochronological study, J. Geol., 93,41– 57.

Yin, A., and T. M. Harrison (2000), Geologic evolutionof the Himalayan-Tibetan orogen, Annu. Rev. EarthPlanet. Sci., 28, 211 –280.

Zhou, M.-F., J. Malpas, P. T. Robinson, and P. H.Reynolds (1997), The dynamothermal aureole ofthe Donqiao ophiolite (northern Tibet), Can. J.Earth Sci., 34, 59–65.

Zonenshain, L. P., M. I. Kuzmin, and L. M. Natapov(1990), Central Asian foldbelt, western part: CentralKazakhstan and Tien Shan, in Geology of theUSSR: A Plate-Tectonic Synthesis, Geodyn. Ser.,vol. 21, edited by B. M. Page, pp. 55–72, AGU,Washington. D. C.

��F. Chen, W. Frisch, M. Schwab, and W. Siebel,

Institut fur Geowissenschaften, Universitat Tubingen,72076 Tubingen, Germany.

V. Lutkov and V. Minaev, Geological Institute ofthe Tajik Academy of Science, 734063 Dushanbe,Tajikistan.

M. McWilliams, Geological and EnvironmentalSciences, Stanford University, Stanford, CA 94305-2115, USA.

B. Nelson, Department of Geological Sciences,University of Washington, Seattle, WA 98195, USA.

L. Ratschbacher and K. Stanek, Institut furGeowissenschaften, Technische Universitat Bergakade-mie Freiberg, 09599 Freiberg, Germany. ([email protected])

J. L. Wooden, U.S. Geological Survey, 345Middlefield Road, Menlo Park, CA 90425, USA.


31 of 31

TC4002

Assembly of the Pamirs: Age and origin of magmatic belts from the southern Tien Shan to the southern Pamirs and their relation to Tibet

Documents