revision of the cretaceous–paleogene stratigraphic …Ž‹成善.pdfrevision of the...

of 19 /19
Revision of the CretaceousPaleogene stratigraphic framework, facies architecture and provenance of the Xigaze forearc basin along the Yarlung Zangbo suture zone Chengshan Wang a, , Xianghui Li b , Zhifei Liu c , Yalin Li a , Luba Jansa d , Jingen Dai a , Yushuai Wei a a State Key Laboratory of Biogeology and Environmental Geology, Research Center of QinghaiTibet Plateau Geology, China University of Geoscience (Beijing), Beijing 100083, China b State Key Laboratory of Mineral Deposit Research, School of Earth Sciences and Engineering, Nanjing University, Nanjing 210093, China c State Key Laboratory of Marine Geology, Tongji University, Shanghai 200092, China d Department of Earth Sciences, Dalhousie University, Halifax, and Geologic Survey of Canada-Atlantic, Dartmouth, N.S., Canada abstract article info Article history: Received 16 March 2011 Received in revised form 22 August 2011 Accepted 8 September 2011 Available online 17 October 2011 Keywords: Cretaceous-Paleogene Xigaze Group Cuojiangding Group Forearc basin Yarlung Zangbo suture zone New eld investigation, structural analysis, review of biostratigraphy, stratigraphic framework, facies archi- tecture, provenance, and ow direction of the CretaceousPaleogene sedimentary strata in southern Tibet resulted in reinterpretation of the evolution of the Xigaze forearc basin (XFB). The XFB is comprised by ysch-dominant Xigaze Group (consisting of Ngamring, Chongdoi, and Sangzugang Formations) and of the shallow marine Cuojiangding Group (Padana, Qubeiya, Quxia, and Gyalaze Formations), with their ages here- in revised as the AlbianConiacian and the SantonianYpresian respectively. The deep-sea ysch represented by the Ngamring Formation constitutes the early sedimentary ll of the XFB. It was subdivided into ve megasequences MS1MS5 in order: the early submarine fan facies stage, channel submarine fan stage, the mature submarine fan stage, the highstand submarine fan stage, and outer fan-pelagic stage. The paleocur- rent ows were eastwards and westwards during the early and late stages of basin development and north- wards and southwards during its middle stage, indicating typical forearc basin sedimentation with longitudinal supply and lateral lling. Provenance studies point to multiple sources, with andesitic debris sourced from the JurassicLower Cretaceous Yeba Formation and Sangri Group formed in the volcanic island arc chains in the southern Gangdese. The four periods of the XFB development are: 1) the formation of the XFB basement by the Xigaze ophiolite, and completed before the Aptian. The coeval deposits on the shelf are the Sangzugang Formation carbonate; 2) the earliest XFB deposits of the earliest Aptian age (ca. 125/ 120 Ma) represented by the Chongdoi Formation, related to the residual forearc; 3) composite forearc devel- oped, represented by the late Albianlate Cenomanian ysch megasequences (MS1MS2) and the TuronianConiacian ysch megasequences (MS3MS5), respectively; and 4) the nal stage of the XFB development started the terminal forearc stage represented by the late Late CretaceousEarly Paleogene shallow marine Cuojiangding Group and terminated during middle Ypresian, when the Gyalaze Formation was deposited. © 2011 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved. 1. Introduction Extensive studies were carried out in the Yarlung Zangbo suture zone (YZSZ) (e.g. Allègre et al., 1984; Gao and Tang, 1984; Coulon et al., 1986; Gao, 1988; Dewey et al., 1990; Yu and Wang, 1990; Yin et al., 1994; Yin and Harrison, 2000; Liu et al., 2010; Dai et al., 2011b; etc.). However, only few were concerned with the XFB, and only few hypotheses on its evolution have been published (e.g. Dürr, 1993, 1996; Einsele et al., 1994; Wang et al., 1999), despite that the sedimentary strata appear to be well preserved and subaeri- ally exposed. As part of the YarlungZangbo subduction zone com- plex, the sediments of XFB document the late subduction processes of the Neo-Tethys Ocean crust and the forearc basin development. The biostratigraphy of XFB sedimentary strata has been established during the 1970s1980s (Wen, 1974; Cao, 1981; Wu, 1984; Liu et al., 1988), and lithofacies and provenance studies have been pub- lished by Liu et al. (1993), Dürr (1993, 1996), and Einsele et al. (1994). However, some recently published biostratigraphic data (e.g. Wan et al., 1998, 2001; Li et al., 2008a,b) and present works on structure and sedimentology demonstrate that previous hypotheses on the time, thickness and sequences of the XFB sediments must be revised. Consequently, new interpretations on facies architecture and provenance will result in a reinterpretation of the XFB evolution. To improve our understanding of the XFB development, we have con- ducted new geological eld investigations of the Xigaze Group. Newly described stratigraphic proles and outcrops, combined with structural analyses, were used to systematically calibrate the stratigraphy, to rein- terpret sedimentary megasequence and provenance, and especially to reconstruct the sedimentary evolution of the XFB as well as its relation- ships to the YZSZ on the basis of the revised stratigraphic framework. Gondwana Research 22 (2012) 415433 Corresponding author. E-mail address: [email protected] (C. Wang). 1342-937X/$ see front matter © 2011 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved. doi:10.1016/ Contents lists available at SciVerse ScienceDirect Gondwana Research journal homepage:

Author: phunghanh

Post on 20-Apr-2018




2 download

Embed Size (px)


  • Gondwana Research 22 (2012) 415433

    Contents lists available at SciVerse ScienceDirect

    Gondwana Research

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

    Revision of the CretaceousPaleogene stratigraphic framework, facies architectureand provenance of the Xigaze forearc basin along the Yarlung Zangbo suture zone

    Chengshan Wang a,, Xianghui Li b, Zhifei Liu c, Yalin Li a, Luba Jansa d, Jingen Dai a, Yushuai Wei a

    a State Key Laboratory of Biogeology and Environmental Geology, Research Center of QinghaiTibet Plateau Geology, China University of Geoscience (Beijing), Beijing 100083, Chinab State Key Laboratory of Mineral Deposit Research, School of Earth Sciences and Engineering, Nanjing University, Nanjing 210093, Chinac State Key Laboratory of Marine Geology, Tongji University, Shanghai 200092, Chinad Department of Earth Sciences, Dalhousie University, Halifax, and Geologic Survey of Canada-Atlantic, Dartmouth, N.S., Canada

    Corresponding author.E-mail address: [email protected] (C. Wang).

    1342-937X/$ see front matter 2011 International Adoi:10.1016/

    a b s t r a c t

    a r t i c l e i n f o

    Article history:Received 16 March 2011Received in revised form 22 August 2011Accepted 8 September 2011Available online 17 October 2011

    Keywords:Cretaceous-PaleogeneXigaze GroupCuojiangding GroupForearc basinYarlung Zangbo suture zone

    New field investigation, structural analysis, review of biostratigraphy, stratigraphic framework, facies archi-tecture, provenance, and flow direction of the CretaceousPaleogene sedimentary strata in southern Tibetresulted in reinterpretation of the evolution of the Xigaze forearc basin (XFB). The XFB is comprised byflysch-dominant Xigaze Group (consisting of Ngamring, Chongdoi, and Sangzugang Formations) and of theshallow marine Cuojiangding Group (Padana, Qubeiya, Quxia, and Gyalaze Formations), with their ages here-in revised as the AlbianConiacian and the SantonianYpresian respectively. The deep-sea flysch representedby the Ngamring Formation constitutes the early sedimentary fill of the XFB. It was subdivided into fivemegasequences MS1MS5 in order: the early submarine fan facies stage, channel submarine fan stage, themature submarine fan stage, the highstand submarine fan stage, and outer fan-pelagic stage. The paleocur-rent flows were eastwards and westwards during the early and late stages of basin development and north-wards and southwards during its middle stage, indicating typical forearc basin sedimentation withlongitudinal supply and lateral filling. Provenance studies point to multiple sources, with andesitic debrissourced from the JurassicLower Cretaceous Yeba Formation and Sangri Group formed in the volcanic islandarc chains in the southern Gangdese. The four periods of the XFB development are: 1) the formation of theXFB basement by the Xigaze ophiolite, and completed before the Aptian. The coeval deposits on the shelfare the Sangzugang Formation carbonate; 2) the earliest XFB deposits of the earliest Aptian age (ca. 125/120 Ma) represented by the Chongdoi Formation, related to the residual forearc; 3) composite forearc devel-oped, represented by the late Albianlate Cenomanian flysch megasequences (MS1MS2) and the TuronianConiacian flysch megasequences (MS3MS5), respectively; and 4) the final stage of the XFB developmentstarted the terminal forearc stage represented by the late Late CretaceousEarly Paleogene shallow marineCuojiangding Group and terminated during middle Ypresian, when the Gyalaze Formation was deposited.

    2011 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved.

    1. Introduction

    Extensive studies were carried out in the Yarlung Zangbo suturezone (YZSZ) (e.g. Allgre et al., 1984; Gao and Tang, 1984; Coulonet al., 1986; Gao, 1988; Dewey et al., 1990; Yu and Wang, 1990; Yinet al., 1994; Yin and Harrison, 2000; Liu et al., 2010; Dai et al.,2011b; etc.). However, only few were concerned with the XFB, andonly few hypotheses on its evolution have been published (e.g.Drr, 1993, 1996; Einsele et al., 1994; Wang et al., 1999), despitethat the sedimentary strata appear to be well preserved and subaeri-ally exposed. As part of the YarlungZangbo subduction zone com-plex, the sediments of XFB document the late subduction processesof the Neo-Tethys Ocean crust and the forearc basin development.

    ssociation for Gondwana Research.

    The biostratigraphy of XFB sedimentary strata has been establishedduring the 1970s1980s (Wen, 1974; Cao, 1981; Wu, 1984; Liu etal., 1988), and lithofacies and provenance studies have been pub-lished by Liu et al. (1993), Drr (1993, 1996), and Einsele et al.(1994). However, some recently published biostratigraphic data(e.g. Wan et al., 1998, 2001; Li et al., 2008a,b) and present works onstructure and sedimentology demonstrate that previous hypotheseson the time, thickness and sequences of the XFB sediments must berevised. Consequently, new interpretations on facies architectureand provenance will result in a reinterpretation of the XFB evolution.

    To improve our understanding of theXFBdevelopment,wehave con-ducted new geological field investigations of the Xigaze Group. Newlydescribed stratigraphic profiles and outcrops, combined with structuralanalyses, were used to systematically calibrate the stratigraphy, to rein-terpret sedimentary megasequence and provenance, and especially toreconstruct the sedimentary evolution of the XFB as well as its relation-ships to the YZSZ on the basis of the revised stratigraphic framework.

    Published by Elsevier B.V. All rights reserved.[email protected]://

  • 416 C. Wang et al. / Gondwana Research 22 (2012) 415433

    2. Geological background

    The YZSZ comprises four major tectonic units (from north tosouth): Gangdese magmatic arc of CretaceousTertiary plutonic andvolcanic rocks (GMA), Xigaze forearc basin (XFB), an ophiolite beltand a TriassicEocene accretionary wedge (Fig. 1).

    The GMA in southern Tibet, together with the Ladakh magmaticarc in the western Himalayas, comprise a large LadakhGangdesemagmatic belt over 2000 km along the strike (i.e. the Trans-Himalaya). The GMA consists of intrusive (including gabbro, diorite,tonalite, and granodiorite) and extrusive (including basalt, andesite,and rhyolite) rocks, in which the Cretaceous rocks with geochemicalcompositions of I-type granitoids and adakites are thought to havebeen generated by the northward subduction of the Neo-Tethyanoceanic floor (cf. Wen, 1974; Zhu et al., 2006, 2009a, 2011; Zhang etal., 2010). The intrusive rocks were emplaced over a long time periodfrom the Late Triassic (ca. 205 Ma) to Miocene (ca. 10 Ma) (cf. Chunget al., 2009; Ji et al., 2009; Lee et al., 2009; Zhu et al., 2011). The extru-sive rocks, represented by the volcanic exposures in the Early JurassicYeba Formation (cf. Zhu et al., 2008b), Late JurassicEarly CretaceousSangri Group (cf. Zhu et al., 2009a), and Early Paleogene LinzizongGroup (cf. Zhou et al., 2004; Mo et al., 2008), extend about 1500 kmin a WE direction in southern Tibet. The Gangdese thrust faultforms the contact between the GMA and the XFB in the eastern sec-tion (Yin and Harrison, 2000).

    The ophiolite belt is composed of ultramafic rocks with gabbro cu-mulates, sheeted dyke complexes and pillow lava as well asradiolarian-bearing chert. Radiolarians date the chert as of the lateAlbianearly Cenomanian age (Marcoux et al., 1982). The bulk of radio-metric data from ultrabasic and basic rocks yielded an age of 12010Ma (Gpel et al., 1984). It had previously been thought that theshortage of cumulates and sheeted dykes implies that the ophioliteswere product of an abnormal oceanic crust resulting from the slowlyspreading mid-ocean ridge (Nicolas et al., 1981; Xiao, 1984; Girardeauet al., 1985a,b). However, new evidence increasingly indicates that theophiolites come from the supra-subduction zone (SSZ) (Bdard et al.,2009; Bezard et al., 2011; Hbert et al., 2011; Dai et al., 2011b). TheRenbuZedang thrust fault in Xigaze is the contact between the ophio-lites and the southern accretionary wedge (Yin and Harrison, 2000).

    The accretionary wedge is a narrow elongate belt up to 1000 kmlong and 1050 km wide. It consists of rocks derived from the

    Fig. 1. Geological map of the IndusYarlung Zangbo suture zone showing location of the GMal. (2006), accretionary wedge after Wang et al. (1999). Inset: Simplified 1:2,500,000 geolo

    ophiolitic complex represented by highly deformed and serpenti-nized pyroxene peridotite with developed schistosity and by tectonicmlange (Gao and Tang, 1984; Gao, 1988). The accretionary wedgecrops out at Ngamring, Geding, Bailang, and Zedang (Fig. 1) where itis characterized by olistoliths of various sizes ranging up to several kilo-meters, emplacedwithin thematrix of deformed and fragmented flyschsandstone and shale that comprises the Upper TriassicCretaceousXiukang Group.

    The strata of the XFB crop out along the Yarlung Zangbo in aneastwest oriented belt, over 500 km long, which begins at Renbuin the east and ends at Zhongba in the west. The maximum width ofthe belt is 22 km. Our study area was focused on the western part ofthe Xigaze region (Figs. 1 and 2A), where the flysch-dominatedXigaze Group and shallowmarine Cuojiangding Group are subaeriallywell exposed. The Cuojiangding Group overlies the Xigaze Group inthe Zhongba area (Liu et al., 1988; Wan et al., 2001).

    3. Methods

    Three key methods were adopted in this study. The first methodwas review of past works. Previous data was refined and reinter-preted based on newly described data on biostratigraphy, isotopicchronology, petrology, sedimentary megasequence, paleocurrent di-rection, and provenance. The new results have been applied through-out most of the text for both the Xigaze Group and the CuojiangdingGroup.

    The second method was the field geological investigation. Duringfield work, observations and descriptions of profiles, measurementof paleocurrents, and sampling for thin-sections and fossils (Figs. 2and 3), were used to constrain the stratigraphy and to analyze theprovenance of the Ngamring Formation. Especially, five profiles previ-ously studied by Drr (1993) and Einsele et al. (1994) were selectedfor further re-examination: Zhaxigang, Geding, KardoiGyangqingze,NadangDeri, and XishanBeishan profiles (Fig. 2A). Several comple-mentary profiles to the south and west of Xigaze were also studied(Fig. 2A). Of these, the typical profile is the KardoiGyangqingze pro-file which presents the structure and stratigraphy of the NgamringFormation (Fig. 2B). Traditional structures were analyzed for the pro-file by the deformation and faulting approach, and the results ofstructural interpretation were completed in Section 4.

    A, XFB, ophiolite belt, and accretionary wedge. The Sangri Group location is after Zhu etgical map of Tibet (Pan et al., 2004).

  • Fig. 2. A. Location of profiles from field studies shown on a geological map of the middle XFB (modified after Wang et al., 1999; Einsele et al., 1994). 1. GMA complex; 2. QiuwuFormation; 3. Gyabulin Formation; 4. Sangzugang Formation; 5. Ngamring Formation; 6. granites; 7. ophiolites; 8. Liuqu Group.; 9. mlange; 10. Quaternary; 11. thrust fault;12. Structural profiles investigated in this work; 13. village and town; and 14. localities studied in detail. ZXS: Zhaxigang profile; GDS: Geding profile; GKS: KardoiGyangqingze profile(modified after Drr, 1993: see panel B); DNS:NadangDeri profile; XGS: XigazeXishanNanshan profile. B. Interpreted structure of the KardoiGyangqingze profile (modified after Drr,1993; Einsele et al., 1994). Note the asymmetry, deformation and dipping of the two flanks. Megasequences MS1MS5: for description see text. S1S65: locations of field studies. Short-ening ratio (e) is from Ratschbacher et al. (1992).

    417C. Wang et al. / Gondwana Research 22 (2012) 415433

    The third method was the use of aerial photos at 1:75,000 scale,which in addition to tectonic features studied in the field allowed to lo-cate and map regional distribution of lithological markers (Fig. 3) andmegasequences in the Ngamring Formation. This process focused ontwo lithologicalmarkers: a sandstoneconglomerate assemblagemark-er that appears as steeply sloping, high, sharp ridges that are darkish incolor, indicating the channel-fills of the submarine fan system; and a pe-lagicmarlstonemarker, which is identifiable on the images as a narrow,

    Fig. 3. Locations of samples on geological map (simplified after Einsele et al., 1994 and Wanal., 1998). 2. Numbers and localities of ammonite fossils: (a) and (d) after Wiedmann anddetrital zircon samples (after Wu et al., 2010). 4. Limestone layers (modified after Drr, 1993

    lighter-colored stripe. All the results of the lithological markers areshown in Fig. 3.

    4. Structures of Ngamring Formation

    The Ngamring Formation forms a synclinorium striking WE(Fig. 2). It is asymmetric, with the northern flank 1015 km wideand southern flank 68 km wide (Fig. 2B). This asymmetry may

    et al., 1998). 1. Numbers and localities of the foraminiferal fossil samples (after Wan etDrr (1995); (b) after Xiao (1984); (c) after Wen (1974). 3. Numbers and localities ofand Einsele et al., 1994); solid line: observed, dashed line: inferred; 5. village and town.

    image of Fig.2image of Fig.3

  • 418 C. Wang et al. / Gondwana Research 22 (2012) 415433

    have been result of difference in the thicknesses of sediments depos-ited in the distal and proximal part of the basin and/or by a northwardoverriding thrust on the southern flank, and/or by the lack of theAptianAlbian sediments on the southern flank.

    The structural deformation increases northward as shown by theprofile (Fig. 2B). To the south, the folds are open, associated withsmaller sliding folds striking westeast, commonly having axial cleav-age evident in the weak layers. In the center of the synclinorium thefolds become small-scale subordinate folds with axial cleavages, andstrata are horizontal; in the north, the strata are strongly deformedwith tightly-closed, imbricated folds.

    All strata strike WE as result of contraction in the NS direction.The folds are asymmetric, with the structural style observed along thestrike at both, the southern and northern flanks of the basin. The foldaxial surfaces dip to the south at the northern flank, and northward atthe southern flank (Fig. 2B). Pale strips on images which mark pelagicmarlstones outcrop near the center of the synclinorium. Two of theseidentified on the southern flank and one at the northern flank belongto the same layer (Figs. 2B and 3), as documented by the foraminiferalbiostratigraphy (Wan et al., 1998). Therefore, they are not two differ-ent horizon as previously indicated by Einsele et al. (1994).

    A set of imbricated thrusts are located at the northern boundary ofthe synclinorium (Fig. 2B), where the mid-Cretaceous calcareousSangzugang Formation and the Tertiary terrestrial Gyabulin Forma-tion are present in the thrust (Yin et al., 1994). At the southernboundary area, the contact of the Ngamring Formation is less clearthan in the other areas, as most of the outcrop is covered by theQuaternary sediments. However, sedimentary contact of the Ngamr-ing flysch with the underlying pillow lava of the ophiolitic seriescan be seen in the Dazhu area. Also at Qunrang and Naxia villages,the flysch overlies the Chongdoi Formation radiolarian cherts. An al-most vertical fault separates the Tertiary terrestrial Liuqu Group andNgamring (Yin et al., 1988a) in the KardoiGyangqingze profile.

    5. Revision of stratigraphy

    Verification of previously established litho-, bio-, and chronostra-tigraphy for the XFB is critical for basin analysis. Below we presentnew data that will revise previous studies re-interpret the ages andrelationships of the stratigraphic units (Fig. 4). Two main lithostrati-graphic associations represent the XFB fill. They are the XigazeGroup dominated by flysch of the Ngamring Formation, and the shal-low marine Cuojiangding Group.

    5.1. Xigaze Group

    The Xigaze Group was originally defined by Wen (1974) as sedi-mentary flysch about 6000 m thick, unconformably overlying theYanshanian granites with basal conglomerates of CenomanianTuronianage at its base. Later, the shallow marine carbonates of the SangzugangFormation were included into this group (Wu et al., 1977). In thispaper, besides the Ngamring Formation and Sangzugang Formation,the Chongdoi Formation is included as part of the Xigaze Group (Fig. 4).

    5.1.1. Chongdoi FormationThis formation was named by Cao (1981) after Chongdoi village,

    about 20 km SE of Xigaze, where it is composed by two members.The lower member is ophiolite breccia interbedded with purplish-red radiolarian chert. The upper member is tuffaceous siltstone inter-bedded with shale that ranges in thickness from 73 to 197 m (XizangBGMR, 1993, 1997). Its basal contact is depositional upon the ophio-lite (Fig. 5) and pillow lava and its upper contact is conformablewith the overlying flysch of the Ngamring Formation (Fig. 5). InNaxia locality, the planktonic foraminifer Rotalipora spp. was foundin the limestone at the top of the Chongdoi Formation (Wu, 1984)which indicates that it was deposited no later than Cenomanian.

    This age was also confirmed by radiolarian association found in theChongdoi (Wu and Li, 1985; Wu, 1986), which together with the de-trital zircon UPb isotope age of 116 Ma (Wu et al., 2010) from thesiltstone in the upper of the Chongdoi Formation indicates that thestrata are late Aptian to the early Cenomanian in age.

    Previously, the Chongdoi Formation has been thought to be a part ofan oceanic arc system (e.g. Cao, 1981; Xizang BGMR, 1997; Aitchisonet al., 2000; Ziabrev et al., 2004). However, in this paperwe have includ-ed it into the Xigaze Group instead of the oceanic crust association forthree reasons: (1) The Chongdoi is conformable with the overlyingflyschoid Ngamring Formation (Aitchison et al., 2000; Ziabrev et al.,2004), which we have also observed during our field investigation;(2) Limestone fragments of the Sangzugang Formation are found inthe Chongdoi, as well as in the Ngamring Formation (Einsele et al.,1994; Drr, 1996;Wang et al., 1999); and (3) This assemblage is similarto the sequence of radiolarian chert and clastic rocks in the forearc basinof the Great Valley, western USA, which is an example of the typicalforearc basin sequence.

    5.1.2. Sangzugang FormationThe name of this formation is after the village of Sangzugang, Sajia

    County, east of Xigaze, where it crops out as tectonic fragments. Itconsists of dark gray, thick-bedded bioclastic limestones with abun-dant foraminifera and rudists, indicating shallow carbonate platformfacies. The unit is 60230 m thick (Wu et al., 1977), and decreasingto a few meters thickness about 10 km east of Xigaze, and disappearsfarther east at Dazhuka. It has fault contacts with either the NgamringFormation or the Gyabulin Formation. The Sangzugang is of AptianAlbian age (Cherchi and Schroeder, 1980; Bassoullet et al., 1984).

    5.1.3. Ngamring FormationThe namewas used byWu et al. (1977) to describe Cretaceous deep-

    sea flysch in Ngamring County. Later it was redefined as the Cretaceousdeep-sea flysch occurring above the carbonate Sangzugang Formation(Yin et al., 1988b; Xizang BGMR, 1993, 1997). The Ngamring Formationis composed by sandstone and shale, and is characterized by a series oflarge channelized conglomerates in the lower part. It is conformablewith the underlying Chongdoi Formation (Fig. 5) as well as the overly-ing Padana Formation, west of Ngamring County (Fig. 6). But it has faultcontact with the Xigaze ophiolite in the south and with the GyabulinFormation or Sangzugang Formation in the north (Fig. 2A). However,the stratigraphic sequence, thickness and dating of the Ngamring hasbeen misinterpreted, mainly due to the strongly deformed strata andthe scarcity of fossils. Below they are calibrated on the combination ofthe recently published biota and age maximum of detrital zircon UPbisotope plus the present observation and structural interpretation inSection 4.

    It is generally accepted that the structural analyses shows thatrocks of the Ngamring Formation become progressively older fromthe synclinorium axis towards the southern and northern flanks.However, there are problems associated with the interpretation ofstratigraphic sequence and thicknesses. From the axial center of thebasin towards the north, the Ngamring Formation was dated as ofthe late Turonian age (samples 46-1, 46-2), and the middle-earlyTuronian (samples 47-3, XB-4, XB-5) (Fig. 3) using foraminiferalmicrofossils, and at eastward locations by the occurrence of ammo-nite Mammites sp. (Wen, 1974). Farther to the north, the rockswere dated as Cenomanian by the ammonite Turrilites (T.) acutus(Wiedmann and Drr, 1995). The AptianAlbian age is indicated bythe occurrences of larger benthic foraminifera Orbitolina (Wan et al.,1998) in the sample 14-1 (Fig. 3) and of an ammonite Neophlycticeras(T.) brottianum (Wiedmann and Drr, 1995). On the northern bound-ary, the Sangzugang Formation was dated by foraminifera as ofAptianAlbian age (Cherchi and Schroeder, 1980; Bassoullet et al.,1984), indirectly indicating that the deposition of the Ngamringcould not have started prior to the AptianAlbian.

  • Fig. 4. Diagram showing the stratigraphic framework of the XFB with data of biostratigraphy, maximum age of detrital zircon UPb isotopes, and lithofacies. The chronostratigraphicsystem is cited from Gradstein et al., 2008. The column of lithostratigraphy and lithology is composite from the review (details see in text). In the column of the biostratigraphy, thefauna of the Cuojiangding Group are quoted fromWan et al. (2001), and the biozones of the Ngamring Formation are cited fromWan et al. (1998). The column of the age maximumis cited from Wu et al., 2010.

    419C. Wang et al. / Gondwana Research 22 (2012) 415433

    Most of the Ngamring Formation in the southern KardoiGyangqingze profile (Fig. 2) are of Cenomanian age based on the fora-minifera Rotalipora cushmani (sample 43-11) (Caron, 1985). The pres-ence of an ammonite from Brancoceratidae Family found to the west

    of Sagoi indicates the AlbianCenomanian age of the strata (Xiao,1984). Foraminifera and radiolarian fossils in samples 43-16 and43-17 (Fig. 3) indicate the early Late Cretaceous age (Yin et al.,1988b). Abundant foraminiferal assemblage (e.g. Whiteinella

    image of Fig.4

  • Fig. 5. Field photograph showing the sequence from the ophiolite belt to the XFB at Qiongrang village (290934E, 890157N), SE Xigaze. A. Reversed stratigraphic sequence;folded succession of the Chongdoi Formation (rhythmic turbidite sandstone and shale Bouma-cde, shale with ash intercalation, red thin radiolarian cherts) up to the ophiolites.B. Enlargement of contact relationship between radiolarian cherts and ophiolites. The figures in the photograph are shown at the contact.

    420 C. Wang et al. / Gondwana Research 22 (2012) 415433

    archaeocretacea, Marginotruncana renzi, M. pilediformis) in samples44-4 and 44-5 (Fig. 3) indicate deposition during the CenomanianTuronian (Wan et al., 1998). Similarly, foraminifera microfossils insamples ND-15 and ND-16 (Fig. 3), south of the NadangDeri profile(Fig. 2), indicate the Turonian age (Wan et al., 1998). Unfortunately, theSangzugang Formation has not yet been observed on the southernflank of the basin, as only the CenomanianTuronian component of theNgamring Formation crops out on the southernflank of the synclinorium.

    The youngest strata of Ngamring Formation are of the Coniacianage and are present at the axial part of the synclinorium. The forami-niferal assemblage M. sinuosa in sample 46-2, between samples 44-5and 47-3 (Fig. 3) indicates an age younger than the late Turonian(Wan et al., 1998). To the east, the foraminifera Dicarinella primitivain sample XB-1 at the Xiaburi profile (Fig. 2) is an index species ofthe early Coniacian (Caron, 1985). D. concavata is very abundant insample XB-3, indicating the late Coniacian age (Wan et al., 1998).

    The thickness of the Ngamring Formation on the northern flank hasbeen revised to approximately 4100 m down from 6000 to 8000 m pre-viously estimated by Einsele et al. (1994), and it changes to 10004100 m thick in the region. The biostratigraphic data indicatesAptianConiacian age for the Ngamring Formation (Wan et al., 1998).Considering the age maxima of detrital zircon UPb isotopes: 107 Ma,101 Ma, and 88 Ma from the lower, middle, and upper part of this for-mation (Wu et al., 2010), the age of the Ngamring Formation can be re-stricted to late AlbianConiacian in age (ca. 106 to about 86 Ma).

    5.2. Cuojiangding Group

    The Cuojiangding Group which overlies the Ngamring flysh is in-cluded here as part of XFB sedimentary fill to provides auxiliary infor-mation about forearc basin development. The name of this Group isderived after a small lake (ca. 5400 m high) approximately 40 kmnorth of the new Zhongba town. The group consists of the PadanaFormation, Qubeiya Formation, Quxia Formation, and the GyalazeFormation (Fig. 4), which crop out to the west of Ngamring county.It has a total thickness of 1472 m and is comprised of predominantlyby shallow marine sediments (Liu et al., 1988) with their age rangingfrom the Late Cretaceous to the Early Paleogene (Qian et al., 1982; Liuet al., 1988; Wan et al., 2001; Li et al., 2008a,b).

    In previous studies the Padana Formation and Qubeiya Formationhave been included in theXigazeGroup. However, both of themwere de-posited in a shallowmarine environment (details in Section 9), thereforethe depositional setting is very different from the deep-sea flysch of theXigaze Group. In addition, they are similar to the Quxia Formation andGyalaze Formation of the Cuojiangding Group. Therefore based on thelithofacies similarity we have changed the stratigraphic classification ofthe Padana andQubeiya and included them into the Cuojiangding Group.

    5.2.1. Padana FormationThe Padana Formation was named after the Padana valley south-

    east of Sangsang village (Liu et al., 1988). The lower part of this

    image of Fig.5

  • 421C. Wang et al. / Gondwana Research 22 (2012) 415433

    formation is comprised by gray sandstone interbedded with shale andin the upper part by red shale interbedded with gray sandstone andshale. The thickness of this formation ranges from about 733 to2000 m (Xizang BGMR, 1997). It has a conformable contact withboth the underlying Ngamring Formation (Fig. 6) and the overlyingQubeiya Formation. No age diagnostic fossils were found in thePadana Formation, therefore the Santonian age suggested is an inter-pretation from the age of underlying Ngamring Formation and over-lying the Qubeiya Formation as well as a maximum age of 88 Maderived from detrital zircon UPb isotopes (Wu et al., 2010).

    5.2.2. Qubeiya FormationThe Qubeiya Formation, named at the same place as the Padana

    Formation, is composed of yellowish-gray blocky sandy wackestoneintercalated with sandstone, shale, mudstone and packstone. It is430 m thick with the thickness increasing eastward up to ca.1000 m in Saga County (Xizang BGMR, 1997). In the Zhongba area,it has gradational contact with underlying Padana Formation and ahiatus separates it from the overlying Gyalaze Formation. Thebenthic assemblages (Fig. 4) represented by foraminifera, bivalves,ammonites, gastropods, and crinoids (Qian et al., 1982; Liu et al.,1988; Wan et al., 2001), plus the maximum 78 Ma age of detritalzircon UPb isotope from an indeterminate stratigraphic interval inthe Qubeiya (Wu et al., 2010), indicate a CampanianMaastrichtianage of the strata.

    5.2.3. Quxia FormationThe Quxia Formation, named at Quxia valley in the Cuojiangding

    area, is dominated by conglomerates with shale and sandstone inter-calations. At the typical section the formation is 107 m thick and hasparaconformable contact with the underlying Qubeiya Formation andconformable contact with the overlying Gyalaze Formation (Liu et al.,1988; Wan et al., 2001). Similar to the Padana Formation, no age-diagnostic fossils have been found in the formation since it was estab-lished. Therefore, its age has been assumed to be Danian (Wan et al.,2001).

    5.2.4. Gyalaze FormationThe Gyalaze Formation has been also named at Cuojiangding,

    where it is composed of gray and dark gray foraminiferal wackestoneand packstone interbedded with sandstone and conglomerate. Thethickness is over 145 m at the typical section, and the top of the for-mation is at the center of a syncline so its contact with the overlyingunit is unclear. The age is relatively better constrained than the agesof other units of the Cuojiangding Group because foraminifera areabundant in the formation. Assemblages of the large foraminiferalMiscellaneaDaviesi and NummulitesDiscocyclina (Wan et al., 2001)

    Fig. 6. Photograph showing the conformable contact between the Padana Formation ofthe Cuojiangding Group and Ngamring Formation of the Xigaze Group, in the Padanagroove (292132.8N, 864352.8E), Ngamring area. For scale, see figures circled atlower right-hand corner.

    found in the strata indicates their late Paleocene to early Eocene(Selandianearly Ypresian, ca. 6151 Ma) age. Because the top ofthe stratum is not exposed, its age may extend to the latest Ypresian,i.e. 5049 Ma, even younger.

    6. Megasequences of Ngamring Formation

    Before discussing the evolution of the XFB, we provide revision ofmegasequence of Ngamring Formation as it has impact on the basin fill-ing process interpretation. Previous studies documented that theNgamring Formation sediments were deposited in a submarine fan sys-tem (e.g. Drr, 1993; Liu et al., 1993; Einsele et al., 1994). To better de-scribe the submarine fan system, seven basic lithofacies terms (A, B, C,D, E, F, G) of Shanmugam and Moiola (1985) were used here. Allabyand Allaby (1999) defined a megasequence as comprised by unitsdeposited during a distinct phase of basin evolution, separated by themajor unconformities that mark changes in the basin-controlling sedi-mentary deposition processes. We have distinguished the megase-quences of the Ngamring chiefly from lithological associations becauseof the difficulty in recognizing unconformities in a deep-sea basin envi-ronmental setting.

    Combining structural analysis and revised stratigraphy of theNgamring Formation, we distinguish five megasequences, termedMS1 to MS5 (Figs. 2B, 7, 8), in difference to the three megasequencessuggested indirectly by Einsele et al. (1994). The five flyschoid mega-sequences we have recognized have different distributions in space,according to their proximal/distal positions in the basin (Fig. 8). Forexample, MS2 and MS3 are confined to the western area of the prox-imal northern flank of XFB, where lateral lithofacies lack channelizedsandstoneconglomerate beds. In addition, MS3 is not easily separa-ble from MS4 because both were deposited in a distal basin environ-ment where sediments were much thinner than in the proximal partof the basin.

    6.1. MS1, early submarine fan megasequence

    MS1 occurs on both flanks of the synclinorium (Fig. 2B) and appearsto be truncated by faults in the southeastern part of the XFB. On thesouthern flank, MS1 crops out between Kardoi and Geding where it isabout 23 km wide and 300 m thick and is in fault contact with theophiolites (Fig. 8) and has a sharp contact with the MS2 channel con-glomerate (Fig. 7). On the northern flank themegasequence iswell pre-served in an outcrop zone about 5 km wide, and it is in fault contactwith the Gyabulin Formation and/or Sangzugang Formation. It can besubdivided into two sub-megasequences MS11 and MS12.

    6.1.1. MS11, slope apron facies sub-megasequenceThis sub-megasequence is comprised by slope facies C and D

    (mudstonesandstone), intercalated with facies F (debris-flow brec-cias and gravelly shale). Facies C and D are comprised by alternatinggraded, fine-grained sandstones, siltstones and shales that drape theslope. Thin black siliceous limestone also occurs as intercalations,probably formed during interruption in terrigenous supply into thebasin. A characteristic feature of this sub-megasequence is the pres-ence of facies F (debris-flow sediments).

    On the northern flank, facies F was found in the profiles atGyangqingze (sample S14) and Deri (S59). There are two debris-flow layers in the northern GyangqingzeKardoi profile at Gyalie vil-lage (Fig. 2A): the first layer is debris-flow and slump sedimentsabout 46 m thick, enclosing giant olistoliths up to 30 m15 m insize; the second layer, about 20 m thick, cropping out north of thethrust slice, contains some limestone gravels from the SangzugangFormation. Also some Orbitolina fragments were found in sampleS14, located about 3 km east of Gyangqingze village. The up-sectionprofile is composed of back shale (5 m thick), deformed sandstone(7 m) and paraconglomerate (10 m). The paraconglomerate gravels

    image of Fig.6

  • Fig. 7. Vertical stratigraphic diagram of sedimentary sequence of flysch, Ngamring Formation of Xigaze Gp. Sample numbers of fossils as in Fig. 2. Note the thickness discrepancybetween southern and northern flanks. 1. Shale; 2. mudrock; 3. sandstone and mudrock; 4. sandstone; 5. conglomerate; 6. marlstone; and 7. limestone.

    422 C. Wang et al. / Gondwana Research 22 (2012) 415433

    are predominantly from basic volcanics, granite and sedimentaryrock. Pillow structures and chilled borders can be observed in someof the gravels.

    Fig. 8. Lateral extent of Xigaze Group submarine fan megasequences. Geological backgroun2. megasequence MS0; 3. megasequence MS1; 4. megasequence MS2; 5. megasequence Mand MS4; 9. megasequences MS2, MS3 and MS4; 10. megasequence MS5; 11. axis of synet al., 1994); other symbols as in Fig. 2A.

    On the southern flank, the counterpart of MS11 is the lower-middle Chongdoi Formation in Naxia. Because it is located fartherfrom the source, the facies are lenticularly distributed gravity-flow

    d map as in Fig. 2. Note the axis of the synclinorium. 1. Megasequences and numbers;S3; 6. megasequences MS2 and MS3; 7. megasequence MS4; 8. megasequences MS3clinorium; 12. village and town; and 13. localities of supply channels (after Einsele

    image of Fig.8image of Fig.7

  • 423C. Wang et al. / Gondwana Research 22 (2012) 415433

    sandstone, forming lenses less than 20 m long and representing chan-nelized turbidities of facies B. This facies was deposited on the north-ern slope of the accretionary wedge in the southern XFB. Orbitoliniafragments found on both the northern and southern flanks of the syn-clinorium and associated late Albian ammonite Neophlycticears spp.7 km SSW of Gyangqingze village (Wiedmann and Drr, 1995) indi-cate that sub-megasequence MS11 was deposited during the lateAlbian, probably up-to the Cenomanian.

    6.1.2. MS12, overbank facies sub-megasequenceMS12 is dominated by fine-grained, non-channelized sediments

    with many small-scale coarse trough fills (facies B), and lacks thelarge-scale channels of MS2. This implies that there was no point-source supply from the continent during formation of MS12, whichwas related by Drr (1993) to an infant submarine fan stage.

    A characteristic feature of this sub-megasequence is that no coarseclasts are observed on either the southern or northern flanks. Both fa-cies D and G occurring between Naxia and Kardoi villages on thesouthern flank of the synclinorium are comprised by fine-grained tur-biditic sandstones and pelagic sediments, indicating a relatively sta-ble tectonic period. A slump layer approximately 50 m thick appearsin the upper part of facies D. Slumps, sandy pillows and convolutionsassociated with this layer suggest deposition on the lower-slope. Theslumped layer is overlain by a calcareous layer 34 m thick, which sug-gests a lower terrigenous supply due to rising sea level. The MS12 istruncated by the channel conglomerate of the MS2. Facies F is alsofound in the upper sub-megasequence MS12 on the southern flank(Fig. 7), differing from the MS11 in that it is not associated with chan-neled facies. The gravel within this facies is mostly angular and of var-iable composition, indicating a proximal source from the slope apronduring earlier basin development. Facies F outcrops are locally foundin MS12 throughout the XFB. On the northern flank, MS12 is charac-terized by the presence of overbank and interchannel facies, that is,by facies C and D comprising graded beddings of centimeter-thickfine-grained sandstone/siltstone and shale (Tc-e or Te-f). These twofacies extend laterally for up to 100 m. Locally, medium-thick(10100 cm) coarse greywackes with suspended muddy gravelsoccur below facies C, which could be misinterpreted as either widechannel or channelized high-density turbidity. However, we havenot yet found in the field any main channel or supply channel for sed-iments in the MS12. We thus interpret this facies association as over-bank and interchannel facies, with the channels being small, similarto those of the Eocene Hecho Group in the Pyrenees (Mutti, 1977;in Geding by Drr, 1996).

    MS12 is dominated by fine-grained, non-channelized sedimentswith many small-scale coarse trough fills (facies B) but lacking thelarge-scale channels of MS2. This implies that there was no point-source supply from the continent during formation of MS12, whichcorresponds to an infant submarine fan (Drr, 1993).

    Presence of the planktonic foraminifera R. cushmani at the top ofMS12 (Wan et al., 1998) indicates a Cenomanian age for MS1 and MS2.

    6.2. MS2, channeled submarine fan megasequence

    MS2 is easily recognized in the field and distinguishable from theunderlying MS1 in the XFB (Fig. 7), being very thick and dominatedby coarse-grained terrigenous clastics. In the field it is recognizableby the strata forming high, steep relief. Five sets of channel systemsat 2 to 10 km spacing were identified by mapping and remote-sensing image interpretation within a westeast oriented zone120 km long. The MS2 in the Kardoi and Xigaze submarine fan sys-tems is described in detail below.

    6.2.1. Channel system of Xigaze submarine fanOutcrops of the channel system of the Xigaze submarine fan are

    mainly present 1020 km west and east of Xigaze town. The widest

    channel is about 16 km8 km west of the town, and about 8 km tothe east. The total thickness of the channel system is about 1250 m,with a single channel fill up to 100 m thick, making it one of the larg-est submarine fans reported in geological history (Shanmugam andMoiola, 1985).

    Five channel systems represented by sandstoneconglomeratebodies were identified in the Xigaze region: Quarry Ridge, MinorRidge, Castle Ridge, Monastery Ridge, andWest Ridge (Fig. 9A). Litho-facies in Quarry Ridge change upwards from abundant channel sand-stoneconglomerate to interchannel fine-grained sediment depositedon the lower slope: channel sandstoneconglomerate, transitionalfine-grained intrachannel sediment, sandstone and shale betweenmid-fan and lobe, to middle-outer fan sediment. The Minor Ridge sys-tem comprises channel sandstoneconglomerate and interchannelsediment was deposited on the slope by episodic supply from theshelf. The Castle Ridge system appears to be a channelized sandstoneconglomerate composition overlain with an uneven surface by inter-channel slope sediments (Figs. 9B, 10); the Monastery Ridge system iscomposed by superimposed channel fills and slope facies, and theWest Ridge system consists of two channel-filled sequences.

    6.2.2. Channel system of Kardoi submarine fanA 3 km-wide channel sequence of the Kardoi Channel system ex-

    posed near Kardoi, south of the synclinorium, is the only evidenceof this system deposited in a distal basin environment. They are cov-ered by Quaternary sediments. Despite that, two sets of channel sand-stone and conglomerate sequences can be distinguished asconstructing the North and South Kardoi Ridges.

    The South Kardoi Ridge is composed of channel facies approxi-mately 82 m thick, displaying fining-upward and thinning-upwardstrata. Six superimposed channel sandstoneconglomerate bodiesrepresent six incisions, one of which is 2 m deep. A lateral progressionwas observed in the first channel body present in the lower part ofthe channel sequence, indicating eastward migration of the channelaxis. A gravel layer partly composed of boulders was found at thebase of the channel system. The largest boulder is about 1.3 m in di-ameter, with flute casts up to 0.4 m wide and 0.15 m deep. Flutecasts and the imbrication orientation of the gravel indicate depositionby a SW-flowing paleocurrent.

    It has been puzzling that the sandstone and conglomerate of theSouth Kardoi Ridge are located within the distal basin lithofacies.Two possible explanations can be suggested: 1) the submarine fanswere built directly onto the basin facies when the sea level dramati-cally fell below the shelf break, resulting in exposure of continentalshelf; and 2) the proximal channel coarse sediments directly over-lapped on the distal basin facies due to basinward migration of itsdepocenter. We prefer the first explanation, which implies that thebasin center was located farther to the south and close to the SouthKardoi Ridge.

    6.3. MS3, mature submarine fan megasequence

    The MS3 megasequence is present on both flanks of the synclinor-ium. At the southern flank it merges with the MS4 between Kardoivillage and Sino-Nepal highway, where it is very difficult to separatethem (Fig. 8) apart; and at the northern flank, MS3 becomes narrow,incorporating MS4 to the west of Geding village.

    This megasequence is dominated by middle-outer fan facies B andG, with a few occurrences of basin facies C and D. It represents an as-semblage of interchannel, overbank and channel sediments similar toMS11-but, despite these similarities, the style of channel superimpo-sition is different, indicating northward source of the sediments.

    The lithofacies differ between the southern and northern flanksbecause of the difference in the distance from the source. In thenorth, MS3 facies change from retrogradation to progradation: faciesB (branched channel) and facies C (lobe) grade into thinly bedded,

  • Fig. 9. A. Channel systems in MS2 of lower Ngamring Formation west of Xigaze town. 1. Rhythmic alternation of sandstone and shale; 2. channel conglomerate; 3. town district;4. highway; and 5. Quaternary. B. Sketch of the lithology at Castle Ridge at eastern terminal, NW corner of Xigaze town. Note the slumped sandstone block. The largest cobble is>80 cm in diameter.

    424 C. Wang et al. / Gondwana Research 22 (2012) 415433

    fine-grained turbidite-shale and medium-bedded lobe sandstone,with thinly bedded pelagic limestone at the top. In the south, MS3consists mainly of facies C and D (outer fan-basin). Towards the topof the megasequence, sediments have a pelagic character and thinlybedded limestone occurs in the Kardoi profile.

    6.4. MS4, highstand submarine fan megasequence

    This megasequence is well exposed on the northern flank, butmerges with the MS3 on the southern flank, partly as result offaulting.

    MS4 is characterized by pelagic sediments, which can be subdividedinto three sequences. The lower sequence is dominated by facies Bbranched channel sandstone and facies C lobe sandstone (Fig. 11). To-ward the south the sediments become progressively finer, as noted insouthern Xigaze (Nanshan). In themiddle ofMS4, the strata are gradingabruptly to pelagic carbonate, showing as a pale stripe on remote-sensing images and indicating a scarcity of outer fan-basin facies Cand D between the lower andmiddle sequence, perhaps due to eustaticsea level change. In the upper sequence, the strata are typical pelagicbasin facies, that is, rhythmically interbedded black/siliceous shale, sili-ceous marlstone and siltstone, alternating on a centimeter scale, andenclosing few siliceous concretions. The lack of carbonates suggests

    image of Fig.9

  • Fig. 10. Lithostratigraphic column of the sedimentary sequence at the Castle Ridge in the MS2.

    425C. Wang et al. / Gondwana Research 22 (2012) 415433

    deepening of the basin,with deposition perhaps occurring below calcitecompensation depth (CCD).

    6.5. MS5, outer fan and basin megasequence

    Megasequence MS5 occurs in the core of the synclinorium (Fig. 8),where it comprises the following three sequences:

    The lowest sequence, about 240 m thick, consists of facies B and Csandstones. It can be further subdivided into three sub-sequences,the lowest of which contains many coarsening-upwards andthickening-upwards cycles, with poorly sorted mostly matrix-supported sandstones, containing muddy intraclasts. The middlesub-sequence contains beds that are fining-upwards and thinning-upwards, resulting from retrogradation and abandoned of the chan-nels. The uppermost of the sub-sequences is composed of dark shaleinterbedded with facies E (thin sandstone), representing overbankor basin-outer fan sediments.The middle sequence is about 55 m thick and is comprised by faciesB sandstones. Sandstones in the channels have an erosional base,with the source being proximal to the north.In the uppermost sequence, facies C and D sandstones are interca-lated with shale. The occurrence of subfacies Tc-e indicates devel-opment of an outer fan facies. Foraminifera microfossils (Wanet al., 1998) suggest Coniacian age for MS5, which thus representsthe youngest deep-sea flysch deposit of the Ngamring Formationof the Xigaze Group.

    7. Basement of XFB

    Previous geologic studies suggested that the Xigaze ophioliteforms the basement of the XFB (e.g. Einsele et al., 1994; Wang et al.,1999). Results of new studies concur with this interpretation, but in-dicate that their age, nature, relation to the Neo-Tethys ocean crust, ismore complex and less certain. For the convenience of discussion, thecentral segment of the YZSZ will here be called the Xigaze ophioliticmassif, to include the Dazhuka, Bailang, Geding, Lhaze, Ngamring,Sangsang, and Saga submassifs to the west.

    In the 1980s, the Xigaze ophiolitic massif was thought to be a rem-nant of oceanic lithosphere formed at the spreading center (e.g. Nicolaset al., 1981; Allgre et al., 1984; Girardeau et al., 1985a,b; Girardeau andMercier, 1988; Pearce and Deng, 1988) but later studies have indicatedthat the Xigaze ophiolites were formed in the supra-subduction zone(SSZ) (e.g. Aitchison et al., 2000; Wang et al., 2000; Huot et al., 2002;Hbert et al., 2003; Xia et al., 2003; Ziabrev et al., 2003; Dubois-Ctet al., 2005; Dupuis et al., 2005; Guilmette et al., 2008, 2009; Bdardet al., 2009). Recent studies note that the Xigaze ophiolitic massifshave different ages and tectonic settings.

    In the Dazhuka submassif, a zircon UPb age of 1261.5 Mareported from quartz diorite (Malpas et al., 2003) is consistent withthe late Barremianlate Aptian radiolarian age (Ziabrev et al., 2003).Geochemical studies of the same submassif indicate that it wasformed in the SSZ environment (Xia et al., 2003). Subsequent studiesof mafic rocks support this conclusion, and it was proposed that themafics were formed in the back-arc basin spreading center (Dubois-Ct et al., 2005).

    image of Fig.10

  • Fig. 11. Photograph showing typical sequence of lobe facies inMS4 andMS5, 8 km SW of Xigaze town. a, shale intercalatedwith thin beds of fine-grained sandstone and siltstone;b, poorly sorted turbidite or rhythmic alternation of sandstone, siltstone, and shale. The circled Tibetan prayer flag on the top of the mountain is about 3 m high.

    426 C. Wang et al. / Gondwana Research 22 (2012) 415433

    In the Bailang submassif, 40Ar/39Ar dating of hornblendes fromthree amphiobolite samples yielded ages of 123.62.9 Ma, 127.72.2 Ma and 127.42.3 Ma (Guilmette et al., 2008, 2009), in goodagreement with the zircon SHRIMP UPb age of the gabbro (125.60.8 Ma) (Li et al., 2009). Guilmette et al. (2008, 2009) interpretedthese ages as indicating the initiation of intra-oceanic subduction. Thisconclusion is supported by geochemical studies of the ultramafic andmafic rocks that indicate the presence of a north direction Early Creta-ceous intra-oceanic subduction system within the Neo-Tethyan Ocean(Huot et al., 2002).

    In the Geding submassif, a gabbro yielded zircon UPb age of 1282 Ma (Wang et al., 2006). This result was interpreted as the age of sea-floor spreading in the Neo-Tethys. However, the mineral chemical datain contrast indicates that the Geding ophiolite was generated in an SSZsetting (Hbert et al., 2003). Geochemical studies of the Geding basalticrocks suggest that they represent a fragment of mature mid-ocean-ridge basalt (MORB) lithosphere that has been modified in the SSZ en-vironment (Chen and Xia, 2008).

    In the Ngamring submassif, the elemental and SrNbPb isotopicgeochemical data of the mafic rocks show an affinity of N-MORBwithout NbTa anomalies, indicating a mid-ocean-ridge or matureback-arc tectonic setting (Niu et al., 2006).

    Fig. 12. Map of paleocurrent flow directions in the Ngamring Formation flysch. 1. Flute cast;non-polar); 6. numbered data points: 148 for data based on the present study, and 4994

    Zircon SHRIMP UPb dating of diabase in the Sangsang submassifshows that it was formed around 125.23.4 Ma (Xia et al., 2008).Geochemical studies of mafic rocks from the Saga and Sangsang sub-massifs indicate that they are similar to back-arc basin basalts, withsmall negative NbTa anomalies (Bdard et al., 2009). In combinationwith the composition of the peridotites, Bdard et al. (2009) pro-posed that the mafics were formed in an intra-oceanic SSZ.

    In summary, the above data show that the intra-oceanic subductionzone was located within the central segment of YZSZ during the lateEarly Cretaceous, which was temporally formed as the basement ofthe XFB. However, the western and eastern segments of YZSZ aremore complex. In the eastern segment, the ages of rocks from the Luo-busa and Zedang massifs range from 177 to 152 Ma (McDermid et al.,2002; Zhou et al., 2002; Zhong et al., 2006), making them older thanthe Xigaze ophiolites. This may explain why there was not a similarforearc basin developed over there. In the western segment, the agesof the ophiolitic rocks from Zhongba (Dai et al., 2011b), Xiugugabu(Wei et al., 2006), and Yungbwa (Li et al., 2008a,b,c; Bezard et al.,2011), show that most of them were formed around 120130 Ma. Theoccurrence of the Devonian mafic rocks (Zhou, 2002; Dai et al., 2011a)that were inferred as the remnant of Paleo-Tethys indicate a muchmore complex geologic evolution of the ophiolites in YZSZ.

    2. groove cast; 3. current ripple; 4. over-banking direction; 5. pebble orientation (uni-/for data from Drr (1993); and 7. village and town.

    image of Fig.12image of Fig.11

  • Within-plate lavas

    Volcanic arc lavas


    Volcanic pebbles Sandstones

    Gangdese Volcanic rocks

    10 100 1000






    Fig. 13. TiZr covariation diagram for the volcanic pebbles and sandstones from theXigaze Group superimposed on fields of Mid-Ocean Ridge Basalt (MORB), Volcanic-arc lavas and Within-plate lavas (Pearce et al., 1981). All the Xigaze volcanic pebblesand sandstones show compositions of volcanic arc rocks, resembling those of Jurassicto Cretaceous Gangdese volcanic rocks. Data sources of andesite volcanic pebbles andsandstones are from Drr (1996), Gangdese volcanic rocks (Sangri Group and YebaFormation) from Dong et al. (2006), Kang et al. (2009), andZhu et al. (2008a,b,2009a,b).

    427C. Wang et al. / Gondwana Research 22 (2012) 415433

    8. Provenances of XFB sediments

    8.1. Paleocurrent pattern

    Paleocurrent analysis provides important information about prov-enance, geometry of the depositional units, and depositional process-es. During field work, paleocurrent directions were obtained at 48localities (Table S1). Another set of 46 paleocurrent data was pub-lished by Drr (1993). Both of these sets of data demonstrate a pre-dominantly southward paleocurrent flow and subordinatelywestward and eastward directed flows (Fig. 12). This has previouslybeen interpreted as indicating that the XFB was chiefly supplied bysediments from the GMA located to the north, and partly from theophiolite belt to the south (e.g. Drr, 1993, 1996). Our data supportsthis conclusion. For example, much of the ophiolitic detritus-basicvolcanic rocks, chert, and calcite- and epidote-replaced pyroxenesfrom the MS11 in the south were founded in the sandstones in theNaxia profile where the paleocurrent flow directions are NNE 10.However, another group of paleocurrents in MS1 indicated an east-ward flow direction (Fig. 12), and that at both the southern andnorthern flanks. This suggests a change in flow direction after currentreached a southerly point, where it might have been obstructed by ei-ther morphological elevation on the seafloor, or by the presence of anaccretionary wedge that constrained the current to flow in the axialdirection of the basin. The observed eastward paleoflow supportsbasin asymmetry and a WE axial direction of the XFB.

    Paleocurrent flow direction was unified during MS2, when the flowwas southward, withminor SW flows, indicated by flute cast and gravelimbrications in the Kardoi and Naxia area (Table S1, Fig. 12). This can beattributed to the complete infilling of the basin, causing sediment to by-pass it into the trench, and therefore indirectly implying that the basinsubsidence center had migrated farther to the south.

    After MS2 deposition, both eastward and westward paleocurrentflow are found in MS3MS5, as well as the southward and northwardflow. The highly variable paleocurrent flow directions can beexplained by source supplies from the south and the north super-posed on westerly and easterly sources, in a similar fashion to theGreat Valley forearc basin, in western USA (e.g. Ingersoll, 1979;Moxon, 1988, 1990; Williams, 1997).

    8.2. Sources of andesitic component

    Petrological texture, mineralogical composition and geochemistry(Fig. 13) have shown that volcanic detritus in the Ngamring Forma-tion was mainly derived from andesites (Drr, 1996). The shortageof rhyolitic and dacitic detritus indicates that the source of the detri-tus could have been remelting of originally basaltic mantle (e.g. lowercrust or subducted oceanic lithosphere), but not from remelting of themiddle to upper crust (Gill, 1981), which is dominated by SiO2-richmagma, as seen in northern Chile, northern Peru, and western Amer-ica (Schmincke, 1986). Previous studies indicate that a thinned ortransitional continental crust could have existed in the southernGMA during the Early Jurassic through the Early Cretaceous (e.g.Dong et al., 2006; Geng et al., 2006; Zhu et al., 2008a,b, 2009a,b,2011), where andesitic volcanic rocks could have been formed (e.g.the Sangri Group andesites, Zhu et al., 2006, 2009a) and thereaftereroded.

    What could be the sources of the andesitic components in theNgamring Formation? A probable derivation is that they weresourced from older volcanic strata in the GMA, as represented bythe Yeba Formation and Sangri Formation. Macrofossils (Yin et al.,1998) and isotopic chronology indicate that the Yeba Formation isearly to middle Jurassic (e.g. Geng et al., 2006; Zhu et al., 2008a,b,2009a,b) and that the Sangri Formation is late Jurassic to early Creta-ceous (Dong et al., 2006; Zhu et al., 2009a). These ages match the twoprincipal magmatic periods obtained from detrital zircon studies of

    the Ngamring Formation (Wu et al., 2010). The Yeba Formation ischaracterized by the rare occurrence of intermediate rocks (Zhu etal., 2008b) while the Sangri Group is dominated by the presence ofandesites and andesitic breccias with dacites (Dong et al., 2006;Kang et al., 2009; Zhu et al., 2009a). This means that the sedimenta-ryvolcanic sequence of the Late JurassicEarly Cretaceous SangriGroup in the southern GMA was the primary source of andesitic ma-terial from the Ngamring Formation flysch. This conclusion is consis-tent with the recent recognition of Wu et al. (2010), who identified aprincipal magmatic period between 130 Ma and 80 Ma (peaking ataround 110 Ma) and a subordinate period between 190 Ma and150 Ma (peaking at 160 Ma) from detrital zircons from sedimentaryrocks in the XFB.

    The explanation could be the presence of older volcanic stratasouth of the GMA, as represented by the Yeba Formation and SangriFormation. Macrofossils (Yin et al., 1998) and isotopic chronology in-dicate that the Yeba Formation is early to middle Jurassic (e.g. Geng etal., 2006; Zhu et al., 2008a,b, 2009a,b) and that the Sangri Formationis late Jurassic to early Cretaceous (Dong et al., 2006). These agesmatch the two principal magmatic periods obtained from detrital zir-con studies of the Ngamring Formation (Wu et al., 2010). The SangriGroup is dominated by andesitic lava and andesitic breccia withdacites, indicating that their origin is related to arc-type volcanicrocks formed in the lower crust during the Neo-Tethys oceanic crustsubduction (Dong et al., 2006; Kang et al., 2009). The above ages ofthe Yeba and Sangri formations strongly suggest that the two Meso-zoic sedimentaryvolcanic sequences in the southern GMA were thesource of andesitic material for the Ngamring flysch.

    The composition and geochemistry of the sediments also showthat the andesitic and other volcanic lithiclastics are common in thelowest sequence of the Ngamring Formation, but their occurrence de-creases upwards. A possible explanation is that as increasing amountsof terrigenous materials were transported into the XFB, older rockswere eroded as the GMAwas denuded. As a result, the younger volca-nic rocks were the source for the lower part of the sequence, whilethe upper part of the XFB strata was supplied from progressivelyolder rocks during unroofing of GMA. This reversal is supported by re-cent results of detrital zircon UPb and Hf isotopic analyses. Most zir-con ages from the lower flysch of the Ngamring Formation are

    image of Fig.13

  • 428 C. Wang et al. / Gondwana Research 22 (2012) 415433

    characterized by positive Hf(t) values similar to those of magmaticzircons from the Gangdese batholith and from the upper Ngamringand PadanaQubeiya formations. The younger sequences record ei-ther pre-Mesozoic UPb zircon ages or negative Hf(t) values (Wuet al., 2010).

    9. Discussions

    9.1. Sedimentary evolution of XFB

    The forearc basins develop as part of the trencharc system abovethe subduction zone. Dickinson and Seely (1979) classified the fore-arc basins into five types: intramassif basin, residual basin, accretion-ary basin, constructed basin, and composite basin. Combination of therevised stratigraphic framework and of the basin basement, lithofa-cies architecture (megasequences), and provenance study haveresulted in our recognition of four distinct stages in the XFB develop-ment: basement formation, residual forearc, composite forearc, andterminal forearc. The basement formation period is represented bythe Sangzugang Formation which was deposited on a narrow shelfat an active continental margin. Residual and composite forearcbasin periods are represented by the Ngamring Formation, compris-ing five successive megasequences of the submarine fan. The terminalforearc period is represented by the Cuojiangding Group. The changeof the forearc basin type could indicate a linkage to the termination ofthe subduction of the New-Tethys oceanic lithosphere.

    As discussed in Section 8.2, andesitic clastics are common withinsediments of the XFB, indicating that partial melting of lower crustor subducted oceanic lithosphere should have been processed in itsnorth before the XFB formation (SU1, Fig. 14A). Such melting eventcould be attributed to the subduction of the Neo-Tethyan oceanic lith-osphere that resulted in extensive volcanism occurred in the southernmargin of the Lhasa terrane (Zhu et al., 2006, 2009a,b, 2011), asrepresented by the Sangri Group that could have provided andesiticmaterials for the XFB. By the standing subduction, an intra-oceanicsystem (including the northward-subducted zone, oceanic island-arc and intra-arc basin) was formed in south of the open Neo-Tethys ocean (SU2, Fig. 14A, for details refer to Hbert et al., 2011),during which the SSZ-type Xigaze ophiolites were generated. Accord-ing to the ages of the Xigaze ophiolite (see in the Section 7) and XFB(see in Section 5.1), it is reasonable to conclude that the subductionmay have terminated about 120 Ma ago.

    In comparison to the older normal oceanic crust, the SSZ-typeXigaze ophiolite has a light density, for which it would have neverbeen intermixed with the mantle lithosphere. Therefore, it wasobducted toward the south and became the basement of the XFB inthe late Early Cretaceous (SU2, Fig. 14B and C). At the same time,the Sangzugang Formation carbonates were deposited on a narrowcontinental shelf. During the same time, the residual forearc basinwas formed by the initiation of the typical arctrench system of theYZSZ.

    The residual forearc basin represents the earliest stage of the XFBdevelopment, represented by the late Albianmid CenomanianNgamring Formation flysch (MS1) and by the Chongdoi Formation.During this period, submarine fan and slope facies spread out intothe basin. As described in Section 6.1, the evolution of the megase-quence MS1 can be subdivided into two stages (Fig. 7): in the firststage, MS11 is characterized by slope facies mudrock and siltstone in-tercalated with debris-flow breccia and gravelly mudrock, dominatedby non-channelized debris-flow sediment in apron facies in the northand by rare lenticular and sheet turbidite sandbodies of facies B. Dur-ing the second stage, MS12 was formed by overbank sediments, and itfeatures non-channelized fine-grained sediment facies, during whichsupply channels were not incised for transportation of detritus fromeither the GMA or the carbonate platform on the narrow margin, in-dicating no point sources, despite the fact that the XFB contains

    sediments from both the carbonate platform and GMA in the northand from the Xigaze ophiolite and the accretionary wedge in thesouth.

    Frequent debris of orbitolinid-bearing limestone found in the ba-sinal Chongdoi Formation indicates that erosional clastics from thenorthern carbonate platform reached as far as the southern XFB.Such interpretation is supported by positive initial Hf(t) values of zir-cons derived from the GMA (Wu et al., 2010). The eastwest paleo-current direction (Fig. 12) indicates that the accretionary wedgeformed a morphologic high on the sea floor and that the subsidencecenter was located south of the XFB.

    During the middle-late Cenomanian, the XFB changed from a re-sidual basin to a composite basin, and the accretionary wedge also be-came a part of the basement of the basin. At this time, largesubmarine fan channelized systems developed in the deep-sea basin(Figs. 7, 8, 9), while the carbonate platform was exposed to subaerialconditions and eroded when sea level dropped sharply. Subsequently,large amounts of sediment were brought up to the accretionarywedge, and in places channels approached the trench. This is evidentby the occurrence of only two sets of channel fill sequences depositedin the distal basin setting where they directly downlapped onto pe-lagic limestone in the Kardoi region. During the stage, the directionof sediment supply to the basin changed from lateral to longitudinal,attributable to the tectonic barrier caused by emplacement of the ac-cretionary wedge and trapping sediments in the XFB. This tectonic ac-tivity resulting from ocean crust subduction extended farther to thenorth, as evidenced by folding of the Takena Formation in the Lhasaretroarc basin (Coulon et al., 1986).

    The composite forearc basin matured during the early-middle Tur-onian (Fig. 7) when MS3 was built by middle and outer fan facies.During this time period sediment supply decreased. On the north ofthe basin, the megasequence is formed by branched channel sand-stone of facies B and lobe sandstone of the lower part of facies C, grad-ing upward into fine-grained turbiditic and lobe sandstones; the topstrata contain thin pelagic limestone layers, indicating a pelagic envi-ronmental setting. In the south, MS3 is dominated by facies C lobesandstone and facies D outer fan-basin fine-grained sediments.

    During the middle-late Turonian, the composite forearc basin wasconstructed by MS4 (Fig. 7) consisting of branching channel lobe sand-stone and abundant basin facies. At the top of this megasequence, light-gray pelagic limestone provides significant late-Turonian marker forcorrelation between both flanks of the synclinorium. The compositeforearc basin development stage weakened during the Coniacian. It isrepresented by MS5, which is comprised by the outer-fan and basinfacies (Fig. 7).

    As revised in Section 5, the change from the deep-sea facies toshallow-sea facies lies between the Ngamring Formation of theXigaze Group and the Padana Formation of the Cuojiangding Group.That means either the age at the base of the Padana or the age atthe top of the Ngamring represent the termination of the deep-seadeposition. This time would mark either the termination of the com-posite forearc basin or the initiation of the terminal forearc basin. Therevision of stratigraphy (details refer to Section 5) shows that theNgamring Formation is of late AlbianConiacian age and the PadanaFormation is of Santonian age, indicating that deposition in the XFBchanged from deep-sea facies to shallow-sea facies at ca. 86 Ma.

    The terminal sedimentary succession of the XFB is represented bythe Cuojiangding Group comprised by Late CretaceousEarly Paleo-gene shallow marine carbonates and terrigenous rocks that crop outwest of Ngamring. This shallowing-up sequence could be attributedto the decreasing accommodation space for basin filling material(Fig. 14C). The hypothesis is supported by the conformable contactbetween the Xigaze Group and the Cuojiangding Group (Fig. 6).

    Relatively, it is simple to date the termination of the XFB. The Gya-laze Formation at the top of the Cuojiangding Group is generallythought to be the marine sediment record of the XFB. We then

  • Fig. 14. Model showing the evolution of the XFB and the middle YSZS. A. First (early) subduction (SU1) in the earliest Barremian (~130 Ma), indicating possible bi-directional sub-duction of the Neo-Tethys oceanic crust and an intra-oceanic obduction (OB) before the formation of XFB, resulting in spreading of southern back-arc and the forming of theobducted Xigaze ophiolites. Note the potential overriding thrust in northern back-arc and ophiolitic mlange in forearc basin. B. Second subduction (SU2) and XFB formation duringthe early Albian. New typical trench, arc, and forearc basin were formed by persistent subduction, during which the Xigaze ophiolite was obducted onto the forearc area of theGangdese arc and became the basement of the XFB. Note three contemporaneous geological processions: the retro-arc basin had already started in the northern Gangdese arc(Leier et al., 2007), the shallow carbonate platform (Sangzugang Formation) had been built up on the narrow shelf in the southern Gangdese arc, and the prism had been addedto the southern edge of the XFB blocking the southward transportation of clastics into the trench. C. Complete filling of the XFB in the Late CretaceousEarly Paleogene(8650 Ma) by shallow sea Cuojiangding Group sediments. The accretionary wedge may be itself raised above sea level after the transition of the deep-sea flysch to shallow-seafacies, being a potential source area of the XFB, although not the main source. D. Present structure of the YZSZ, showing the locations of the YZSZ and the accretionary wedgeafter the collision (5045 Ma) of India with Eurasia. Note width of accretionary wedge, the inferred localities of MORB ophiolites, and the change in source of volcanic rocks toboth volcanic and batholithic rocks.

    429C. Wang et al. / Gondwana Research 22 (2012) 415433

    suggest that the termination of the Gyalaze Formation depositionrepresents the termination of XFB. As revised in Section 5.2, theGyalaze Formation is of SelandianYpresian age (top age, ca.5049 Ma), so we propose that the XFB terminated at about 5049 Ma.

    As a whole, the sedimentary succession of the XFB represents atypical composite basin (Dickinson, 1995). The XFB developed inthe forearc area with sediments extensively supplied from the conti-nental magmatic arc. The sedimentary succession of pelagic radiolar-ian chert, shale and microfossil-bearing tuff in the lowest layer, flyschsubmarine fan facies in the middle, and costal-shallow facies in theupper strata documents the evolution of the basin subsidence center.

    9.2. Comparison between Indus Forearc Basin and XFB

    In LadakhIndus, the suture is named the Indus Suture Zone. Con-sidered together with the YZSZ, it was named the IndusYarlungZangbo Suture Zone (Searle et al., 1997; Henderson et al., accepted

    for publication), and is regarded as the site of the collision betweenthe Indian and Eurasian continents (e.g. Gansser, 1977, 1980;Shackleton, 1981; Tapponnier et al., 1981). Similar to the XFB, theIndus Forearc Basin developed in the Indus Suture Zone. Starting onthe Ladakh magmatic arc in the mid-Cretaceous, development of theIndus Forearc Basin terminated in the early Eocene, as marked by anintermontane molasse (Van Haver, 1984; Garzanti and Van Haver,1988). It is obviously instructive to compare the two.

    As is evident from their present structure, both are asymmetrical,with the one exception of near-symmetry appearing in the easternXFB. This could have resulted from thrusting in collision and/orfrom different thicknesses in the original deposition.

    A complete terrigenous to carbonate succession 1500 m thick wasdeposited in the southern Indus Forearc Basin, similar to the XFB. Fa-cies change through time show similar sequences of shallowing-upwards megasequences (turbidite to nummulitic limestone), and al-luvial to deltaic sediments occur in both forearc basins. Platform

  • 430 C. Wang et al. / Gondwana Research 22 (2012) 415433

    deposition began with orbitolinid and rudist carbonates at almost thesame time, shown by the Khalsi Formation in Indus and the Sangzu-gang Formation in Yarlung Zangbo.

    Further, the terrestrial coarse successions above the forearc basinthat started to develop following the early Eocene show a great dealof similarity in both regions. A set of conglomerates and sandstones2501600 m thick, nonconformable on the marine Gyalaze Formationlimestones and sandstones of the top XFB in Cuojiangding Lake,northern Zhongba, correlate with those of the Hemis conglomeratein the Indus basin. It is interesting that the terrestrial successionwas transported westward in the Indus basin, mostly fed from thenortheast (Frank et al., 1977; Baud et al., 1982) and became more dis-tal toward the west and south. It is inferred that this change repre-sents the initiation of collision between two continents and impliesan oblique collision from east to west; however, much work needsto be done in the future to test this hypothesis.

    There are two distinctions between the two basins. One is that theposition of shallow water carbonate deposition is different: the KhalsiFormation is located between the Indian Ophiolite Complex and Indi-an Forearc Basin, whereas the Sangzugang Formation is positioned onthe northern margin of the XFB. The other difference is the timing ofthe beginning of the deep-sea flysch; the deposition of the NgamringFormation started earlier than the late Albian in XFB, and the Tar For-mation (deep-sea turbidite) began in the Maastrichtian or even in thePaleocene. This distinction is profound, and implies either diachrone-ity or obliquity of the Neo-Tethyan oceanic crust subduction. This isalso evident in the Indus Basin. Shallow-sea nummulitic limestonewas deposited in the west followed by deltaic succession, withdeep-sea turbidite occurring in the south-east (Garzanti and VanHaver, 1988).

    10. Summaries and conclusions

    The XFB resembles a typical forearc basin in its filling pattern andbasement lithology (Dickinson, 1995; Dickinson and Seely, 1979). Thebasin has a shallowing-upwards sequence of pelagic and turbiditicdeep-sea facies, from near-shore shelf through to continental facies,dated from the late Aptian through the Ypresian (ca. 12049 Ma)with a duration of about 71 Myr. The basin is narrow (2022 kmwide in the northsouth direction) and 550 km long in the eastwest direction. The sedimentary fill is asymmetric. The NS shorten-ing ratio has been estimated as 31% for the Gyabulin Formationand 65% for the Ngamring Formation (Ratschbacher et al., 1992),which makes the dimensions of the original XFB were 71 km wideand 50100 km long, thus comparable in size to a large forearcbasin (Dickinson, 1995; Dickinson and Seely, 1979).

    The combination of the revised stratigraphic framework, reinter-pretation of lithofacies architecture, provenance, and basin basement,leads to the following conclusions:

    1) The XFB consists of two main lithostratigraphic associations: theXigaze Group and the Cuojiangding Group. The former group con-sists of the Sangzugang Formation, Chongdoi Formation, andNgamring Formation, with their ages revised as the late Aptianearly Albian, middle AptianAlbian, and middle AlbianConiacian,respectively. The thickness of the Ngamring Formation was revisedto less than 4000 m on the basis of the structural analyses of studiedprofiles. The Cuojiangding Group is composed of the Padana Forma-tion, Qubeiya Formation, Quxia Formation, and Gyalaze Formation,ages ofwhichwere revised as Santonian, CampanianMaastritchian,Danian, and SelandianYpresian, respectively. Most of the forma-tions have conformable contactswith each other except of the Sang-zugang Formation, which is disconformable with others.

    2) The Xigaze Group is dominated by deep-sea fan flysch lithofacies,except for the Sangzugang Formation, which is comprised by shal-low carbonate platform lithofacies. The Cuojiangding Group is

    characterized by shallow-marine mixed terrigenous and carbon-ate rocks. The gradual upward transition from the deep-sea flyschfacies of the Ngamring Formation into the shallowmarine facies ofthe Padana Formation indicates progressive basin filling ratherthan tectonic activity, which is tentatively dated at the boundary(ca. 86 Ma) of the Santonian and Coniacian .

    3) The deep-sea fan represented by the Ngamring Formation consistsof sandy turbidites, channeled sandstones and conglomerateswith a few thin pelagic limestone and calcareous shale. Sedimen-tary beds are arranged into five megasequences (MS1MS5), com-prised by the early submarine fan, channeled submarine fan,mature submarine fan, highstand submarine fan, and outer fanand pelagic sediments. MS1 has been further subdivided into theslope apron facies sub-megasequence (MS11) and the overbankfacies sub-megasequence (MS12). Megasequence MS2 can be sub-divided into the channel system of the Xigaze submarine fan andthe channel system of the Kardoi submarine fan.

    4) Former data plus ours show that the paleocurrent flows were east-wards and westwards during the early and late stages of basin de-velopment and northwards and southwards during its middlestage, indicating typical forearc basin sedimentation with longitu-dinal supply and lateral filling. This change in directional sedimentsupply to the basin can be attributed to the tectonic barrier causedby emplacement of the accretionary wedge and trapping sedi-ments in the XFB.

    5) The provenance analysis supports a new interpretation that theandesitic detritus of the XFB may have been derived from theJurassicEarly Cretaceous arc-volcanic Yeba Formation and SangriGroup south of the GMA, but not from the batholith or/and Paleo-gene volcanic rocks in GMA, implying that obduction of the SSZXigaze ophiolite may have been directed toward to the south.

    6) Four distinct phases of the XFB development we recognize are thebasement formation, residual forearc, composite forearc, and ter-minal forearc. The main basement of the XFB is formed by theXigaze ophiolite, formed prior the Aptian (ca. 120 Ma), with thecoeval counterpart on the continent being the carbonate Sangzu-gang Formation. The residual forearc may have initiated in theearliest Aptian (ca. 125/120 Ma), as represented by sedimentarydeposits of the Chongdoi Formation. Then the residual forearcand composite forearc developed, represented by the late AlbianConiacian flysch megasequences of the Ngamring Formation. Theterminal forearc continued through the late Late CretaceousEarlyPaleogene (ca. 8649 Ma), represented by the shallow marineCuojiangding Group, and would have not terminated until thelatest Ypresian (ca. 5049 Ma) by the Gyalaze Formation.

    7) The studies of stratigraphy, lithofacies and provenance as well ascomparison to other typical forearc basins not only can contributeto the understanding of the sedimentary evolution of the XFB, butalso are significant to the evolution of YZSZ. From the records ofXFB, we suggest that the Neo-Tethys oceanic crust may havebeen subducted at two different times and obducted once.

    Supplementary material related to this article can be found onlineat doi:10.1016/


    We acknowledge the help of many colleagues in carrying out thefield work: Zhou Xiang, Lu Yan, Wan Xiaoqiao, Chen Jianping, HeZhengwei, M. Dobs, Li Anren, and I. Westlake. G. Einsele and S. Drrprovided some materials during this study. Comments from R. Hbertwere useful in shaping the manuscript. We also thank Zhu Dichengfor constructive suggestions and comments. This project is fundedby National Natural Science Foundation of China (41072075), Nation-al Basic Research Program of China (2012CB822000), 111 Project of

  • 431C. Wang et al. / Gondwana Research 22 (2012) 415433

    China Grant (B07011), and Zhongba 1:50,000 geological mappingprojects by Chinese Bureau of Geological Survey (12112011086037).


    Aitchison, J.C., Davis, A.M., Liu, J., Luo, H., Malpas, J.G., McDermid, I.R.C., Wu, H.Y., Ziabrev,S.V., Zhou, M.F., 2000. Remnants of a Cretaceous intra-oceanic subduction systemwithin the YarlungZangbo suture (southern Tibet). Earth and Planetary ScienceLetters 183, 231244.

    Allaby, A., Allaby, M., 1999. Megasequence A Dictionary of Earth

    Allgre, C.J., Courtillot, V., Tapponnier, P., Hirn, A., Mattauer, M., Coulon, C., Jaeger, J.J.,Achache, J., Schrer, U., Marcoux, J., Burg, J.P., Girardeau, J., Armijo, R., Garipy, C.,Gpel, C., Li, T.D., Xiao, X.C., Chang, C.F., Li, G.Q., Lin, B.Y., Teng, J.W., Wang, N.W.,Chen, G.M., Han, T.L., Wang, X.B., Den, W.M., Sheng, H.B., Cao, Y.G., Zhou, J., Qiu,H.R., Bao, P.S., Wang, S.C., Wang, B.X., Zhou, Y.X., Xu, R.H., 1984. Structure and evo-lution of the HimalayaTibet orogenic belt. Nature 307, 1722.

    Bassoullet, J.P., Colchen, M., Mascle, G., Wang, N., 1984. Les ensembles sedimentaires dela zone du Tsangpo (Lhaze, Lhasa, Linzhu). In: Mercier, J.L., Li, Guangcen (Eds.),Mission Franco-Chinoise au Tibet, 1980. Centre National de ia Recherche Scientifi-que, Paris, pp. 133153.

    Baud, A., Arn, R., Bugnon, P., Crisnel, A., Dolivo, E., Escher, A., Hammerschlag, J.G.,Marthaler, M., Masson, H., Steck, A., Tieche, J.C., 1982. Le contact Gondwanaperi-Gondwana dans Ie Zanskar oriental (Ladakh, Himalaya). Bulletin Socit Go-logique de France 24 (2), 341361.

    Bdard, ., Hbert, R., Guilmette, C., Lesage, G., Wang, C.S., Dostal, J., 2009. Petrologyand geochemistry of the Saga and Sangsang ophiolitic massifs, Yarlung ZangboSuture Zone, Southern Tibet: evidence for an arcback-arc origin. Lithos 113,4867.

    Bezard, R., Hbert, R., Wang, C., Dostal, J., Dai, J., Zhong, H., 2011. Petrology and geo-chemistry of the Xiugugabu ophiolitic massif, western Yarlung Zangbo suturezone, Tibet. Lithos 125, 347367.

    Cao, R.L., 1981. Petrological characteristics of sediments within the trench and theophiolitic belts of the Yarlung Zangbo suture zone and their geological signifi-cances. Geosciences 3, 247254 (in Chinese with English abstract).

    Caron, M., 1985. Cretaceous planktic foraminifera. In: Bolli, H.M., Saumders, J.B., Perch-Nielsen, K. (Eds.), Plankton Stratigraphy. Cambridge University Press, pp. 1786.

    Chen, G.W., Xia, B., 2008. Platinum Group elemental geochemistry of mafic and ultra-mafic rocks from the Xigaze ophiolite, southern Tibet. Journal of Asian Earth Sci-ences 32, 406422.

    Cherchi, A., Schroeder, R., 1980. Parorbitolinoides hedini n. gen. n. sp., grand foramini-fere du Crtac infrieur du Tibet mridional. Compte Rendus de l'Academie desSciences, Paris, Serie D291, 385388.

    Chung, S.L., Chu,M.F., Ji, J.Q., O'Reilly, S.Y., Pearson,N.J., Liu, D.Y., Lee, T.Y., Lo, C.H., 2009. Thenature and timing of crustal thickening in Southern Tibet: geochemical and zircon Hfisotopic constraints from postcollisional adakites. Tectonophysics 477, 3648.

    Coulon, C., Maluski, H., Bollinger, C., Wang, S., 1986. Mesozoic and Cenozoic rocks fromcentral and southern Tibet: 39Ar40Ar dating, petrologic characteristics and geody-namic significance. Earth and Planetary Science Letters 79, 281302.

    Dai, J., Wang, C., Hbert, R., Li, Y., Zhong, H., Guillaume, R., Bezard, R., Wei, Y., 2011a.Late Devonian OIB alkaline gabbro in the Yarlung Zangbo Suture Zone: remnantsof the Paleo-Tethys? Gondwana Research 19, 232243.

    Dai, J.-G., Wang, C.-S., Hbert, R., Santosh, M., Li, Y.-L., Xu, J.-Y., 2011b. Petrology andgeochemistry of peridotites in the Zhongba ophiolite. Yarlung Zangbo SutureZone: implications for the Early Cretaceous intra-oceanic subduction zone withinthe Neo-Tethys. Chemical Geology 288, 133148.

    Dewey, J.F., Shackleton, R.M., Chang, C.F., Sun, Y.Y., 1990. Tectonic evolution ofQinghaiTibet Plateau. Sino-Britain Investigation Team, Geological Tectonic Evolu-tion of QinghaiTibet Plateau. Science Press, Beijing, pp. 384415 (in Chinese withEnglish abstract).

    Dickinson, W.R., 1995. Forearc basins. In: Busby, C.J., Ingersoll, R.V. (Eds.), Tectonics ofSedimentary Basins. Cambridge, Massachusetts, Blacwell Science, pp. 211261.

    Dickinson, W.R., Seely, D.R., 1979. Structure and stratigraphy of forearc regions. AmericanAssociation of Petroleum Geologists Bulletin 63, 231.

    Dong, Y.H., Xu, J.F., Zeng, Q.G., Wang, Q., Mao, G.Z., Li, J., 2006. Is there a Neo-Tethys'subduction record earlier than arc volcanic rocks in the Sangri GrGroup? ActaPetrologica Sinica 22 (3), 661668 (in Chinese with English abstract).

    Dubois-Ct, V., Hbert, R., Dupuis, C., Wang, C.S., Li, Y.L., Dostal, J., 2005. Petrologicaland geochemical evidence for the origin of the Yarlung Zangbo ophiolites, south-ern Tibet. Chemical Geology 214, 265286.

    Dupuis, C., Hebert, R., Dubois-Cote, V., Wang, C.S., Li, Y.L., Li, Z.J., 2005. Petrology andgeochemistry of mafic rocks frommelange and flysch units adjacent to the YarlungZangbo Suture Zone, southern Tibet. Chemical Geology 214, 287308.

    Drr, S., 1993. Sedimentation and sediment composition of the Cretaceous Xigaze fore-arc basin, South Tibet. Tiibinger Geowiss. Arb., Reihe A, Geol. Paleiontol., InstitutTiibingen, pp. 1119.

    Drr, S., 1996. Provenance of Xigaze fore-arc basin clastic rocks (Cretaceous, SouthTibet). Geological Society of America Bulletin 108, 669684.

    Einsele, G., Liu, B., Drr, S., Frisch, W., Liu, G., Luterbacher, H.P., Ratschbacher, L., Ricken,W., Wendt, J., Wetzel, A., Yu, G., Zheng, H., 1994. The Xigaze forearc basin; evolu-tion and facies architecture (Cretaceous, Tibet). Sedimentary Geology 90, 132.

    Frank, W., Gansser, A., Trommsdorff, V., 1977. Geological observations in the Ladakharea (Himalayas)a preliminary report. Schweizerische Mineralogische undPetrographische Mitteilungen 57, 89133.

    Gansser, A., 1977. The great suture zone between Himalaya and Tibeta preliminary ac-count. Colloq. Intern., 268. Ecologie et Gologie de l'Himalaya C.N.R.S, pp. 181191.

    Gansser, A., 1980. The Peri-Indian suture zone. Texte integral du collogue C5: G-ologiedes chaines alpines issues de la Tthys du 2b GI, pp. 140148.

    Gao, Y.L., 1988. Evedence for plate tectonics and evolution of the middle segment ofthe Yarlung Zangbo suture zone, Tibet. Regional Geology of China 4, 341347 (inChinese with English abstract).

    Gao, Y.L., Tang, Y.Q., 1984. Melanges in the southern Xizang (Tibet). Himalaya GeologyEditing Commission ed. Himalaya Geology (II). Geological Publishing House, Beijing,pp. 2744 (in Chinese with English abstract).

    Garzanti, E., Van Haver, T., 1988. The Indus clastics: forearc basin sedimentation in theLadakh Himalaya (India). Sedimentary Geology 59, 237249.

    Geng, Q.R., Pan, G.T., Wang, L.Q., Zhu, D.C., Liao, Z.L., 2006. Isotopic geochronology ofthe volcanic rocks from the Yeba Formation in the Gangdise zone, Xizang. Sedi-mentary Geology and Tethyan Geology 26 (1), 17 (in Chinese with Englishabstract).

    Gill, J., 1981. Orogenic Andesites and Plate Tectonics. Springer, Berlin, pp. 1390.Girardeau, J., Mercier, J.C.C., 1988. Petrology and texture of the ultramafic rocks of the

    Xigaze ophiolite (Tibet): constraints for mantle structure beneath slow-spreadingridges. Tectonophysics 147, 3358.

    Girardeau, J., Mercier, J.C.C., Cao, Y.G., 1985a. Origin of the Xigaze ophiolite, YarlungZangbo suture zone, southern Tibet. Tectonophysics 119, 407433.

    Girardeau, J., Mercier, J.C.C., Wang, X.B., 1985b. Petrology of the mafic rocks of theXigaze ophiolite, Tibet: implications for the genesis of the oceanic lithosphere.Contributions to Mineralogy and Petrology 90, 309321.

    Gpel, C., Allgre, C.J., Xu, Ronghua, 1984. Lead isotopic study of the Xigaze ophiolite(Tibet): the problem of the relationship between magmatites (gabbros, dolerites,lavas) and tectonites (harzburgites). Earth and Planetary Science Letters 69,301310.

    Gradstein, F.M., Ogg, J.G., van Kranendonk, M., 2008. On the geologic time scale 2008.Newsletters on Stratigraphy 43 (1), 513. doi:10.1128/0078-0421/2008/0043-0005.

    Guilmette, C., Hebert, R., Dupuis, C., Wang, C.S., Li, Z.J., 2008. Metamorphic history andgeodynamic significance of high-grade metabasites from the ophiolitic melangebeneath the Yarlung Zangbo ophiolites, Xigaze area, Tibet. Journal of Asian EarthSciences 32, 423437.

    Guilmette, C., Hbert, R., Wang, C., Villeneuve, M., 2009. Geochemistry and geochronol-ogy of the metamorphic sole underlying the Xigaze Ophiolite, Yarlung Zangbo Su-ture Zone, South Tibet. Lithos 112, 149162.

    Hbert, R., Huot, F., Wang, C.S., Liu, Z.F., 2003. Yarlung Zangbo ophiolites (SouthernTibet) revisited: geodynamic implications from the mineral record. In: Dilek, Y.,Robinson, P.T. (Eds.), Ophiolites in Earth History: Geological Society, London, Spe-cial Publications, 218, pp. 165190.

    Hbert, R., Bezard, R., Guilmette, C., Dostal, J., Wang, C.S., Liu, Z.F., 2011. The IndusYarlungZangbo ophiolites fromNanga Parbat toNamche Barwa syntaxes, southern Tibet:firstsynthesis of petrology, geochemistry and geochronology with incidences on geody-namic reconstructions of Neo-Tethys. Gondwana Research, in revision.

    Henderson, A.L., Najman, Y., Parrish, R., BouDagher-Fadel, M., Barford, D., Garzanti, E.,And S., accepted for publication. The geology of the Cenozoic Indus Basin sedi-mentary rocks; palaeoenvironmental interpretation of sedimentation from thewestern Himalaya during the early phases of IndiaEurasia collision. Tectonics 29.

    Huot, F., Hebert, R., Varfalvy, V., Beaudoin, G., Wang, C.S., Liu, Z.F., Cotten, J., Dostal, J.,2002. The Beimarang melange (southern Tibet) brings additional constraints inassessing the origin, metamorphic evolution and obduction processes of the Yar-lung Zangbo ophiolite. Journal of Asian Earth Sciences 21, 307322.

    Ingersoll, R.V., 1979. Evolution of the Late Cretaceous forearc basin, northern and cen-tral California. Geological Society of America Bulletin 90, 18131826.

    Ji, W.Q., Wu, F.Y., Chung, S.L., Li, J.X., Liu, C.Z., 2009. Zircon UPb chronology and Hf iso-topic constraints on the petrogenesis of Gangdese batholiths, southern Tibet.Chemical Geology 262, 229245.

    Kang, Z.Q., Xu, J.F., Chen, J.L., Wang, B.D., 2009. Geochemistry and origin of Cretaceousadakites in Mamuxia Formation, Sangri GrGroup, South Tibet. Geochemica 38 (4),331341 (in Chinese with English abstract).

    Lee, H.-Y., Chung, S.L., Lo, C.-H., Ji, J.Q., Lee