accretionary tectonics of the western kunlun orogen, china...

[The Journal of Geology, 2005, volume 113, p. 687–705] � 2005 by The University of Chicago. All rights reserved. 0022-1376/2005/11306-0005$15.00

687

Accretionary Tectonics of the Western Kunlun Orogen, China: APaleozoic–Early Mesozoic, Long-Lived Active Continental Margin

with Implications for the Growth of Southern Eurasia

W. J. Xiao, B. F. Windley,1 D. Y. Liu,2 P. Jian,2 C. Z. Liu, C. Yuan,3 and M. Sun4

State Key Lab of Lithospheric Evolution, Institute of Geology and Geophysics, ChineseAcademy of Sciences, Beijing 100029, China

(e-mail: [email protected])

A B S T R A C T

Our new SHRIMP U-Pb zircon ages from the Western Kunlun Orogen allow us to constrain the history of an activecontinental margin developed on the southern boundary of the Tarim block from the Ordovician to the Triassic. A

-Ma dacite from Yixieke extrusive rocks that contain -Ma reheated zircons is interpreted as an intra-492 � 7 220 � 5oceanic arc complex that accreted to the Tarim block. The Yirba granodiorite has a continental arc geochemicalsignature, a -Ma U-Pb crystallization age, and -Ma inherited zircons. It formed during the first, early471 � 5 491 � 3Paleozoic stage of an active continental margin arc that was juxtaposed to the south against the Kudi high-gradegneiss complex, the Buziwan ophiolite, and the Yixieke volcanic and sedimentary rocks. Zircons from a paragneissin the Kudi gneiss complex range in age from to Ma; the oldest reflect protolith ages of a gneissic398 � 12 1345 � 31continental block (incorporated into the trench), and the youngest may represent the age of a refoliated high-gradefabric created during accretion. The Buziwan ophiolite occupies a thrust sheet tectonically overlying the Kudi gneisscomplex. A leuco-gabbro pegmatite, with a zircon age of Ma and ca. 490-Ma inherited zircons, and the North403 � 7Kudi granite, with a zircon age of Ma, were emplaced during the second mid-Paleozoic stage of the active408 � 7continental margin. The Akarz subduction-related granite that has a -Ma zircon crystallization age formed214 � 1during the final, early Mesozoic stage of the active margin. The long-lasting active continental margin in the westernKunlun forms a key, well-documented section of the Andean-type margin that extends from the Caucasus to theQinling.

Online enhancements: color versions of figures 3 and 4.

Introduction

The Western Kunlun Orogen (WKO), located alongthe northern periphery of the Tibetan plateau, is a1000-km-long mountain belt extending from thePamir syntaxis in the west to the Altyn–East Kun-lun Orogen in the east (fig. 1). Its Paleozoic to earlyMesozoic orogenic history is of considerable im-portance for the reconstruction of paleo-Asia be-cause it occupies a key tectonic position between

Manuscript received June 3, 2004; accepted January 28, 2005.1 Department of Geology, University of Leicester, Leicester

LE1 7RH, United Kingdom.2 SHRIMP Laboratory Beijing, Institute of Geology, Chinese

Academy of Geological Sciences, Beijing 100037, China.3 Guangzhou Institute of Geochemistry, Chinese Academy

of Sciences, Guangdong, Guangzhou 510640, China.4 Department of Earth Sciences, University of Hong Kong,

Hong Kong SAR, China.

the Tarim block to the north and the Tethyan do-main to the south, and it sheds light on the tectonicarchitecture of the Tibetan plateau immediately toits south. However, the Paleozoic tectonic evolu-tion of the WKO has been contentious (see Jiang etal. 1992; Yao and Hsu 1994; Matte et al. 1996; Mat-tern et al. 1996; Pan 1996; Yang et al. 1996; Yuan1999; Mattern and Schneider 2000; Xiao et al.2002a, 2002b, 2003a; Yuan et al. 2002a, 2002b) be-cause of the lack of reliable isotopic ages of certainkey tectonic units, in particular the Kudi ophiolite,which is situated in a possibly middle Paleozoic(Akaz) suture (Matte et al. 1996; Mattern andSchneider 2000; Pan 1996; Sobel and Arnaud 1999;Cowgill et al. 2003) that extends eastward to theLapeiquan suture in the Altyn–East Kunlun (fig. 1).

688 W . J . X I A O E T A L .

Figure 1. Schematic map of the Kunlun-Qinling ranges and adjacent regions showing the Paleozoic–early Mesozoicactive continental margin, marked by crosses (modified after Sobel and Arnaud 1999; Xiao et al. 2002b; Cowgill etal. 2003; Roger et al. 2003); . Area of figure 2 is indicated by the box.F p fault

This suture is a remnant of the paleo-Tethyanocean in central Asia (Pan 1996; Sobel and Arnaud1999). Knowledge of the isotopic ages of the Kudiophiolite and spatially associated granites is essen-tial to understand the tectonic evolution of paleo-Tethys (Pan 1996; Zhou and Graham 1996; Wang1997; Sobel and Arnaud 1999; Yue and Liou 1999;Yuan et al. 2002a, 2002b).

In this study, we report new zircon U-Pb SHRIMPages from the Kudi ophiolite and two key granitesthat have well-established structural age relation-ships with the ophiolitic rocks, in order to docu-ment specific stages in the crustal evolution of theWKO. We have integrated all these SHRIMP zirconages with our recent structural, geochemical, pet-rological, geochronological, and tectonic data fromthe WKO (Yuan 1999; Xiao et al. 2002a, 2002b,2003a; Yuan et al. 2002a, 2002b). We use publishedgeochronological (Xu et al. 1994, 1996; Bi et al.1999; Sobel and Arnaud 1999; Cowgill et al. 2003,2004a, 2004b; Gehrels et al. 2003a, 2003b) and geo-chemical data (Deng 1995; Zhang et al. 1996; Yuan1999; Jiang et al. 2002; Wang et al. 2002; Yuan etal. 2002a, 2002b, 2004) to reevaluate the origin ofthe plutonic and volcanic rocks and their interre-

lationships within a supra-subduction zone setting.Significant advances in understanding the WKOhave been made in the last several years, includingimprovements in understanding of the timing andpatterns of deformation (Zhou et al. 2000; Xiao etal. 2002a, 2002b, 2003a), the origin of the graniticplutons (Yuan et al. 2002a, 2002b), the nature ofthe Ordovician arc (Sobel and Arnaud 1999; Cow-gill et al. 2003), and the accretionary tectonics inthe southern part of the WKO (Xiao et al. 2002a,2002b, 2003a). This article summarizes the broadgeological environments and structures of theWKO and presents a new model to explain the ma-jor tectonic events within the context of the ac-cretionary orogens of southern central Asia.

Regional Geology

The WKO is divisible into the North Kunlun, SouthKunlun, and Tianshuihai domains, separated by theAkaz and Mazar-Kangxiwar faults, respectively (fig.2). The North Kunlun domain represents the base-ment of the Tarim block, and the South Kunlundomain is mainly composed of various tectonic as-semblages, including the Buziwan (Kudi) ophiolite,

Journal of Geology K U N L U N A C C R E T I O N A R Y T E C T O N I C S 689

Figure 2. Schematic tectonic map of the Kunlun ranges and adjacent regions showing the position of the Kudi areaand major pluton age distribution (modified after Xu et al. 1996; Sobel and Arnaud 1999; Xiao et al. 2002b; Cowgillet al. 2003; Gehrels et al. 2003a, 2003b). Area of figure 3 is marked by a box. fault.MKF p Mazar-Kangxiwar

the Yixieke arc, the Kudi gneiss, and the Xianan-qiao arc (Xiao et al. 2002a, 2002b, 2003a), the mainsubjects of this article. The Tianshuihai domain ismainly a huge accretionary wedge that records alate Paleozoic–early Mesozoic subduction-relatedorogenic process (Xiao et al. 2002a, 2003a). Formore detailed stratigraphy and structures, readersare referred to Matte et al. (1996), Pan (1996), Mat-tern and Schneider (2000), and Wittlinger et al.(2004). Figure 3 shows the main rock units, de-scribed below mainly in time sequence. In this ar-ticle, we use the new geological time scale (Grad-stein et al. 2004) to analyze tectonic stages orgeological events.

The northernmost unit, near Akaz Daban (pass;fig. 3), comprises the basement of the Tarim block(Matte et al. 1996; Mattern and Schneider 2000; Pan1996), which is composed of Proterozoic gneisses,schists, migmatites, stromatolite-bearing lime-stones, clastics, and cherts, overlain by Sinian con-glomerates, tillites, clastics, and carbonates (fig. 3).On the southern side of the gneisses, a fault (figs.3, 4) marks the southern boundary of exposed rocksof the Tarim block. South of the fault, greenschistsform 20- to 1100-m-thick layers within an 800-m–1.5-km-thick succession (the Sailajaz Group ofYuan et al. 2004) of stromatolite- and crinoid-bearing Neoproterozoic to Early Cambrian beddedmarbles that are cut by 30-cm-thick basic dikes (ofunknown age) and contain layers of slate at least100 m thick. Yuan et al. (2004) show that the green-schists are large-ion lithophile and light rare earthelement (LREE) enriched, have relatively high Th/Nb and La/Nb ratios, and within-plate character-

istics, together with a wide range of Zr/P2O5 ratiosand a concomitant increase in Th/Nb ratios, sug-gesting crustal contamination. We interpret theSailajaz belt as a whole as a remnant of a conti-nental rift shelf in which shelf limestones are cutby rift dikes and are interbedded with mudstonesand rift lavas. The belt is situated on the southernmargin of the Tarim block and represents a half-graben-shelf succession that bordered an ocean tothe south. The Sailajaz Group rests unconformablyon the Middle Proterozoic Tarim basement rocksthat consist of metamorphosed clastic, carbonate,and volcanic rocks intruded by a 2.2-Ga graniticpluton (Pan 1996; Yuan et al. 2004); this confirmsa continental basement beneath the shelf riftsuccession.

South of the Sailajaz belt (fig. 3), Zhang (1997)reported Ordovician crinoids in marbles, which, to-gether with greenschists, were interpreted to befragments of a volcanic seamount in an accretion-ary wedge (Xiao et al. 2002b). This means that theremust have been ocean floor on which the seamountwas built, and therefore we suggest that the south-ern side of the Sailajaz belt marks a new positionfor a suture zone along the Akaz fault (fig. 3). Clo-sure of the ocean resulted in the Akaz suture, whichseparates continental and continental-margin rocksof the Tarim craton (North Kunlun domain) to thenorth from accreted oceanic-derived ophiolitic andarc rocks of the orogen to the south (South Kunlundomain; fig. 3).

On the southern side of the accretionary wedgeis the Yirba arc-related, lineated, and foliated grano-diorite (fig. 3), which has U/Pb zircon ages of

Figure 3. Tectonic map of the Kudi area, Western Kunlun Orogen (based on our field data, incorporated with thoseof Matte et al. 1996; Mattern and Schneider 2000; XBGMR 1993; and Yin and Bian 1995). A color version of thisfigure is available in the online edition of the Journal of Geology.


Figure 4. Cross section A–A′ along the line shown in figure 3. Key as in figure 3, except as indicated. KP ppoint where section direction changes. See text for discussion. A color version of this figure is available in theknick

online edition of the Journal of Geology.

and Ma (Yuan et al. 2002a, 2002b).491 � 3 471 � 5This forms the older active margin developed alongthe southern margin of the Tarim block (Xiao et al.2002b).

Kudi Arc-Ophiolite-AccretionaryWedge Assemblage

An important assemblage of the South Kunlun do-main was formerly known as the “Kudi ophiolitesuite” (Matte et al. 1996; Pan 1996; Mattern andSchneider 2000) and is mainly composed of ultra-mafic rocks and volcanic and volcaniclastic rocksand forearc sediments of the former Yishak Group(Xiao et al. 2002a). The ultramafic rocks includea southward-thrusted ultramafic-gabbroic klippeabout 3 km long and 1.5 km wide in the Buziwanvalley and 2–3-m-thick tectonic slices now imbri-cated within the Kudi gneiss in unnamed valleyssouth of Kudi. The volcanic and volcaniclasticrocks and forearc sediments crop out mainly to thenorth of the ultramafic-gabbroic rocks, with a verygood section in the Yixieke Valley (fig. 3). Becausethere are various rock types with different struc-tures and geochemical signatures in this formerophiolite suite, we use the term “Buziwan (or Kudi)ophiolite” of Xiao et al. (2002a, 2003a) and Wanget al. (2001, 2002) to encompass ultramafic and gab-broic rocks in the Buziwan valley and “Yixiekeforearc” for volcanic and volcaniclastic rocks in theYixieke valley. Accordingly, we describe the maincomponents of the accreted ophiolite-arc rock as-semblage under these locality names.

The Yixieke extrusive rocks consist of massiveand pillowed basalts, boninites, tuffs, welded an-desitic breccias and agglomerates, and calc-alkaline

lavas intruded by uncommon dolerite dikes; thereare no sheeted dikes (Matte et al. 1996; Pan 1996;Sobel and Arnaud 1999; Yuan 1999). Three groupsof tholeiitic lavas were recognized by Wang et al.(2002). Group 1 basalts have LREE-enriched,chondrite-normalized REE patterns and Cr-Y val-ues typical of island arc tholeiites and La/Sm-TiO2

ratios similar to those of the Mariana arc. Group 2basalts have low K contents, marked negative Nbanomalies, flat to slightly LREE-depleted REE pat-terns typical of transitional midocean ridge basalt(T-MORB), Cr-Y values akin to those of island arcbasalts, La/Sm-TiO2 ratios comparable to those ofthe Lau back-arc basin, and a supra-subductionzone signature in Hf/3-Th-Nb/16 space. Buziwangabbros and diabase dikes that transect extrusiverocks have geochemical signatures similar to thoseof the group 2 tholeiites. Our dacitic sample 123occurs within the group 2 basalts near the bottomof the Yixieke volcanic pile. Group 3 basalts arecharacterized by high Cr and low Y indicative of arelatively high degree of partial melting derivedfrom a depleted mantle source, have low TiO2

(0.16–0.38 wt%), high Mg ( ), normalMg# p 62–72midocean ridge basalt (N-MORB)–normalized traceelement patterns, La/Sm-TiO2 ratios similar tothose of forearc boninites in the Izu-Bonin-Marianaarc, and U-shaped REE patterns that are typical ofmany boninites (Yuan 1999). The three groups ofYixieke basalts are overlain by calc-alkaline lavasthat include basaltic andesites, andesites, and vol-caniclastic rocks such as tuffs, welded andesiticbreccias, and agglomerates.

However, a recent detailed field and geochemicalstudy along the Yixieke valley has provided a newtectonic framework in which five units are recog-

692 W . J . X I A O E T A L .

nized in the central part of the Yixieke extrusiverocks (Yuan et al. 2005). The lowest, unit A, hasN-MORB-like geochemical characteristics, and theoverlying unit B suggests an enriched midoceanridge basalt (E-MORB) affinity. The geochemistryof overlying units C and D reflects the involvementof a slab-derived component, possibly produced bypartial melting of a mantle source modified bymelt-rock interaction during upwelling of E-MORBmantle. The uppermost unit, E, shows geochemicalfeatures that can be explained by mixing of a MORBcomponent with melts from subducted sediments.It is noteworthy that tholeiitic basalts have initial143Nd/144Nd and 87Sr/86Sr isotopic ratios rangingfrom 0.5122 to 0.5123 ( ) and from� p 5.8–8.0Nd

0.7037 to 0.7050, respectively. Boninitic lavas arecharacterized by high Al2O3/TiO2 values of 120, lowTiO2 and Al2O3 values, high SiO2 and Na2O values,LREE-enriched patterns ( ), and(La/Yb)N p 1.5–2.0�Nd values lower than 3.0. These ratios and the dis-tribution of major- and trace-element data point toan origin in an incipient oceanic arc created by pos-sible mixing of fertile oceanic island basalt, de-pleted subarc mantle, and fluids derived from a sub-ducted slab (Yuan 1999). The geochemical datashow that the rocks are akin to evolved boninitesof the Mariana forearc (Yuan 1999; Wang et al. 2002;Xiao et al. 2002b). Such boninites are typical of asupra-subduction zone environment, where magmageneration is strongly influenced by aqueous fluids(Hickey and Frey 1982).

The volcanic rocks are overlain by a 1500-m-thick succession of Yixieke turbidites (fig. 3) thatincludes ophiolite-derived debris flows, tuffaceousand andesitic sandstones, and radiolarian cherts(Wang 1983; Pan 1996). The Yixieke turbidites aresubdivided into lower and upper parts, with a tec-tonic contact between them (Wang 1983; Jiang etal. 1992; Mattern and Schneider 2000; Xiao et al.2002b). A maximum age for the sedimentation isindicated by Late Ordovician–Silurian radiolaria inthe lowermost turbidites. Petrochemical data of thelower turbidites suggest an origin in a forearc basin(Fang 1998; Fang et al. 1998); this is consistent withthe predominantly forearc nature of the underlyingvolcanic rocks.

Above the Late Ordovician–Silurian lower tur-bidites is a thrust, above which are the youngerupper turbidites that contain radiolaria of LateDevonian–Early Carboniferous (Fang 1998; Zhou etal. 2000) and possibly Carboniferous-Permian age(Mattern and Schneider 2000). The thrust sheet iscomposed of imbricated turbidites with second-order thrust faults (Mattern and Schneider 2000;Xiao et al. 2002b, 2003a). We interpret these post-

thrust turbidites as clastic debris deposited in aforearc basin that underwent late thrusting.

The Buziwan ultramafic-gabbroic rocks make upa 3-km-thick slab that contains sheared basal ser-pentinite, layered/foliated chromite-bearing du-nite, harzburgite, clinopyroxenite, and gabbro. Themain body of unserpentinized dunite contains lay-ers of clinopyroxenite and hornblendite. Harzbur-gites and dunites are traversed by veins of olivine-orthopyroxene, clinopyroxenite, and asbestos(Wang et al. 2001, 2002), and gabbros are cut bydikes of gabbro (Jiang et al. 1992). The meaning ofa whole-rock mineral Sm-Nd isochron age of

Ma on dunite, harzburgite, gabbro, and651 � 53plagioclase from the gabbro is uncertain (Ding etal. 1996). The intrusive Akarz early Mesozoicgranodiorite (zircon age of 212–213 Ma; Yuan et al.2002a, 2002b) contains three lenses up to 100 mwide of dunite and gabbro (fig. 4). The main ultra-mafic slab has been thrust southward over the Kudigneiss complex (fig. 4).

Xiao et al. (2002b, 2003a) interpreted the Buzi-wan ophiolite as a substrate of Buziwan oceanfloor overlain by the Ordovician-Silurian supra-subduction Yixieke arc, in turn overlain by a LateOrdovician–Silurian turbiditic forearc basin. How-ever, the isotopic age of the ophiolite is not known,and this uncertainty has led to contrasting specu-lations on its age: late Neoproterozoic (Wang 1983),Proterozoic–early Paleozoic (Matte et al. 1996; Pan1996), and late Paleozoic (Jiang et al. 1992; Yao andHsu 1994; Yang et al. 1996; Yin and Harrison 2000).

The Kudi gneiss complex (figs. 1–3) forms themain ridge of the WKO. It is composed of horn-blende/biotite gneisses that contain minor lensesof schist, marble, phyllite, quartzite, and amphib-olite that are cut by two generations of discordantamphibolite dikes, which in the second generationare subhorizontal and undeformed. The gneissescontain a 10-m2 body of anorthosite (Zhou et al.2001). The gneiss complex has been interpreted asa Proterozoic microcontinent derived from the Ta-rim block (XBGMR 1993; Ding et al. 1996; Pan1996) and as a metamorphosed Paleozoic accretion-subduction complex (Sengor and Okurogullari1991; Sengor and Natal’in 1996; Zhou et al. 2000).The 40Ar/39Ar dates on hornblende ( Ma)452 � 5and biotite ( Ma) from the gneisses, in re-428 � 2lation to kinematic indicators, suggest that theywere affected by Late Ordovician–Early Silurian lo-cal ductile shearing (Matte et al. 1996; Zhou et al.2000).

The North Kudi granite (Matte et al. 1996; Mat-tern and Schneider 2000; Jiang et al. 2002; Yuan etal. 2002a) intrudes the Kudi gneiss complex (Zhou


et al. 2000) and now is in fault contact with theYixieke volcanic rocks (fig. 3). It has high d18O(11.6%) and high 87Sr/86Sr ratios (I pSr

) but slightly lower �Nd(t) values0.7097–0.7119(�3.8 to 1.4; Jiang et al. 2002) than the volcanicrocks from the Yixieke arc, which have �Nd(t) valuesfrom 1.4 to 4.4 and 87Sr/86Sr ratios I pSr

(Deng 1995). These relations suggest0.7054–0.7069that the source region of the granite involved sub-ducted oceanic crustal sediment. The granite hasU/Pb zircon ages of 380.0 �1.9/�0.7 Ma (Xu et al.1994) and Ma (Yuan 1999; Yuan et al.405 � 22002b). To get a more precise age and test the twodifferent ages, we conducted a new analysis usingSHRIMP zircon dating.

The Buziwan main dunite body and the Kudigneiss complex were stitched by the Akarz graniticpluton, which has 40Ar/39Ar ages of and180 � 10

Ma (Xu et al. 1994) and a zircon U-Pb221 � 6.6age of Ma (Yuan et al. 2002a, 2002b).214 � 1

Analytical Procedures

Locations from which the samples were collectedare shown in figure 3. Zircons were separated usingconventional heavy-liquid and magnetic tech-niques. Representative zircons were hand-pickedand, together with several examples of standard zir-con TEM from the Research School of Earth Sci-ences (RSES), Australian National University,mounted in epoxy resin and sectioned approxi-mately in half, and the mount surfaces were pol-ished to expose the grain interiors and then goldcoated.

Zircons were analyzed at the Chinese GeologicalAcademy of Sciences using SHRIMP II. TheSHRIMP data have been reduced according to themethod of Williams and Claesson (1987), Williams(1992), Williams et al. (1996), Compston et al.(1984, 1992), and Huang et al. (2004). Interelementfractionation was estimated relative to the RSESstandard zircon TEM (417 Ma). The U, Th, and Pbconcentrations were determined relative to thosemeasured in the standard zircon SL13, which hasa U concentration of 238 ppm and an age of 572Ma (Claoue-Long et al. 1995). Corrections for com-mon Pb were made using the measured 204Pb/206Pbratios. Because of the small amount of 207Pb formedin young (i.e., !1000 Ma) zircons, which results inlow count rates and high analytical uncertainties,the determination of the ages for young zircons hasto be based primarily on their 206Pb/238U ratios(Compston et al. 1992). Uncertainties in the iso-topic ratios and ages in the data table (table 1) andin the error ellipses in the plotted data are reported

at a 1j level, but the final ages on pooled data setsare all 206Pb/238U ages reported as weighted meansat 95% confidence level. All age calculations andstatistical assessments of the data have been madewith the geochronological statistical softwarepackages ISOPLT/EX (version 2.00) and SQUID 1.0of Ludwig (1999, 2001).

SHRIMP U-Pb Geochronology

Yixieke Volcanic Rocks (Sample 123). We sampledthis dacite from a 5–10-m-thick flow within basal-tic lavas at the southernmost end of the main bodyof extrusive rocks, which is located 50 m south ofthe bridge just north of the North Kudi pluton, inorder to constrain the maximum age of the ophi-olitic lavas and to establish their isotopic age re-lationship with the Buziwan ultramafic-gabbroicrocks. This is one of five dacitic flows near the baseof the extrusives. In the sample, plagioclase dom-inates over alkali feldspar in a 2 : 1 ratio, which istypical of dacites.

Fifteen analyses of zircons yielded U/Pb ages,with a number of analyses showing concordance(fig. 5a). The major analyses plot as a group strad-dling the concordia and give a weighted mean 206Pb/238U age of Ma ( , ), as492 � 9 n p 8 MSWD p 0.96these data are located along or near the concordiaand show little variance. These zircons are mainlyclear and elongate with bipyramidal terminations,and some are fragments of grains with pyramidalterminations. A smaller group gives a weightedmean 206Pb/238U age of Ma ( ,220 � 5 n p 5

; fig. 5a; table 1). These grains areMSWD p 1.16equant and have prismatic magmatic shapes. Be-cause these data are located along or near the con-cordia and have similar error ellipses (fig. 5a; table1), we accept the latter as a concordant date.

Buziwan Pegmatite (Sample 120). We sampled thishornblende leuco-gabbro pegmatite dike in Buzi-wan Valley (fig. 3) below the chromite mine (fig.4), where gabbros are cut by gabbro dikes (Jiang etal. 1992). Thirteen analyses of zircons yielded U/Pb ages, with a number of analyses showing con-cordance, as these data are located along or nearthe concordia and show little variance (fig. 5b).Most analyses plot as a group straddling the con-cordia and give a weighted mean 206Pb/238U age of

Ma ( , ). The zircons403 � 7 n p 11 MSWD p 0.47that yield the 403 Ma age are clean, thin, and long,and nearly all are prismatic; we interpret 403 Maas the age of formation of the pegmatite dike. Twoanalyses (11.1 and 12.1 in table 1) are statisticaloutliers from this group and have an age of a ca.490 Ma; these zircons are equant, with pyramidal

Table 1. Summary of U-Th-Pb SHRIMP data on zircons from the Kudi ophiolite and associated granites

Grain,spot

U(ppm)

Th(ppm)

232Th/238U

206 ∗Pb(ppm)

%206Pbc

206Pb/238Uages (Ma)

/207 ∗ 206 ∗Pb Pb /235U207 ∗Pb /238U206 ∗Pb

Value Error (%) Value Error (%) Value Error (%)

Sample 119, Arkarz pluton (36�48.258�N, 76�56.351�E):1.1 547 298 .56 16.1 1.10 215 � 6 .0620 6.0 .290 6.6 .0339 2.62.1 499 286 .59 14.9 1.67 216 � 6 .0696 14 .327 14 .0341 2.73.1 495 250 .52 14.3 .98 211 � 6 .0710 13 .326 13 .0333 2.74.1 769 353 .47 23.3 .60 222 � 6 .0551 6.4 .266 6.9 .0351 2.65.1 312 155 .51 9.15 3.81 209 � 6 .0600 17 .271 17 .0329 2.86.1 489 298 .63 14.6 1.66 217 � 6 .0555 9.6 .262 9.9 .0342 2.77.1 410 178 .45 12.2 2.28 215 � 6 .0449 16 .210 16 .0339 2.88.1 552 298 .56 16.1 .87 213 � 6 .0482 12 .223 12 .0336 2.79.1 418 152 .38 12.3 2.69 212 � 7 .0540 20 .248 20 .0330 3.410.1 666 143 .22 19.2 1.27 210 � 6 .0523 9.3 .239 9.7 .0331 2.711.1 463 150 .33 13.4 2.24 209 � 6 .0505 9.4 .229 9.8 .0329 2.712.1 339 159 .48 10.4 2.09 222 � 6 .0540 16 .260 16 .0350 2.913.1 622 320 .53 17.0 .87 200 � 5 .0480 5.2 .208 5.9 .0315 2.614.1 2013 1367 .70 60.3 .40 220 � 6 .0500 2.6 .239 3.7 .0347 2.615.1 276 184 .69 14.4 2.17 371 � 10 .0541 9.6 .442 10.0 .0593 2.7

Sample 120, Buziwan leuco-gabbro pegmatite (36�48.105�N, 76�56.721�E):1.1 186 275 1.53 10.2 1.41 391 � 11 .0540 14 .466 15 .0625 2.82.1 276 229 .86 15.5 1.27 403 � 11 .0513 6.9 .456 7.4 .0646 2.73.1 173 117 .70 9.83 2.32 403 � 11 .0556 16 .495 16 .0646 2.94.1 227 134 .61 12.9 1.73 407 � 11 .0587 6.9 .527 7.4 .0651 2.75.1 407 264 .67 23.3 .73 413 � 11 .0552 4.2 .504 5.0 .0662 2.76.1 1030 200 .20 57.4 .29 404 � 10 .0530 2.2 .473 3.4 .0646 2.67.1 968 187 .20 53.7 .30 402 � 10 .0539 6.7 .478 7.2 .0644 2.68.1 411 574 1.44 23.1 .98 404 � 11 .0551 4.5 .491 5.2 .0646 2.79.1 212 139 .68 12.3 1.89 412 � 11 .0523 8.5 .476 8.9 .0660 2.710.1 155 213 1.42 8.77 2.35 403 � 11 .0535 11 .476 11 .0645 2.811.1 292 171 .60 20.1 1.49 488 � 12 .0583 5.0 .632 5.7 .0787 2.612.1 298 169 .59 20.6 .95 494 � 13 .0590 3.9 .648 4.7 .0797 2.613.1 77 140 1.89 4.40 7.45 387 � 12 .0570 25 .490 25 .0618 3.3

Sample 121, Kudi biotite gneiss (36�48.188�N, 76�59.751�E):1.1 537 115 .22 69.3 .31 900 � 22 .0789 1.2 1.630 2.9 .1498 2.62.1 484 98 .21 57.9 .65 836 � 20 .0751 1.6 1.433 3.1 .1384 2.63.1 937 126 .14 93.7 .35 707 � 18 .0720 1.1 1.151 2.8 .1160 2.64.1 482 171 .37 79.0 .41 1,120 � 26 .0842 1.5 2.204 3.0 .1898 2.65.1 532 122 .24 62.0 .46 818 � 20 .0787 1.5 1.468 3.0 .1353 2.66.1 601 46 .08 50.3 .41 597 � 15 .0680 2.1 .910 3.3 .0971 2.67.1 99 110 1.15 5.76 6.08 398 � 12 .0620 19 .550 19 .0637 3.18.1 430 155 .37 86.1 .44 1,345 � 31 .1013 1.3 3.239 2.9 .2320 2.69.1 1022 27 .03 72.9 .25 513 � 13 .0604 1.5 .690 3.0 .0829 2.610.1 687 66 .10 69.2 .65 711 � 17 .0800 2.3 1.285 3.4 .1165 2.611.1 642 28 .04 50.4 .61 560 � 14 .0672 3.2 .842 4.2 .0908 2.712.1 500 101 .21 61.8 .50 863 � 21 .0783 1.6 1.546 3.0 .1432 2.613.1 765 89 .12 90.4 .51 827 � 21 .0869 2.4 1.641 3.6 .1369 2.614.1 950 8 .01 65.6 2.16 488 � 12 .0642 4.2 .696 4.9 .0786 2.615.1 594 25 .04 48.9 1.44 581 � 14 .0645 4.9 .839 5.6 .0944 2.616.1 1151 116 .10 111 .24 684 � 17 .0693 1.2 1.069 2.8 .1120 2.6

Sample 122, biotite granite of the North Kudi pluton (36�53.905�N, 76�58.909�E):1.1 53 116 2.25 3.05 9.66 376 � 15 .0410 61 .34 62 .0600 4.02.1 105 90 .88 5.68 1.77 385 � 13 .0570 18 .487 18 .0616 3.53.1 233 435 1.93 12.9 1.28 397 � 10 .0551 6.0 .483 6.6 .0635 2.74.1 131 193 1.52 7.43 3.29 400 � 11 .0456 18 .402 18 .0639 2.95.1 282 167 .61 16.1 1.10 409 � 10 .0539 6.0 .488 6.6 .0656 2.66.1 718 425 .61 42.0 .44 423 � 11 .0545 2.5 .509 3.6 .0677 2.67.1 162 147 .94 9.64 1.48 424 � 11 .0525 10.0 .493 10 .0680 2.88.1 335 390 1.20 19.9 1.72 423 � 11 .0520 5.6 .486 6.2 .0679 2.69.1 427 260 .63 23.7 1.12 400 � 10 .0558 5.9 .492 6.4 .0640 2.610.1 400 241 .62 22.8 .98 410 � 10 .0539 4.1 .488 4.9 .0656 2.611.1 96 110 1.19 5.50 2.99 403 � 11 .0551 15 .491 15 .0646 2.912.1 411 282 .71 23.0 .80 404 � 10 .0538 5.0 .479 5.6 .0646 2.6


Table 1 (Continued)

Grain,spot

U(ppm)

Th(ppm)

232Th/238U

206 ∗Pb(ppm)

%206Pbc

206Pb/238Uages (Ma)

/207 ∗ 206 ∗Pb Pb /235U207 ∗Pb /238U206 ∗Pb

Value Error (%) Value Error (%) Value Error (%)

Sample 123, dacite of the Yixieke calc-alkaline rocks (36�57.502�N, 76�59.005�E):1.1 657 204 .32 1.78 1.78 477 � 12 .0558 2.9 .586 3.9 .0761 2.62.1 596 103 .18 .67 .67 491 � 12 .0579 1.6 .630 3.0 .0789 2.63.1 640 112 .18 .62 .62 481 � 12 .585 1.9 .623 3.2 .0772 2.64.1 995 208 .22 69.9 1.00 502 � 12 .0656 1.9 .732 3.2 .0809 2.65.1 685 115 .17 46.9 .98 490 � 12 .0626 3.6 .681 4.4 .0789 2.66.1 580 174 .31 41.9 1.54 512 � 13 .0670 7.3 .763 7.7 .0827 2.67.1 538 226 .43 16.2 1.56 219 � 6 .0528 6.2 .251 6.7 .0345 2.68.1 877 627 .74 26.5 1.54 220 � 6 .0519 4.8 .248 5.5 .0346 2.69.1 578 183 .33 17.5 1.62 220 � 6 .0424 9.2 .203 9.6 .0348 2.710.1 1225 192 .16 85.9 .73 502 � 12 .0582 2.2 .650 3.4 .0810 2.611.1 475 68 .15 32.1 1.27 482 � 12 .0579 4.1 .621 4.9 .0777 2.612.1 1405 313 .23 44.5 1.91 229 � 6 .0511 5.5 .255 6.1 .0362 2.613.1 487 273 .58 14.3 2.03 212 � 6 .0481 8.5 .222 8.9 .0334 2.7

Note. Errors are 1j. Pbc and indicate the common and radiogenic portions, respectively. The error in standard calibration was∗Pb0.66%. Common Pb percentage was corrected using measured 204Pb.

terminations. We consider these to be inheritedfrom the main host gabbro.

Kudi Gneiss Complex (Sample 121). We selectedthis biotite gneiss, which comes from just 300 msouth of Kudi (fig. 3), in order to date the gneisscomplex. Zircons show variable shapes; some areeuhedral and long, others are broken and short, butthe majority are prismatic. Sixteen analyses of zir-cons yielded U/Pb ages that all show severe dis-cordance (fig. 5c), suggesting extensive Pb loss. Asindicated in table 1, the ages range from 398 �

to Ma (for interpretation, see below).12 1345 � 31North Kudi Pluton (Sample 122). We sampled the

North Kudi pluton (fig. 3) in order to compare itsSHRIMP age with published single-zircon U-Pbages of the pluton that range from 380 to 405 Ma.A biotite granite sample yielded a single populationof zircons that are short, idiomorphic, prismatic,and clear. The SHRIMP analyses (table 1; fig. 5d)produced a 206Pb/238U age of Ma ( ,408 � 7 n p 11

) that is identical to the U/Pb zirconMSWD p 1.17age of Ma of Yuan (1999) but signifi-404.0 � 3.1cantly older than the U/Pb zircon age of �1.9380.0�0.7

Ma of Xu et al. (1994).Akarz Pluton (Sample 119). We sampled this bi-

otite granite (fig. 3) where it intruded the main Bu-ziwan dunite body, leaving lenses of chromite-layered dunite (at the chromite mine) in the granite(fig. 4), in order to document its age and to providea more precise upper age limit on the convergenttectonic processes. Zircons are transparent, euhe-dral, prismatic, needle-like crystals that have mag-matic morphologies. Fifteen analyses of zirconsyielded U/Pb ages, with a number of analyses show-ing concordance (fig. 5e). Most analyses plot as agroup straddling the Concordia and give a weighted

mean 206Pb/238U age of Ma ( ,213 � 3 n p 14). One analysis (15.1 in table 1) isMSWD p 1.17

statistically an outlier from this group and indi-cates some inheritance from a ca. 371-Ma source.

Interpretation

Buziwan Ultramafic-Gabbroic Rocks and Yixieke La-vas. The Buziwan ultramafic-gabbroic rocks inthe Kudi ophiolite form an important indicator ofan oceanic basin that was eliminated during theaccretionary process along the southern boundaryof the Tarim block. The unserpentinized Buziwandunites contain layers of hornblendite in what weregard as a metamorphic foliation fabric. The pre-dominant dunite-harzburgite composition of theultramafic rocks suggests that they are metamor-phic restites after high degrees of partial melting ofa lherzolite mantle (Coleman 1977). This is con-sistent with the fact that pyroxenes in the harz-burgites have low Al2O3 and TiO2 contents, sug-gesting that these rocks are residual mantleperidotites after a high degree of partial melting(Wang et al. 2002). Within a cogenetic magma se-quence, restites of dunite-harzburgite compositionare complementary to hornblende gabbros, low-Titholeiites, and high-Mg boninites (Beccaluva et al.1983). All these rocks are different accreted com-ponents within an accretion-subduction complex,although geochemistry suggests that they possiblyall belong to the same cogenetic magmatic se-quence (cf. Khain et al. 2002).

The composition of chromian spinel is diagnosticof particular tectonic environments (Dick and Bul-len 1984). Wang et al. (2002) showed that Cr spinels

Figure 5. Concordia plot of SHRIMP U-Pb data for zircons from (a) sample 123, dacite from the Yixieke arc; (b)sample 120, pegmatite from the Buziwan gabbro; (c) sample 121, biotite gneiss from the Kudi gneiss complex; (d)sample 122, biotite granite of the North Kudi pluton; and (e) sample 119, biotite granite of the Akarz pluton.


in the Buziwan dunites have a Cr number (Cr# p) that is typical of arc-100Cr/(Cr � Al) p 60–67

related ultramafic rocks associated with a subduc-tion zone.

Although Wang et al. (2002) were unsure aboutsome age relationships of the basalts, they foundthat the boninites are locally interbedded with andoverlie the group 2 basalts. In their interpretationof the petrogenetic sequence, these authors placedemphasis on the back-arc signature of the group 2basalts, suggested that continued upwelling in theback-arc allowed hydrous fluids from the subduct-ing slab to trigger remelting of depleted refractorymantle, so forming the boninites, and that group 1island arc tholeiites enriched in LREE were createdlast by renewed subduction, which permitted hy-drous fluids and/or melts from the subducting slabto interact with mantle rocks. They invoked man-tle diapirism to initiate the back-arc spreading, fol-lowing the idea of Karig (1971), and so discountedthe more modern alternative of trench rollback asa mechanism to create the back-arc extension.However, considering the documented geochemi-cal characteristics in relation to current ideas onarc magmatism, we suggest the following evolu-tionary scenario.

The most widely accepted model for the for-mation of back-arc basins and of supra-subductionzone ophiolites depends on hinge rollback (Maru-yama 1997; Shervais 2001). The several phases ofdevelopment of the Kudi ophiolite are precisely pre-dicted by the model of Shervais (2001). The initialphase of ophiolite formation in the forearc givesrise to low-K, LREE-depleted tholeiites, whichrange in composition from basalt to basaltic an-desite and even dacite and often have flat REE pat-terns and trace elements that resemble MORB(group 2 gabbros and tholeiites) and form by melt-ing of MORB source asthenosphere before any in-troduction of fluids from the subducting plate. Ourdated dacite (sample 123) comes from these low-ermost volcanic rocks, and thus its age provides areasonable estimate of the time of initiation ofophiolite formation.

There are several alternative interpretations forthe origin of sample 123 dacite from near the baseof the Yixieke volcanics: (1) the later date (220 �

Ma) is the result of Pb loss during a Triassic ther-5mal event; (2) the dacite is a hypabyssal intrusionrelated to Triassic magmatism, and the -492 � 9Ma-dated zircons are xenocrystic grains; (3)

Ma is the age of crystallization of the da-492 � 9cite, and the 220-Ma age is due to reheating by aTriassic thermal event. Because the -Ma220 � 5date is concordant, as indicated above, we exclude

the first possibility. If we treat the dacite as a hyp-abyssal intrusion related to Triassic magmatismand the -Ma-dated zircons as xenocrystic492 � 9grains, it is hard to reconcile the fact that a similarTriassic granite (sample 119) and a Devonian gran-ite (sample 122), which both occur nearby, do nothave any old xenocrystic grains. In addition, in theTriassic rock assemblages no dacite has ever beenreported in the Kudi area. A close association of thedacite with the Yixieke arc volcanic rocks leads usto accept the last interpretation. This is in accordwith the fact that most early Paleozoic compo-nents, i.e., the leuco-gabbroic pegmatite and theYirba pluton, are closely related to the Kudi arc-ophiolite assemblage, which all have ca. 490-Maages. A more detailed investigation is needed to testthe interpretation that we favor here.

The second phase generates tholeiites that arelow in Ti, enriched in LREE, and depleted in Nband Mg-enriched boninites. These melts arebrought about by an increasing flux of fluids fromthe subducting slab. At this stage of the Shervais(2001) model, an even higher fluid flux would giverise to tholeiites even more enriched in LREE andTi (Shervais 2001), like the group 1 basalts.

The third and final phase in the formation of anaccreted active margin in relation to the supra-subduction ophiolite is the generation of calc-alkaline basaltic andesites, andesites, and rhyolites,together with hornblende-bearing diorites, tonal-ites, and trondhjemites. The Yirba hornblendegranodiorite belongs to this mature phase of de-velopment. Here Ma may be reasonably471 � 5regarded as the age of crystallization (Yuan et al.2002a, 2002b), and we interpret -Ma zir-491 � 3cons as xenocrysts inherited from the early group2 lavas. The 492-Ma dacite in the lowermost lavasis close to the start of formation of the active mar-gin, and 471 Ma (the age of latest granodiorite plu-ton) is close to the mature stage of its development.

We conclude that the Kudi ophiolite wentthrough a typical supra-subduction developmentand has a corresponding geochemical signature, inwhich case there is no need to invoke a single back-arc model or a second-stage subduction in a back-arc basin, as proposed by Wang et al. (2002). Wenote that many of the stratigraphic and geochem-ical characteristics of the Kudi ophiolite are similarto those of the Late Jurassic ophiolite in Hokkaido,Japan, which Takashima et al. (2002) concludedformed in a forearc rift basin above a supra-subduction zone; viz., harzburgites, pyroxenites,and dunites containing chromian spinels with arc-related chemistry, tholeiitic basalts with back-arcbasin-like chemistry interbedded with and overlain

698 W . J . X I A O E T A L .

by boninitic high-Mg andesites, calc-alkaline an-desites, and turbidites in a forearc basin. Similarrelationships occur in the 1020-Ma Dunzhugursupra-subduction ophiolite in Siberia (Khain et al.2002).

The Kudi Gneiss Complex. A paragneiss from thiscomplex contains zircons that have a wide rangeof dates, from to Ma. The lat-398 � 12 1345 � 31est date, Ma, is located along the con-398 � 12cordia, which indicates a concordant age. This

-Ma concordant date is an indication of398 � 12newly formed magmatic zircon in the Kudi com-plex; thus, a possible Early Devonian magmaticevent may predate an early phase of accretion partlyrepresented by the paragneiss. This is in good agree-ment with the similar ages of the North Kudi plu-ton and the Buziwan leuco-gabbroic pegmatite. Inaddition, Matte et al. (1996) reported Ar-Ar ages of380–350 Ma, and Zhou (1998) and Zhou et al. (2000)reported Ar-Ar ages of 452–428 Ma in different seg-ments of the Kudi gneiss complex, which they in-terpreted as records of metamorphism.

The other, earlier dates in the Kudi gneiss areobviously for detrital zircons. Z. Hui (pers. comm.)has obtained a U-Pb zircon date with a lower in-tercept age of Ma and an upper intercept533 � 21age of Ma on a Kudi gneiss. The inter-1251 � 23mediate dates represent original ages modified byextensive Pb loss caused by the youngest meta-morphism. This is consistent with the fact thatgranites intruding the Kudi gneisses yield Ndmodel ages of 1.1–1.5 Ga (Yuan et al. 2002a). Wetherefore interpret the earlier dates to reflect theprotolith ages of the continental gneissic block thatdocked into the trench and the 452–350-Ma datesas the time of peak metamorphism that gave riseto the high-grade mineral assemblage and refoliatedfabric of the gneiss created during accretion.

Discussion

The History of the Kunlun Active Continental Mar-gin. The tectonic history of the WKO has beensummarized by many workers from their availablegeochemical, stratigraphic, structural, and tectonicdata, the most recent of which were Xiao et al.(2002a, 2002b, 2003a) and Yuan et al. (2002a, 2002b,2004). Our SHRIMP zircon dates provide new con-straints on the timing of several key tectonic events,which correspondingly require new interpretations.Below we present a revised Paleozoic–early Meso-zoic tectonic history of the WKO, illustrated in fig-ure 6.

A passive continental margin rift succession ex-isted before the early Paleozoic on the southern

border of the Tarim block, with an ocean (proto-Tethys) to the south (Pan 1996). As we discussedearlier (Xiao et al. 2002b, 2003a), there was a periodof southward subduction of the passive marginalsequence of the Tarim block in the Late Cambrianto earliest Ordovician (fig. 6a). From the Early Or-dovician, the ca. 490-Ma Yixieke arc–Kudi ophio-lite complex accreted to the Tarim block, a north-ward subduction followed beneath the compositeTarim accretionary margin, and thus the early stageof an Andean-type magmatic arc developed on thesouthern margin of the Tarim block (fig. 6b). Thisis consistent with and explains the fact that thedacite from the Yixieke arc volcanic rocks has aformation age of Ma and that ca. 490-Ma492 � 9inherited zircons have been found in both the

-Ma leuco-gabbro pegmatite from the Bu-403 � 7ziwan valley and the -Ma granodiorite from471 � 5the Yirba pluton. The intrusion of the Yirba grano-diorite at Ma (Yuan et al. 2002b) and of a471 � 5similar -Ma pluton nearby (Xu et al. 1996)�2.4460�2.5

marked the mature phase of development of thisAndean-type active margin.

The mid-Proterozoic Kudi continental block wasapproaching the subduction zone in the LateOrdovician–Silurian to Early Devonian when theductile shear zone in the gneisses was created (fig.6c). The Kudi gneiss complex underwent a rela-tively long accretionary process, as it contains ob-viously different tectonic components, including aca. 398-Ma paragneiss. In the Early to Middle De-vonian, the ultramafic rocks of the ophiolite werethrust over the Kudi gneiss complex, probably in atrench. The collision and accretion at the leadingedge of the active margin assisted the creation ofa thrust-thickened mountain belt. Melting of un-derlying metasomatized mantle wedge created the405-Ma lamprophyres (Zhou and Li 2000; fig. 6c).Further development of the active margin gener-ated the 408–380-Ma North Kudi granite and the403-Ma leuco-gabbro pegmatite dikes that retaininherited zircons from the early history of the mar-gin. In the Late Devonian to Early Carboniferous,a forearc basin created over the accretionary beltreceived turbidites that contain clastic debris fromthe eroding mountains (fig. 6d).

The resumption of northward-dipping subduc-tion along the southern margin of the Tarim blockin the Permian to early Mesozoic was contempo-raneous with the collision between the Siberia-Altaid continent and the northern margin of theTarim block (Heubeck 2001; Roger et al. 2003; Xiaoet al. 2003b). The 214-Ma Akarz subduction-relatedgranodiorite represents the third and final stage inthe development of the Andean-type margin (figs.

Figure 6. Sequential diagram showing the Palaeozoic–early Mesozoic tectonic evolution of the WKO. (a) LateCambrian to Early Ordovician; (b) Middle Ordovician; (c) Late Ordovician to Middle Devonian; (d) Late Devonian toEarly Carboniferous; (e) Late Carboniferous–Permian to early Mesozoic. See text for discussion.

700 W . J . X I A O E T A L .

Figure 7. Histogram with cumulative probability of alldates in this study.

6e, 7). The southern Tarim active margin collidedwith the Qiangtang block to the south in the LateJurassic (Roger et al. 2003). This collision termi-nated all subduction-related tectonic processes inthe northern WKO.

The Long-Lived Active Margin of Southern Eurasia.From the distribution of the ages of granitic rocks,a lack of magmatic activity from 350 to 220 Ma(Yuan 1999) was previously interpreted as being re-lated to cessation of subduction (Yuan et al. 2002a,2002b). An analysis of a histogram with cumulativeprobability also shows that there was such a gap(fig. 7). However, the paleogeographic reconstruc-tions of Nie et al. (1990) suggested northward sub-duction under the southern margin of the Tarimblock in the Early Permian, and this idea was sup-ported by the discovery of Permian subduction-related volcanic rock along the southern margin ofthis subduction system (Matte et al. 1996; Matternet al. 1996; Pan 1996). Li et al. (1995) and Bi et al.(1999) summarized Early Permian magmatic activ-ity in the WKO. In the southern part of the SouthKunlun, Middle Devonian to Early Permian grano-diorites were reported (Li et al. 1995; Xu et al. 1996;Bi et al. 1999). Forearc accretion south of this pos-sible late Paleozoic magmatic arc took place in theLate Carboniferous to early Mesozoic (Xiao et al.2002b, 2003a; Schwab et al. 2004).

In the meantime, in the eastern Kunlun, plutonsand volcanic rocks of arc affinity likely formed inthe 370–320-Ma interval (Dewey et al. 1988;Schwab et al. 2004); in the Pamirs to the west, vol-canism related to an arc began at ∼370 and 320 Maand most likely continued into the Triassic(Schwab et al. 2004). Therefore, we propose that

during the late Paleozoic to early Mesozoic, thesouthern active margin of the Tarim block still ex-isted (Xiao et al. 2002a) and that there was only arelatively short cessation of magmatic activity inthe Carboniferous. This kind of magmatic cessa-tion is not uncommon in active margins, such asthe present-day western North American activemargin that is characterized by transform tectonicswith large-scale strike-slip faults and related basins(Dickinson 1995).

This would be consistent with paleomagneticdata that indicate that the Tarim block was movingnorthward as a united plate in the Devonian to LateCarboniferous (Li 1990; Yin and Nie 1996) and thata collision occurred between the Tien Shan andsouthern Siberia in Carboniferous-Permian times(Windley et al. 1990). The period of northward driftof the Tarim block resulted in a reduction in therate of subduction-accretion during the time period,with a possible cessation in the Early Carbonifer-ous, which would explain the relatively smallamount of documented igneous and accretionaryactivity during this time. The WKO was submergedand overlain by a thick pile of marine deposits inthe Late Devonian to Early Carboniferous (XBGMR1993). Several early Paleozoic fossiliferous blocksalong this accretionary complex (Yin and Bian 1995;Mattern et al. 1996; Pan 1996; Xiao et al. 2002b,2003a) were probably incorporated by arc-parallelstrike-slip faulting (or associated extension) duringthis time.

Therefore, we conclude, on the basis of theSHRIMP dates and our other cited evidence, thatthe southern Tarim active margin underwent anearly magmatic event at ca. 490 Ma that was fol-lowed by accretionary processes at ca. 400 Ma andwas finally terminated by Andean-type accretion-ary orogenesis at ca. 214 Ma, as indicated by thehistogram of the SHRIMP dates conducted in thisstudy (fig. 7) and other chronological data sum-marized by Pan (1996), Yuan et al. (2002a, 2002b),Cowgill et al. (2003), and Gehrels et al. (2003a,2003b).

In recent years, forearc accretion has gained in-creasing popularity as a process to explain the evo-lution of many orogenic belts, such as those in cen-tral Asia, the Arabian-Nubian Shield (e.g., Sengorand Natal’in 1996), the Lachlan Orogen of easternAustralia (Gray 1997; Gray and Foster 1998; Fosterand Gray 2000), and the Proterozoic Yavapai Oro-gen south of the Wyoming craton (e.g., Hoffman1988). Based on the Japanese model, accretionaryorogens evolve largely by processes of forearc ac-cretion (Isozaki et al. 1990; Sengor and Okurogul-lari 1991; Windley 1992; Sengor et al. 1993; Xiao


Figure 8. Permian to Early Triassic paleogeography of Eurasia, with emphasis on the continental blocks and relatedorogenic belts in central-east Asia (modified after Heubeck 2001; Xiao et al. 2003b). Dark gray shows some Precambrianblocks in the east.

et al. 2003b). Our geochronological data and tec-tonic interpretation all indicate that the Paleozoicto early Mesozoic WKO was a long-lasting, com-plicated accretionary orogen, because an early Me-sozoic Andean-type active continental margin de-veloped on the Paleozoic-accreted margin of theTarim block. This long-lived active continentalmargin is characterized by a major accretionarycomplex and forearc basin on its southern side.This new information will shed light on an im-portant controversy about the evolution of this partof Asia, viz., whether the WKO was a collisionalorogen that resulted from either the collision ofvarious terranes between the Tarim and Qiangtangblocks (Dewey et al. 1988; Jiang et al. 1992; Matteet al. 1996; Mattern et al. 1996) or collapse of back-arc basins (Yao and Hsu 1994) or an accretionaryorogen (Sengor and Okurogullari 1991; Mattern andSchneider 2000; Xiao et al. 2002a, 2002b, 2003a).From a more regional perspective, Sobel and Ar-naud (1999) and Cowgill et al. (2003) compared theWKO with the East Kunlun, and their work sup-ports a general tectonic model for the northern Ti-betan plateau in which an intermediate island arcor composite terrane was accreted to the Tarimblock along a southward-dipping subduction zone.Gehrels et al. (2003a, 2003b) summarized differentmodels for the formation of the northern Tibetanplateau; they excluded a back-arc model but sup-

ported a general tectonic model like that proposedhere. The general multiple accretionary frameworkof our tectonic model for the WKO is consistentwith all previous investigations in which accre-tionary tectonics played a key role together withsouthward subduction followed by arc accretion,subduction flip (northward subduction), and for-mation of an Andean-type active margin. Recenttectonic analysis of accretionary complexes inother areas of the Tibetan Plateau (Kapp et al. 2000,2003a, 2003b; Yin and Harrison 2000; Aitchison etal. 2001) have greatly increased our understandingof the role of accretion on the southern side of theWKO in this segment of southern Eurasia.

Our proposed tectonic scenario of a long-livedactive continental margin is comparable to that inthe East Kunlun and Qinling (figs. 1, 8; Molnar etal. 1987; He et al. 1999; Sobel and Arnaud 1999;Xiao et al. 2002a; Cowgill et al. 2003; Roger et al.2003; Bian et al. 2004; Schwab et al. 2004), althoughdetailed aspects of the evolution could be different(Gehrels et al. 2003a, 2003b) because of possibleorogen-parallel variations. There are also similari-ties farther east in the Qinling-Dabie Orogen,where the North and South China blocks collidedby the Late Permian (Nie et al. 1990), with the mostactive deformation period in the Late Triassic toEarly Jurassic (Zhao and Coe 1987; Enkin et al.

702 W . J . X I A O E T A L .

1992; Gilder et al. 1999; Roger et al. 2003; Bian etal. 2004; Schwab et al. 2004).

The early to late Paleozoic active continentalmargin along the Kunlun range apparently initiateda tectonic framework that influenced all subse-quent paleogeographic developments in northernTibet (Xiao et al. 2002a; Roger et al. 2003). Heubeck(2001) showed that this active margin extendedfrom the Qinling to the Caucasus from at least theMiddle Devonian to the Late Permian (fig. 8).Therefore, in the late Paleozoic to early Mesozoic,a continuous active margin of southern Eurasia ex-tended for some 10,000 km north of the paleo-Tethys ocean (figs. 1, 8). This subduction-accretionhistory was mostly superimposed on Paleozoic ac-cretion and amalgamation along a northward-dipping subduction zone beneath the southern ac-tive margin of Eurasia (Lin et al. 1985; Reischmannet al. 1990; Kroner et al. 1993; Meng and Zhang1999, 2000; Heubeck 2001; Ratschbacher et al.2003). The amalgamation of the southerly derivedblocks that accreted to Eurasia in Mesozoic-Cenozoic time (including the Qiangtang-Cimmeriablock in the Late Jurassic) took place in this paleo-

geographic framework (Dewey et al. 1988; Kapp etal. 2000, 2003a, 2003b).

A C K N O W L E D G M E N T S

We are grateful to the personnel of the BeijingSHRIMP Laboratory for their kind assistance. Wesincerely thank Q. L. Hou, Z. H. Wang, J. Hao, A.M. Fang, G. C. Zhang, H. L. Chen, and H. Zhou fortheir help in the field and laboratory. W. J. Xiao isgrateful to the University of Hong Kong, where hewas invited as a visiting scholar and prepared thismanuscript. Discussions with Y. S. Pan, W. M.Deng, J. Aitchison, G. C. Zhao, and M.-F. Zhougreatly improved early drafts. We thank two anon-ymous reviewers and A. Anderson for constructivecomments and suggestions that greatly improvedthe manuscript. Funding by Projects of the ChineseNational Science Foundation (40172080, 40234045,40032010, and 40372042), the Chinese Academy ofSciences (KZCX2-SW-119), the Hong Kong Re-search Grants Council (HKU7040/04P), and theRoyal Society of London is gratefully acknowl-edged. This article forms a contribution to IGCP473 and 420.

R E F E R E N C E S C I T E D

Aitchison, J. C.; Badengzhu; Davis, A. M.; Liu, J.; Luo,H.; Malpas, J. G.; McDerimid, I. R. C.; Wu, H.; Ziabrev,S. V.; and Zhou, M.-F. 2001. Remnants of a Cretaceousintra-oceanic subduction system within the Yarlung-Zangbo suture (southern Tibet). Earth Planet. Sci.Lett. 183:231–244.

Beccaluva, L.; DiGirolamo, P.; Macciotta, G.; and Morra,V. 1983. Magma affinities and fractionation trends inophiolites. Ofioliti 8:307–323.

Bi, H.; Wang, Z. G.; Wang, Y. L.; and Zhu, X. Q. 1999.Tectono-magmatic evolution of the western Kunlunorogenic belt. Sci. China 29:398–406.

Bian, Q.-T.; Li, D.-H.; Pospelov, I.; Yin, L.-M.; Li, H.-S.;Zhao, D.-S.; Chang, C.-F.; et al. 2004. Age, geochem-istry and tectonic setting of Buqingshan ophiolites,North Qinghai-Tibet Plateau, China. J. Asian EarthSci. 23:577–596.

Claoue-Long, J. C.; Compston, W.; Roberts, J.; and Fan-ning, C. M. 1995. Two Carboniferous ages: a compar-ison of SHRIMP zircon dating with convectional zir-con ages and 40Ar/39Ar analysis. In W. W. Berggren; D.V. Kent; M.-P. Aubry; and J. Hardenbol, eds. Geochro-nology, time scales and global stratigraphic correla-tion. SEPM Spec. Publ. 54:3–21.

Coleman, R. G. 1977. Ophiolites: ancient oceanic lith-osphere? Berlin, Springer, 229 p.

Compston, W.; Williams, I. S.; Kirschvink, J. L.; Zhang,Z.; and Ma, G. 1992. Zircon U-Pb ages for the EarlyCambrian timetable. J. Geol. Soc. Lond. 149:171–184.

Compston, W.; Williams, I. S.; and Meyer, C. 1984. U-Pb geochronology of zircons from lunar breccia 73217using a sensitive high mass-resolution ion micro-probe. J. Geophys. Res. B 89:525–534.

Cowgill, E.; Arrowsmith, J. R.; Yin, A.; Wang, X.; andChen, Z. 2004a. The Akato Tagh bend along the AltynTagh fault, northwest Tibet 2: active deformation andthe importance of transpression and strain hardeningwithin the Altyn Tagh system. Geol. Soc. Am. Bull.116:1443–1464.

Cowgill, E.; Yin, A.; Arrowsmith, J. R.; Wang, X.; andZhang, S. 2004b. The Akato Tagh bend along the Al-tyn Tagh fault, northwest Tibet 1: smoothing byvertical-axis rotation and the effect of topographicstresses on bend-flanking faults. Geol. Soc. Am. Bull.116:1423–1442.

Cowgill, E.; Yin, A.; Harrison, T. M.; and Wang, X.-F.2003. Reconstruction of the Altyn Tagh fault basedon U-Pb geochronology: role of back thrusts, mantlesutures, and heterogeneous crustal strength in form-ing the Tibetan Plateau. J. Geophys. Res. 108:2346,doi:10.1029/2002JB002080.

Deng, W. M. 1995. The geologic characteristics of theophiolites in the Karakoram-West Kunlun region andtheir tectonic significance. Acta Petrol. Sin. 11(suppl.):98–111.

Dewey, J. F.; Shackleton, R.; Chang, C. F.; and Sun, Y.1988. The tectonic evolution of the Tibetan Plateau.Philos. Trans. R. Soc. Lond. A 327:379–413.


Dick, H. J. B., and Bullen, T. 1984. Chromian spinel asa petrogenetic indicator in abyssal and alpine-type pe-ridotites and spatially associated lavas. Contrib. Min-eral. Petrol. 86:54–76.

Dickinson, W. R. 1995. Forearc basins. In Busby, C. J.,and Ingersoll, R. V., eds. Tectonics of sedimentary ba-sins. Cambridge, MA, Blackwell Science, p. 221–261.

Ding, D. G.; Wang, D. X.; Liu, W. X.; and Sun, S. Q. 1996.The West Kunlun orogenic belts and basins. Beijing,Geological Publishing House, 143 p. (in Chinese withEnglish abstract).

Enkin, R. J.; Yang, Z.; Chen, Y.; and Courtillot, V. 1992.Paleomagnetic constraints on the geodynamic historyof the major blocks of China from Permian to thepresent. J. Geophys. Res. B 97:13,953–13,989.

Fang, A. M. 1998. The sedimentology and tectonic evo-lution of a fore-arc basin flysch in the Kudi ophiolite,west Kunlun, Xinjiang, China. PhD dissertation, In-stitute of Geology and Geophysics, Chinese Academyof Sciences, 108 p.

Fang, A. M.; Li, J. L.; Hou, Q. L.; Zhou, H.; Wan, S.; andLi, H. 1998. The first discovery of radiolarian fossilsin Yixikegou flysch section of Kuda, west Kunlun,China. Sci. Geol. Sin. 33:384.

Foster, D. A., and Gray, D. R. 2000. Evolution and struc-ture of the Lachlan Fold Belt (Orogen) of eastern Aus-tralia. Annu. Rev. Earth Planet. Sci. 28:47–80.

Gehrels, G. E.; Yin, A.; and Wang, X.-F. 2003a. Detritalzircon geochronology of the northeastern Tibetan pla-teau. Geol. Soc. Am. Bull. 115:881–896.

———. 2003b. Magmatic history of the northeastern Ti-betan Plateau. J. Geophys. Res. 108:2423, doi:10.1029/2002JB001876.

Gilder, S.; Leloup, N.; Courtillot, V.; Chen, Y.; Coe, R.S.; Zhao, X. X.; Xiao, W. J.; et al. 1999. Tectonic evo-lution of the Tancheng-Lujiang (Tan-Lu) fault via Mid-dle Triassic to Early Cenozoic paleomagnetic data. J.Geophys. Res. B 104:15,365–15,390.

Gradstein, F. M.; Ogg, J. G.; Smith, A. G.; Bleeker, W.;and Lourens, L. J. 2004. A new geological time scale,with special reference to Precambrian and Neogene.Episodes 27:83–100.

Gray, D. R. 1997. Tectonics of the southeastern Austra-lian Lachlan fold belt: structural and thermal aspects.In Burg, J. P., and Ford, M., eds. Orogeny through time.Geol. Soc. Lond. Spec. Publ. 121:149–177.

Gray, D. R., and Foster, D. A. 1998. Character and ki-nematics of faults within the turbidite-dominatedLachlan Orogen: implications for the tectonic evolu-tion of eastern Australia. J. Struct. Geol. 20:1691–1720.

He, Z.; Tian, S.; Xu, Z.; Yang, J.; and Cui, J. 1999. Dis-covery of the late Palaeozoic radiolarian fossils in themiddle segment of the Altyn Tagh and its implication.Chin. Geol. Rev. 45:246.

Heubeck, C. 2001. Assembly of central Asia during themiddle and late Paleozoic. In Hendrix, M. S., and Da-vis, G. A., eds. Paleozoic and Mesozoic tectonic evo-lution of central and eastern Asia: from continental

assembly to intracontinental deformation. Geol. Soc.Am. Mem. 194:1–22.

Hickey, R. L., and Frey, F. A. 1982. Geochemical char-acteristics of boninitic-series volcanics: implicationsfor their source. Geochim. Cosmochim. Acta 46:2099–2115.

Hoffman, P. F. 1988. United plates of America, the birthof a craton: early Proterozoic assembly and growth ofLaurentia. Annu. Rev. Earth Planet. Sci. 16:543–603.

Huang, X.-L.; Xu, Y.-G.; and Liu, D.-Y. 2004. Geochro-nology, petrology and geochemistry of the granulitexenoliths from Nushan, east China: implication for aheterogeneous lower crust beneath the Sino-Koreancraton. Geochim. Cosmochim. Acta 68:127–149.

Isozaki, Y.; Maruyama, S.; and Furoka, F. 1990. Accretedoceanic materials in Japan. Tectonophysics 181:179–205.

Jiang, C. F.; Yang, J. S.; Feng, B. G.; and Zhu, Z. 1992.Opening and closing tectonics of the Kunlun Moun-tains. Beijing, Geological Publishing House, 217 p.

Jiang, Y.-H.; Jiang, S.-Y.; Ling, H.-F.; Zhou, X.-R.; Rui, X.-J.; and Yang, W.-Z. 2002. Petrology and geochemistryof shoshonitic plutons from the western Kunlun oro-genic belt, Xinjiang, northwestern China: implica-tions for granitoid genesis. Lithos 63:165–187.

Kapp, P.; Murphy, M. A.; Yin, A.; Harrison, T. M.; Ding,L.; and Guo, J. 2003a. Mesozoic and Cenozoic tectonicevolution of the Shiquanhe area of western Tibet. Tec-tonics 22:1029, doi:10.1029/2001TC001332.

Kapp, P.; Yin, A.; Manning, C. E.; Murphy, M.; Harrison,T. M.; Spurlin, M.; Ding, L.; Deng, X.-G.; and Wu, C.-M. 2000. Blueschist-bearing metamorphic core com-plexes in the Qiangtang block reveal deep crustalstructure of northern Tibet. Geology 28:19–22.

Kapp, P.; Yin, A.; Manning, C. E.; Murphy, M.; Harrison,T. M.; Taylor, M. H.; and Ding, L. 2003b. Tectonicevolution of the early Mesozoic blueschist-bearingQiangtang metamorphic belt, central Tibet. Tectonics22:1043, doi:10.1029/2002TC001383.

Karig, D. E. 1971. Origin and development of marginalbasins in the western Pacific. J. Geophys. Res. 76:2542–2561.

Khain, E. V.; Bibikova, E. V.; Kroner, A.; Zhuravlev, D.Z.; Sklyarov, E. V.; Fedotova, A. A.; and Kravchenko-Berezhnoy, I. R. 2002. The most ancient ophiolite ofthe central Asian fold belt: U-Pb and Pb-Pb zircon agesfor the Dunzhugur complex, eastern Sayan, Siberia,and geodynamic implications. Earth Planet. Sci. Lett.199:311–325.

Kroner, A.; Zhang, G.; Zhou, D.; and Sun, Y. 1993. Gran-ulites in the Tongbai area, Qinling belt, China: geo-chemistry, petrology, single zircon geochronology andimplications for the tectonic evolution of east Asia.Tectonics 12:245–255.

Li, Y. A.; Li, X. D.; Sun, D. J.; and Han, Y. L. 1995. Tec-tonic evolution of Qiangtang Block and Kangxiwarstructure zone in Kara-Kunlun Mountains, south-western Xinjiang, China. Urumqi, Xinjiang Sci-Techand Hygiene Publishing House, 107 p.

704 W . J . X I A O E T A L .

Li, Y. P. 1990. An apparent polar wander path from theTarim Block, China. Tectonophysics 181:31–41.

Lin, J.; Fuller, M.; and Zhang, W. 1985. Preliminary Phan-erozoic polar wander paths for the north and southChina blocks. Nature 313:444–449.

Ludwig, K. R. 1999. Using Isoplot/Ex, version 2.00: a geo-chronological toolkit for Microsoft Excel. BerkeleyGeochronological Center Spec. Publ. 1a.

———. 2001. Squid 1.02: a user manual. Berkeley Geo-chronological Center Spec. Publ. 2.

Maruyama, S. 1997. Pacific-type orogeny revisited:Miyashiro-type orogeny proposed. Island Arc 6:91–120.

Matte, P.; Tapponnier, P.; Arnaud, N.; Bourjot, L.; Avouac,J. P.; Vidal, P.; Liu, Q.; Pan, Y.; and Wang, Y. 1996.Tectonics of western Tibet, between the Tarim andthe Indus. Earth Planet. Sci. Lett. 142:311–330.

Mattern, F., and Schneider, W. 2000. Suturing of theproto- and paleo-Tethys oceans in the western Kunlun(Xinjiang, China). J. Asian Earth Sci. 18:637–650.

Mattern, F.; Schneider, W.; Li, Y.; and Li, X. 1996. A tra-verse through the western Kunlun (Xinjiang, China):tentative geodynamic implications for the Paleozoicand Mesozoic. Geol. Rundsch. 85:705–722.

Meng, Q.-R., and Zhang, G.-W. 1999. Timing of collisionof the north and south China blocks: controversy andreconciliation. Geology 27:123–126.

———. 2000. Geological framework and tectonic evo-lution of the Qinling Orogen, central China. Tectono-physics 323:183–196.

Molnar, P.; Burchfiel, C.; Zhao, Z.; Lian, K.; Wang, S.;and Huang, M. 1987. Geologic evolution of north Ti-bet: results from an expedition to Ulugh Muztagh.Science 235:299–305.

Nie, S.; Rowley, D. B.; and Ziegler, A. M. 1990. Con-straints on the location of Asian microcontinents inpalaeo-Tethys during the Late Palaeozoic. In Mc-Kerrow, W. S., and Scotese, C. R., eds. Palaeozoic pa-laeogeography and biogeography. Geol. Soc. Lond.Mem. 12:397–409.

Pan, Y. S. 1996. Geological evolution of the Karakoramand Kunlun Mountains. Beijing, Seismological Press,288 p.

Ratschbacher, L.; Hacker, B. R.; Calvert, A.; Webb, L. E.;Grimmer, J. C.; McWilliams, M. O.; Ireland, T.; DongS.; and Hu, J. 2003. Tectonics of the Qinling (centralChina): tectonostratigraphy, geochronology, and de-formation history. Tectonophysics 366:1–53.

Reischmann, T.; Altenberger, U.; Kroner, A.; Zhang, G.;Sun, Y.; and Yu, Z. 1990. Mechanism and time of de-formation and metamorphism of mylonitic ortho-gneisses from the Shagou shear zone, Qinling Belt,China. Tectonophysics 185:91–109.

Roger, F.; Arnaud, N.; Gilder, S.; Tapponnier, P.; Jolivet,M.; Brunel, M.; Malavieille, J.; Xu, Z.; and Yang, J.2003. Geochronological and geochemical constraintson Mesozoic suturing in east central Tibet. Tectonics22:1037, doi:10.1029/2002TC001466.

Schwab, M.; Ratschbacher, L.; Siebel, W.; McWilliams,M.; Minaev, V.; Lutkov, V.; Chen, F.; et al. 2004. As-

sembly of the Pamirs: age and origin of magmatic beltsfrom the southern Tien Shan to the southern Pamirsand their relation to Tibet. Tectonics 23:TC4002, doi:10.1029/2003TC001583.

Sengor, A. M. C., and Natal’in, B. A. 1996. Paleotectonicsof Asia: fragments of a synthesis. In Yin, A., and Har-rison, T. M., eds. The tectonic evolution of Asia. Cam-bridge, Cambridge University Press, p. 486–640.

Sengor, A. M. C.; Natal’in, B. A.; and Burtman, V. S. 1993.Evolution of the Altaid tectonic collage and Palaeo-zoic crustal growth in Eurasia. Nature 364:299–307.

Sengor, A. M. C., and Okurogullari, A. H. 1991. The roleof accretionary wedges in the growth of continents:Asiatic examples from Argand to plate tectonics. Eclo-gae Geol. Helv. 84:535–597.

Shervais, J. W. 2001. Birth, death, and resurrection:the life cycle of suprasubduction zone ophiolites.Geochem. Geophys. Geosyst. 2(1), doi:10.1029/2000GC000080.

Sobel, E. R., and Arnaud, N. 1999. A possible middlePaleozoic suture in the Altyn Tagh, NW China. Tec-tonics 18:64–74.

Takashima, R.; Nishi, H.; and Yoshida, T. 2002. Geology,petrology and tectonic setting of the Late Jurassicophiolite in Hokkaido, Japan. J. Asian Earth Sci. 21:197–215.

Wang, E. 1997. Displacement and timing along the north-ern strand of the Altyn Tagh fault zone, northern Ti-bet. Earth Planet. Sci. Lett. 150:55–64.

Wang, Y. Z. 1983. The time of the Yishak Group in thewest Kunlun Mountains and its tectonic implications.Xinjiang Geol. 1:1–8 (in Chinese with Englishabstract).

Wang, Z. H.; Sun, S.; Hou, Q. L.; and Li, J. L. 2001. Effectof melt-rock interaction on geochemistry in the Kudiophiolite (western Kunlun Mountains, northwesternChina): implication for ophiolite origin. Earth Planet.Sci. Lett. 191:33–48.

Wang, Z. H.; Sun, S.; Li, J. L.; and Hou, Q. L. 2002. Pet-rogenesis of tholeiite associations in Kudi ophiolite(western Kunlun Mountains, northwestern China):implication for the evolution of back-arc basins. Con-trib. Mineral. Petrol. 143:471–483.

Williams, I. S. 1992. Some observations on the use ofzircon U-Pb geochronology in the study of graniticrocks. Trans. R. Soc. Edinb. Earth Sci. 83:447–458.

Williams, I. S.; Buick, I. S.; and Cartwright, I. 1996. Anextended episode of early Mesoproterozoic metamor-phic fluid flow in the Reynolds Range, central Aus-tralia. J. Metamorph. Geol. 14:29–47.

Williams, W., and Claesson, S. 1987. Isotopic evidencefor the Precambrian provenance and Caledonian meta-morphism of high-grade paragneisses from the SeveNappes, Scandinavian Caledonides. II. Ion microprobezircon U-Th-Pb. Contrib. Mineral. Petrol. 97:205–217.

Windley, B. F. 1992. Proterozoic collisional and accre-tionary orogens. In Condie, K. C., ed. Proterozoiccrustal evolution. Amsterdam, Elsevier, p. 419–446.

Windley, B. F.; Allen, M. B.; Zhang, C.; Zhao, Z. Y.; andWang, G. R. 1990. Paleozoic accretion and Cenozoic


re-deformation of the Chinese Tien Shan Range, Cen-tral Asia. Geology 18:128–131.

Wittlinger, G.; Vergne, J.; Tapponnier, P.; Farra, V.; Pou-pinet, G.; Jiang, M.; Su, H.; Herquel, G.; and Paul, A.2004. Teleseismic imaging of subducting lithosphereand Moho offsets beneath western Tibet. Earth Planet.Sci. Lett. 221:117–130.

XBGMR (Xinjiang Bureau of Geology and Mineral Re-sources). 1993. Memoir of Xinjiang geology. Beijing,Geological Publishing House, 540 p.

Xiao, W. J.; Han, F. L.; Windley, B. F.; Yuan, C.; Zhou, H.;and Li, J. L. 2003a. Multiple accretionary orogenesisand episodic growth of continents: insights from thewestern Kunlun Range, Central Asia. Int. Geol. Rev.45:303–328.

Xiao, W. J.; Windley, B. F.; Chen, H. L.; Zhang, G. C.; andLi, J. L. 2002a. Carboniferous-Triassic subduction andaccretion in the western Kunlun, China: implicationsfor the collisional and accretionary tectonics of thenorthern Tibetan plateau. Geology 30:295–298.

Xiao, W. J.; Windley, B. F.; Hao, J.; and Li, J. L. 2002b.Arc-ophiolite obduction in the western Kunlun Range(China): implications for the Palaeozoic evolution ofcentral Asia. J. Geol. Soc. Lond. 159:517–528.

Xiao, W. J.; Windley, B. F.; Hao, J.; and Zhai, M.-G. 2003b.Accretion leading to collision and the Permian Solon-ker suture, Inner Mongolia, China: termination of thecentral Asian orogenic belt. Tectonics 22:1069, doi:10.1029/2002TC001484.

Xu, R. H.; Zhang, Y.; Xie, Y.; Vidal, P.; Arnaud, N.; Zhang,Q.; and Zhao, D. 1996. Isotopic geochemistry of plu-tonic rocks. In Pan, Y., ed. Geological evolution of theKarakoram and Kunlun Mountains. Beijing, Seismo-logical Press, p. 137–186.

Xu, R. H.; Zhang, Y. Q.; and Xie, Y. W. 1994. A discoveryof an early Paleozoic tectono-magmatic belt in thenorthern part of the Kunlun Mountains. Sci. Geol. Sin.29:313–328.

Yang, J. S.; Robinson, P. T.; Jiang, C. F.; and Xu, Z. Q.1996. Ophiolites of the Kunlun Mountains, China andtheir tectonic implications. Tectonophysics 258:215–213.

Yao, Y., and Hsu, K. J. 1994. Origin of the Kunlun Moun-tains by arc-arc and arc-continent collisions. IslandArc 3:75–89.

Yin, A., and Harrison, T. M. 2000. Geological evolutionof the Himalayan-Tibetan Orogen. Annu. Rev. EarthPlanet. Sci. 28:211–280.

Yin, A., and Nie, S. 1996. A Phanerozoic palinspasticreconstruction of China and its neighboring regions.In Yin, A., and Harrison, T. M., eds. The tectonic evo-lution of Asia. Cambridge, Cambridge UniversityPress, p. 442–485.

Yin, J. X., and Bian, Q. T. 1995. Geological map of theKarakoram–West Kunlun and adjacent regions. Bei-jing, Science Press.

Yuan, C. 1999. Magmatism and tectonic evolution of the

West Kunlun Mountains. PhD dissertation, Univer-sity of Hong Kong, 231 p.

Yuan, C.; Sun, M.; Yang, J. S.; Zhou, H.; and Zhou, M.-F. 2004. Nb-depleted, continental rift–related volca-nism of the Arkarz greenschist (West Kunlun): impli-cation for the break-up of the Tarim craton fromGondwanaland. In Malpas, J.; Fletcher, C. J.; Aitchi-son, J. C.; and Ali, J., eds. Aspects of the tectonic evo-lution of China. Geol. Soc. Lond. Spec. Publ. 226:131–144.

Yuan, C.; Sun, M.; Zhou, M.-F.; Xiao, W. J.; and Zhou,H. 2005. Geochemistry and petrogenesis of the YishakVolcanic Sequence, Kudi ophiolite, West Kunlun (NWChina): implications for the magmatic evolution in asubduction zone environment. Contrib. Mineral. Pet-rol. (forthcoming), doi: 10.1007/s00410-005-0012-0.

Yuan, C.; Sun, M.; Zhou, M.-F.; Zhou, H.; Xiao, W. J.; andLi, J. L. 2002a. Absence of the Archean basement forthe South Kunlun Block: Nd-Sr-O isotopic evidencefrom granitoids. Island Arc 12:13–21

———. 2002b. Tectonic evolution of the West Kunlun:geochronological and geochemical constraints fromKudi granitoids. Int. Geol. Rev. 44:653–669.

Yue, Y., and Liou, J. G. 1999. Two-stage evolution modelfor the Altyn Tagh fault, China. Geology 27:227–230.

Zhang, G. C. 1997. Mesozoic sedimentological and tec-tonic evolution of the West Kunlun–Karakoram fore-land fold-thrust belts. PhD dissertation, Institute ofGeology and Geophysics, Chinese Academy of Sci-ences, 137 p.

Zhang, Y. Q.; Xie, Y. W.; and Xu, R. H. 1996. Geochem-istry of granitoid rocks. In Pan, Y., ed. Geological evo-lution of the Karakoram and Kunlun Mountains. Bei-jing, Seismological Press, p. 94–136.

Zhao, X., and Coe, R. S. 1987. Paleomagnetic constraintson the collision and rotation of north and south China.Nature 327:141–143.

Zhou, D., and Graham, S. 1996. Extrusion of the AltynTagh wedge: a kinematic model for the Altyn TaghFault and palinspastic reconstruction of northernChina. Geology 24:427–430.

Zhou, H. 1998. The main ductile shear zone and thelithosphere effective elastic thickness of the W Kun-lun orogenic belt. PhD dissertation, Institute of Ge-ology and Geophysics, Chinese Academy of Sciences,103 p.

Zhou, H.; Chu, Z. Y.; Li, J. L.; Hou, Q. L.; Wang, Z. H.;and Fang, A. M. 2000. 40Ar/39Ar dating of ductile shearzone in Kuda, West Kunlun, Xinjiang. Sci. Geol. Sin.35:233–239 (in Chinese with English abstract).

Zhou, H.; Jiang, N.; Yan, Q.; and Li, J. L. 2001. Discoveryof anorthosite in the West Kunlun. China J. Geol. 36:490–491 (in Chinese with English abstract).

Zhou, H., and Li, J. L. 2000. Age and geochemical featuresof lamprophyres in Kuda, western Kunlun. Acta Pet-rol. Sin. 16:380–384 (in Chinese with Englishabstract).

accretionary tectonics of the western kunlun orogen, china...

Documents