Lithos 190–191 (2014) 240–263

Contents lists available at ScienceDirect

Lithos

journal homepage: www.elsevier.com/locate/lithos

Petrogenesis of Cretaceous adakite-like intrusions of the Gangdese Plutonic Belt, southern Tibet: Implications for mid-ocean ridge and crustal growth

Yuan-chuan Zheng a,b,⁎, Zeng-qian Hou b, Ying-li Gong c, Wei Liang a,Qing-ZhongSunb,SongZhanga, Qiang Fu a, Ke-Xian Huang a, Qiu-Yun Li a,WeiLia a School of Earth Sciences and Resources, China University of Geosciences, Beijing 100082, PR China b Institute of Geology, Chinese Academy of Geological Sciences, Beijing 100037, PR China c Laboratory of Department of Thermal Engineering, Tsinghua University, Beijing 100084, PR China article info abstract

Article history: We have conducted a whole-rock geochemical, U–Pb zircon geochronological, and in situ zircon Hf–Oisotopic Received 15 August 2013 compositional study of rocks in southern Tibet from the Langxian igneous suite (including a lamprophyre Accepted 16 December 2013 dyke, mafic enclaves, a granodiorite, and a two-mica granite) and the Nuri igneous suite (a quartz–diorite). U– Available online 24 December 2013 Pb zircon dating indicates that the timing of crystallization of the mafic enclaves and host granodiorite of the Langxian suite are ca. 105 Ma and 102 Ma, respectively, that the Langxian lamprophyre dyke and the two- Keywords: – – mica granite were emplaced at ca. 96 Ma and 80 76 Ma, respectively, and that the Nuri quartz diorite was fi U–Pb zircon ages emplaced at ca. 95 Ma. With the exception of the lamprophyre dyke and ma c enclaves in the Langxian area, Zircon Hf–Oisotopes rocks from the Langxian and Nuri igneous suites all show signs of a geochemical affinity with adakite- 18 Adakite-like rocks like rocks. The high Mg-numbers, high abundance of compatible elements, high εNd(t) (2.7 and 2.8) and δ O 87 86 Southern Tibet (8.9 and 9.2‰) values, elevated zircon εHf(t) (11.0–17.0) values, and low Sr/ Sr(i) ratios (0.7040), collectively indicate that the Nuri adakite-like quartz–diorite was derived from of the low temperature altered Neo-Tethyan , and that these dioritic subsequently interacted with as they rose upwards through the overlying mantle wedge. The observation of identical differentiation trends, similar whole-

rock Sr–Nd and zircon Hf isotopic compositions, and consistently low (Dy/Yb)N ratios among the Langxian igneous suite rocks, indicates that the adakite-like granodiorite was produced by low-pressure fractional crystal-

lization of precursor magmas now represented by the (relict) mafic enclaves. However, relatively high Al2O3 contents, low MgO, Cr and Ni contents, and low (La/Yb)N and (Dy/Yb)N values indicate that the two-mica granite was derived from partial melting of the southern Tibetan mafic lower crust in the absence of , while isotopic data suggest that at least 70% of the source region was juvenile materials. Combined with the presence of HT (high temperature) charnockitic magmatism, HT granulite facies , and large volumes of Late Cretaceous batholiths, the oceanic-slab-derived Nuri adakitic rocks indicate a substantial high heat flux in the Gangdese batholith belt during the Late Cretaceous, which may have been related to subduction of a Neo-Tethyan mid-ocean ridge system. According to this model, hot asthenosphere would rise up through the corresponding slab window, and come into direct contact with both the oceanic slab and the base of the overlying plate. This would cause melting of both the oceanic slab and the overlying plate by the addition of heat that was ultimately linked with peak magmatism and the significant growth and chemical differentiation of juvenile crust in southern Tibet during the Late Cretaceous (105–76 Ma). In addition, the petrogenesis of the Langxian adakite-like two-mica granite indicates that the southern Tibetan crust was still of normal thickness prior to the emplacement of these intrusions at ca. 76 Ma. This probably means that large parts of southern Tibet were not very highly elevated prior to the Indian–Asian collision. Crown Copyright © 2013 Published by Elsevier B.V. All rights reserved.

1. Introduction

The growth of is commonly ascribed to two distinct processes: subduction-zone magmatism and mantle plume-related ⁎ Corresponding author at: School of Earth Sciences and Resources, China University of magmatism (Rudnick, 1995; Taylor, 1977). However, geochemical Geosciences, 29# Xueyuan Road, Haidian District, Beijing 100082, PR China. Tel.: +86 15011361152. calculations indicate that most of the continental crust (more than E-mail address: [email protected] (Y. Zheng). ~80%) was generated by subduction-zone magmatism (Barth et al.,

0024-4937/$ – see front matter. Crown Copyright © 2013 Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.lithos.2013.12.013 Y. Zheng et al. / Lithos 190–191 (2014) 240–263 241

2000; Plank and Langmuir, 1998); consequently, an understanding of The Gangdese batholiths extend for more than 1500 km across magmatism in the subduction-zone is essential to unraveling processes southern Tibet (Fig. 1A), and they have generally been thought to of crustal growth and maturation (Lackey et al., 2005, 2006, 2008; represent a typical Andean-type convergent margin before the collision Nelson et al., 2013). It was widely assumed that melting of mantle of the Indian and Asian continents (Chu et al., 2006, 2011; Chung et al., peridotite, induced by the addition of water and “hyperfusible mate- 2005; Yin and Harrison, 2000; Zhang et al., 2014). Geophysical studies rials” (Na, Si, Al) derived from subducted oceanic slab and sedimentary indicate that the Tibetan crust beneath the Gangdese belt is twice as rocks, could be the most important crust-forming process in post- thick as average continental crust (~65–80 km) (Murphy et al., 1997; Archean subduction-zones (e.g., Rudnick, 1995; Tang et al., 2012a; Nelson et al., 1996; Priestley et al., 2006). Recent whole-rock geochem- Taylor, 1967, 1977). However, there is increasing evidence that ridge ical, Sr–Nd isotopic and in-situ zircon Hf–O isotopic studies indicate that subduction, accompanied by ridge–trench interactions, may also impact the southern Tibetan crust is characterized by a juvenile crust with strongly on magmatic activity and crustal growth along convergent mantle contributions up to 50–90% (Hou et al., 2012, 2013; Zheng margins (Bourdon and Eissen, 2003; Delong et al., 1979; Guivel et al., et al., 2012a; Zhu et al., 2011), which means that southern Tibet may 2006; Kinoshita, 1995; Sisson et al., 2003; Tang et al., 2012a, b). have the thickest juvenile continental crust on Earth. Extensive studies of the effects of ridge subduction have been conduct- The growth of juvenile crust in southern Tibet was generally thought ed along the modern PacificRim(Karsten et al., 1996; Sisson et al., to have been associated with the northwards subduction of the Neo- 2003). These studies have mainly involved interpretations of geophysi- Tethyan oceanic slab beneath the Lhasa terrane and the consequential cal data and the disruption of surficial materials, and there is little melting of the subduction-modified mantle wedge (Chu et al., 2006; information about deep-seated processes; what little there is has been Mo et al., 2005; Ravikant et al., 2009; Zhu et al., 2011). However, some largely inferred from indirect evidence and from material exhumed authors have recently emphasized that subduction of the Neo-Tethyan from great depths (Pavlis and Sisson, 1995). If an ancient ridge subduc- mid-ocean ridge system may also play an important role in the Late tion system is exhumed from great depth, the complex structural, Cretaceous (Zhang et al., 2010b; Zhu et al., 2009b). The Late Cretaceous metamorphic, igneous, and sedimentary events revealed can provide Gangdese granitoids and metamorphic complex are particularly well insights into the evolutionary processes of the ridge subduction system. exposed and accessible in southeastern Tibet, which could provide However, there are only a few well-documented cases of ridge subduc- critical constraints on the evolutionary processes of ridge subduction tion systems in the ancient geological record (Pavlis and Sisson, 1995; and ridge–trench interaction, as well as the growth and chemical differ- Sisson et al., 2003; Tang et al., 2012a, b). entiation mechanisms of the thickest juvenile continental crust on Earth.

A 75°E 81°E 87°E 93°E 99°E 38°N Tarim Craton North China Kunlun-Qilian Terrane Craton 36°N Golmu Xining

S ongPan-Ganze Ter 34°N

Q rane BNS iangtang Terrane 32°N Shiquanhe Yare BNS STD YTS NL MFF CL 30°N Lhasa Terrane SL Lhasa Linzhi Yangtze 0 100 200 km Namche Xigaze YTS Barwa Craton 28°N N-S normal faults Himalayan Fig.1C Fig.1B Linzizong volcanic successions Qomolangma Gangdese batholiths STD GangdeseOther Mesozoic and correlated batholiths batholiths MFF 26°N Xigazi sediments 84°E Indian Plate90°E

o o o o o B 93 15' 93 30' 93 45' 94 00' 95 15' 96°E 102°E '

15 91°48´ 91°49´ o o Milin C 29°18´ 29 29 15' 95 Ma 88 Ma Quateanary Pt rock series 76 Ma 97 Ma 2-3 Wolong Two-mica granite 80 Ma Granodiorite 5' 76 Ma Lilong 86-90Ma 1 km o 80 Ma Charnockite 90

2 103 Ma Diorite-gabbro Quateanary Cretaceous limestone Langxian 93o 15'983 Ma 3o30'93o45' 94o 00' Fault 10 km Quartz-diorite Cretaceous granite

Fig. 1. (A) Simplified geological maps of the Tibetan–Himalayan orogen showing outcrops of the Gangdese batholiths along the south edge of the Lhasa terrane (after Chung et al., 2009; Zhao et al., 2009; Wu et al., 2010). Geological map showing outcrops of (B) the Langxian and (C) the Nuri suites on the southern margin of the Lhasa terrane (after Zhang et al., 2010b). Abbreviations: BNS = Bangong–Nujiang suture; YTS = Yarlung–Tsangpo suture; STD = south Tibet detachment system; MFF = main frontal fault, CL = central Lhasa subterrane, NL = northern Lhasa subterrane, SL = southern Lhasa subterrane. 242 Y. Zheng et al. / Lithos 190–191 (2014) 240–263

In this paper, we present whole-rock geochemical, Sr–Nd–O isotopic in this subterrane consist mainly of sandstone, slate, limestone, and data, zircon U–Pb ages, and in situ Hf–O isotopic data, for the Langxian mudstone (Fig. 1A) (Pan et al., 2004, 2012). and Nuri adakite-like igneous intrusive suites from the eastern part of Prior to the Indian–Asian collision, the Lhasa terrane underwent an southern Tibet. These new data, together with previously published Andean-type orogeny, during which the Neo-Tethyan oceanic slab data, provide new constraints on the mechanisms of growth and was subducted northward beneath the Eurasian plate in the Jurassic differentiation of the southern Tibetan crust during the Late Cretaceous, through Early Paleogene. The subsequent continent–continent collision which might be related to subduction of a Neo-Tethyan mid-ocean ridge between the Indian and Eurasian plates began in Late Paleocene times. system. Ultimately, these data provide new insights into the timing of Throughout this protracted and varied tectonic evolution, emplacement thickening of the continental crust of southern Tibet to twice the global of voluminous Gangdese arc batholiths and eruption of widespread average, and the time scales involved in uplift of the Tibetan Plateau. Linzizong volcanic successions took place across a vast tract of the southern Lhasa subterrane (red and blue in Fig. 1A) (Mo et al., 2005, 2007; Zhu et al., 2009b, 2011). For instance, the Gangdese batholiths 2. Geological setting record a long, punctuated history of magmatism spanning ~210 Ma to ~40 Ma, with peak episodes of granitoid plutonism at 109–80 Ma and The Tibetan Plateau is a tectonic collage of four continental blocks: 55–45 Ma, while the Linzizong volcanic successions were mainly the Songpan–Ganze belt, the Qiangtang terrane, the Lhasa block, and erupted at 65–42 Ma (Chu et al., 2006; Mo et al., 2005, 2007; Zhu the Himalaya. The Lhasa terrane represents the southernmost Asian et al., 2009b, 2011). Collectively, this Gangdese plutonism extends for block, and is bounded to the north by the Banggong–Nujiang suture N1500 km along an E–W trending belt within the Lhasa terrane that is and to the south by the Yarlung–Tsangpo suture (Fig. 1A). The sub-parallel to the Yarlung–Tsangpo suture (Fig. 1). A younger stage Banggong–Nujiang suture formed as a result of of granitoid plutonism that took place during the interval ca. 30 Ma to between the Lhasa and Qiangtang terranes during the Late Jurassic– 8 Ma (i.e., after the closure of the Neo-Tethyan ocean), comprising Early Cretaceous upon closure of the Bangong–Nujiang ocean (Hou ultrapotassic and adakite-like intrusions, also occurs within the Lhasa et al., 2009; Yin and Harrison, 2000). In contrast, the Yarlung–Tsangpo terrane (Miller et al., 1999; Sun et al., 2013; Turner et al., 1996; suture resulted from the northward drift of the Indian plate and its con- Williams et al., 2004; Zhao et al., 2009; Zheng et al., 2012a, b). sequential collision with the Eurasian plate at ca. 55–50 Ma (Besse et al., This study focuses particularly on understanding the petrogenesis of 1984; Klootwijk et al., 1992; Leech et al., 2005; Patriat and Achache, the Langxian and Nuri adakite-like granitoids of southern Tibet. The 1984). The INDEPTH II deep profiling results for southern Tibet indicate Langxian granitoids, which are located between Langxian and Milin that the continental crust there is 65–75 km thick (Molnar et al., 1998), (Fig. 1B), intrude into older country gneisses, and consist primarily of reaching a maximum thickness of 80 km beneath the Lhasa terrane granodiorite and two-mica granite intrusions. The country gneiss is (Kind et al., 1996). The crustal thickness of the Qiangtang terrane, how- mainly composed of amphibolite, marble, and schist (Zhang et al., ever, is commonly less than that of the Lhasa terrane (~50–60 km; 2010b). The Langxian granodiorite pluton is composed primarily of McNamara et al., 1995). All of these various estimates of crustal massive granodiorite, but contains decimeter-scale maficenclavesand thickness across the Tibetan Plateau are consistent with the results is crosscut by late, narrow, N–S trending lamprophyre dykes (Fig. 2A). determined from a number of other independent seismological studies Ten samples, including seven mafic enclaves and three host rocks (i.e., (Nelson et al., 1996; Priestley et al., 2006). massive granodiorite), were collected from this granodiorite pluton. In detail, the Lhasa terrane can be further subdivided into northern, The granodiorite host rocks are coarse-grained and consist mainly of central, and southern subterranes (Zhu et al., 2011). The central plagioclase (30%–40%), quartz (20%–30%), K-feldspar (10%–15%), subterrane was once a microcontinent containing a basement of ancient biotite (10%–15%), and hornblende (5%–10%) (Fig. 2B). The mafic Proterozoic and Archean rocks. This microcontinent acted as a nucleus enclaves themselves have a medium-grained igneous texture and are onto which juvenile Phanerozoic crust was accreted in the northern composed mainly of plagioclase (50%–60%), hornblende (15%–25%), and southern subterranes, during its journey drifting across the Tethyan and biotite (5%–10%) (Fig. 2C). Rock samples of the lamprophyre dyke Ocean basins and its ultimate continental collision. The Precambrian generally have a fine-grained and equigranular texture (Fig. 2D), with metamorphic basement of the central Lhasa subterrane is covered constituent mineral components that are similar in abundance to with widespread Permo-Carboniferous metasedimentary rocks, which those of the mafic enclaves in the Langxian granodiorite. The two- contain continental arc volcanic rocks and abundant glaciomarine mica granite rocks are muscovite-bearing and consist mainly of plagio- diamictites. Also present as cover successions are Upper Jurassic– clase (30%–35%), K-feldspar (30%–35%), quartz (20%–30%), muscovite Lower Cretaceous sedimentary rocks that are interlayered with (5%–10%), biotite (5%–10%), and epidote (b3%) (Fig. 2E), but show a abundant volcanic rocks and minor amounts of Ordovician, Silurian, distinct lack of hornblende that is so common in the other three previ- Devonian, and Triassic limestones (cf. Pan et al., 2004, 2012). Precam- ous types of igneous rocks described. The Nuri adakite-like suite brian basement rocks are most likely absent from the northern Lhasa (quartz–diorite in Fig. 1C), which is located on the hanging wall of the subterrane. The oldest sedimentary cover in this subterrane is Middle north-dipping Gangdese thrust fault, is intruded into a sequence of to Upper Triassic in age, and it consists mainly of slate, sandstone, and Jurassic sedimentary rocks and Cretaceous volcaniclastics and limestone radiolarian chert (Nimaciren et al., 2005; Pan et al., 2004, 2006). This (Harrison et al., 2000). This quartz–diorite has a medium-grained sequence is unconformably overlain by a Middle Jurassic unit composed igneous texture and is composed mainly of plagioclase (50%–60%), of coarse-grained clastic rocks, above which lies Upper Jurassic strata hornblende (10%–50%), quartz (5%–10%), and biotite (~5%), with comprising fine-grained clastic rocks (Nimaciren et al., 2005). The minor K-feldspar (Fig. 2F). Lower Cretaceous sequence includes slate, siltstone, and limestone with abundant volcanic rocks (Zhu et al., 2009b, 2011), and is uncon- 3. Results formably overlain by an Upper Cretaceous sequence interpreted as a terrigenous molasse deposit (Kapp et al., 2005, 2007; Pan et al., 2004, 3.1. U–Pb zircon ages of granitoids 2006). Precambrian basement rocks occur in the eastern part of the southern Lhasa subterrane. The volcano–sedimentary cover succession Samples BB–45 and BB–112 were collected from the Langxian mafic in the southern Lhasa subterrane–composed mainly of Late Carbonifer- enclaves and host granodiorite, respectively, while sample BB–42 was ous–Early Permian or Triassic clastic sedimentary rocks and abundant collected from a lamprophyre dyke. Samples BB–55, BB–113, BB–114, volcanic rocks–is largely restricted to the eastern part of this crustal and BB–116 were collected as representative samples of the Langxian segment. The Upper Jurassic–Cretaceous volcano–sedimentary strata two-mica granite, and sample NR–14 was collected from the Nuri Y. Zheng et al. / Lithos 190–191 (2014) 240–263 243

400um Bt Amp Kf Pl Ep

Mf Ep Amp Pl Lm Mf Gd Gd Ep Pl Qz A BB

400um 400um

Qz Amp

Qz Pl Bt Amp Bt Bt Qz Qz Pl Bt Amp Bt Pl Amp Pl CC DD

400um 400um Qz

Kf Qz Pl Qz Pl Amp Ep Pl Bt+Mus Amp Pl Pl Qz Qz Amp Pl EE F

Fig. 2. (A) Field photograph showing the Langxian granodiorite and lamprophyre. Microphotographs of (B) a granodiorite, (C) a mafic enclave, (D) a lamprophyre, (E) a two-mica granite from the Langxian granitoids, and (F) a quartz–diorite from the Nuri granitoids, southern Tibet. Abbreviations: Amp = amphibole, Bt = biotite, Ep = epidote, Gd = granodiorite, Kf = K-feldspar, Lm = lamprophyre dyke, Mf = mafic enclave, Mus = muscovite, Pl = plagioclase, Qz = quartz.

quartz–diorite. The zircons separated from all of these samples are 114, and BB–116 that showed clearly developed oscillatory growth characterized by euhedral to subhedral crystal shapes exhibiting long zoning. Seventeen spot analyses carried out on zircons from sample prismatic forms. Most of the zircons are transparent and colorless, and BB–55 yielded similar 206Pb/238U ages ranging from 73.6 to 80.8 Ma show euhedral oscillatory zoning in cathodoluminescence (CL) images with a mean age of 78.1 ± 0.9 Ma (MSWD = 1.2; Fig. 3D), with four (Suppl. Fig. 1). additional concordant data points yielded anomalously old 238U/206Pb All of the results from LA-ICP-MS and SIMS U–Pb zircon dating study zircon ages ranging from 89.4 to 351.0 Ma. In total, eleven spot analyses are listed in Supplementary Table 1 and shown on concordia diagrams carried out on zircons from sample BB–113 yielded similar 206Pb/238U in Fig. 3. Ten analyses of sample BB–45 have similar 206Pb/238U ages ages ranging from 74.8 to 80.4 Ma with a mean age of 76.3 ± 1.9 Ma ranging from 106 Ma to 108 Ma with a mean 206Pb/238U age of (MSWD = 0.28; Fig. 3E), although three analyses yielded concordant 106.4 ± 2.6 Ma (MSWD = 0.15; Fig. 3A). Twenty analyses of sample data points having significantly older 206Pb/238U ages. In all, nineteen BB–112 zircons yielded 206Pb/238U ages ranging from 99.8 Ma to concordant or nearly concordant data points were determined from 106.5 Ma, with a mean 206Pb/238U age of 103.1 ± 1.8 (MSWD = 0.28; spot analysis of sample BB–114 zircons, yielding 206Pb/238U ages that Fig. 3B) for 19 spots. Zircons from sample BB–42 yielded 206Pb/238U range from 75.0 to 91.7 Ma, and which can be divided into two distinct ages ranging from 95 to 99 Ma for a total of 20 spot analyses, which age groups (Fig. 3F). The younger age group has a mean 206Pb/238Uage have a mean 206Pb/238U age of 96.9 ± 0.7 Ma (MSWD = 1.5; Fig. 3C). of 79.7 ± 1.8 Ma (n = 13; MSWD = 0.55), while the older age group In order to constrain the timing of emplacement of the Langxian has a mean 206Pb/238U age of 89.7 ± 3.0 Ma (n = 6; MSWD = 0.15; two-mica granites in this study, we focused on LA-ICP-MS and SIMS Fig. 3F). Fifteen spot analyses of zircons from sample BB–116 yielded U–Pb dating of only those zircons from samples BB–55, BB–113, BB– mostly concordant data with similar 206Pb/238U ages ranging from 244 Y. Zheng et al. / Lithos 190–191 (2014) 240–263

135 118 A 0.0185 B 0.021 114 0.0175 125 110

106 U 0.019 8 8 0.0165 102 Pb/ U Pb/ 115 06 23 06 23 2 2 0.0155 98 0.017 94 105 0.0145 Mean= 106.4±2.6 Ma 90 Mean= 103.1±1.8 Ma n=10, MSWD=0.15 n = 19, MSWD = 0.28 0.015 0.0135 0.08 0.10 0.12 0.14 0.16 0.18 0.07 0.09 0.11 0.13 0.15 0.0164 104 C 0.019 D 120 0.0160 102 110 0.017 100 0.0156 100 98 0.015 0.0152 90 Pb /U

Pb /U 96 206 238

206 238 0.013 0.0148 80 94 without point 13 0.011 70 0.0144 92 Mean= 96.9 ± 0.7 Ma Mean=78.1±0.9 Ma n=21, MSWD = 1.5 n = 12, MSWD=1.2 90 60 0.0140 0.009 0.07 0.09 0.11 0.13 0.15 0.04 0.06 0.08 0.10 0.12 0.14 0.020 0.017 120 Mean = 89.7±3.0 Ma E F n = 6, MSWD = 0.15 0.018 110 98 0.015 94 0.016 100 90

U 86 8 90 38 0.014 0.013 82 b/ U Pb/ P 62 623 78 20

20 80 0.012 74 70 0.011 70 0.010 66 60 Mean = 76.3±1.9 Ma Mean = 79.7±1.8 Ma n = 11, MSWD = 0.28 n = 13, MSWD = 0.55 0.008 0.009 0.04 0.06 0.08 0.10 0.12 0.14 0.065 0.075 0.085 0.095 0.105 0.015 0.0185 G H 116 112 0.014 88 0.0175 108 84 0.013 0.0165 104 80 U 100 0.012 76 0.0155 Pb/ U Pb/

6238 96 206 238 72 20 0.011 0.0145 92 68 88 64 0.010 0.0135 Mean = 95.9±0.9 Ma Mean=76.1±2.1 Ma 84 n = 15, MSWD=1.5 n = 29, MSWD = 0.43

0.009 0.0125 0.066 0.074 0.082 0.090 0.00 0.04 0.08 0.12 0.16 0.20 0.24 207Pb/ 235 U 207Pb/ 235 U

Fig. 3. U–Pb ages of zircons from (A) maficenclaves(BB–45), (B) granodiorite (BB–112), (C) lamprophyre (BB–42) from the Langxian granitoids, and (D–J) two-mica granite (BB–55, BB– 113, BB–114 and BB–116), (H) quartz–diorite from the Nuri granitoids (NR–14), southern Tibet.

70.5 to 84.0 Ma, and a mean 206Pb/238U age of 76.1 ± 2.1 Ma yielded similar 206Pb/238U ages ranging from 94 to 99 Ma with a mean (MSWD = 1.5; Fig. 3G), although five additional concordant to nearly 206Pb/238U age of 95.9 ± 0.9 Ma (MSWD = 0.43; Fig. 3H). concordant data points yielded anomalously old 238U/206Pb zircon ages ranging from 95.2 to 322.6 Ma. 3.2. Geochemistry Thirty-two spot analyses were carried out on zircons from sample NR–14, and with the exception of one anomalously young and two The major and trace element data, and the whole-rock Sr–Nd anomalously old spot ages, the other twenty-nine spot analyses all isotopes for the rocks are listed in Table 2 and Table 3, respectively. Y. Zheng et al. / Lithos 190–191 (2014) 240–263 245

Table 1 Zircon Hf–O isotopic data for the Langxian and the Nuri granitoids, southern Tibet.

Sample # 176Hf/177Hf ±2σ 176Lu/177Hf ±2σ Age εHf(t) TDM (Ma) TDMC(Ma) δ18O(‰)±2σ

Maficenclave BB-45-01 0.283003 0.000018 0.001847 0.000152 106.0 10.4 361 505 BB-45-02 0.282997 0.000011 0.000742 0.000009 105.0 10.2 359 514 BB-45-03 0.283015 0.000015 0.000423 0.000005 109.0 11.0 330 467 BB-45-04 0.282983 0.000017 0.000540 0.000014 110.0 9.8 376 541 BB-45-05 0.282980 0.000015 0.000441 0.000004 104.0 9.6 380 551 BB-45-06 0.282999 0.000021 0.001962 0.000107 103.0 10.1 368 516 BB-45-07 0.282979 0.000020 0.001934 0.000023 107.0 9.5 396 557 BB-45-08 0.282984 0.000018 0.002891 0.000036 124.0 10.0 400 542 BB-45-09 0.282991 0.000017 0.001318 0.000073 107.0 10.0 372 527 Lamprophyre BB-42-1 0.283090 0.000032 0.100944 0.002308 98.0 6.9 −137 722 BB-42-2 0.283022 0.000024 0.014007 0.000558 96.0 10.1 498 515 BB-42-3 0.283095 0.000030 0.048962 0.002160 95.0 10.4 −792 498 BB-42-4 0.282933 0.000025 0.024805 0.000654 97.0 6.2 1234 757 BB-42-5 0.283014 0.000021 0.014505 0.000265 96.0 9.7 527 536 BB-42-6 0.283087 0.000039 0.080989 0.002879 97.0 8.1 −206 648 BB-42-7 0.283007 0.000026 0.030723 0.001082 97.0 8.5 1671 613 Two-mica granite BB-55-1 0.28296 0.00003 0.00081 0.00004 76.2 8.1 417 625 6.7 0.2 BB-55-2 0.28298 0.00003 0.00038 0.00002 77.0 9.2 374 560 6.8 0.3 BB-55-3 0.28289 0.00002 0.00231 0.00006 80.8 5.8 531 776 6.4 0.2 BB-55-4 0.28282 0.00004 0.00230 0.00001 78.8 3.5 627 925 6.6 0.4 BB-55-5 0.28242 0.00002 0.00178 0.00001 351.0 −5.2 1203 1684 10.0 0.2 BB-55-6 0.28297 0.00002 0.00180 0.00011 79.0 8.5 414 602 7.0 0.3 BB-55-7 0.28298 0.00004 0.00173 0.00005 77.7 8.8 399 582 6.6 0.3 BB-55-8 0.28302 0.00003 0.00065 0.00000 111.8 11.3 319 446 6.4 0.3 BB-55-9 0.28287 0.00002 0.00055 0.00003 78.3 5.2 532 812 6.3 0.2 BB-55-10 0.28289 0.00002 0.00166 0.00002 78.6 5.7 528 784 6.3 0.3 BB-55-11 0.28297 0.00004 0.00260 0.00009 103.2 9.2 413 577 6.0 0.3 BB-55-12 0.28282 0.00004 0.00329 0.00004 76.9 3.4 646 930 6.1 0.2 BB-55-13 0.28288 0.00004 0.00109 0.00000 73.6 5.4 527 796 6.7 0.3 BB-55-14 0.28293 0.00003 0.00044 0.00001 89.4 7.4 454 680 6.1 0.3 BB-55-15 0.28287 0.00003 0.00138 0.00004 77.7 5.0 550 825 6.6 0.2 BB-55-16 0.28291 0.00003 0.00209 0.00002 79.6 6.7 493 722 6.6 0.2 BB-55-17 0.28291 0.00003 0.00209 0.00002 76.5 6.6 493 723 6.0 0.2 BB114-01 0.283006 0.000018 0.002352 0.000155 83.2 10.0 361 511 BB114-02 0.282991 0.000016 0.001317 0.000005 79.8 9.4 372 543 BB114-04 0.283048 0.000018 0.000324 0.000006 90.9 11.7 283 404 BB114-05 0.283027 0.000020 0.002486 0.000002 79.4 10.6 332 467 BB114-09 0.283054 0.000020 0.002039 0.000010 78.8 11.6 288 404 BB114-10 0.283030 0.000018 0.002206 0.000003 80 10.8 325 459 BB114-11 0.282992 0.000018 0.001343 0.000018 79.3 9.5 371 541 BB114-12 0.283066 0.000020 0.000792 0.000018 75 12.0 261 374 BB114-13 0.283033 0.000019 0.000665 0.000006 80.2 11.0 307 446 BB114-14 0.282974 0.000019 0.000723 0.000002 80.4 8.9 391 580 BB114-15 0.283039 0.000017 0.000431 0.000003 82.2 11.2 297 431 BB114-16 0.283010 0.000020 0.000559 0.000018 79.5 10.1 339 499 BB114-17 0.28288 0.000017 0.001288 0.000008 76.3 9.3 377 553 BB114-22 0.283058 0.000019 0.001066 0.000016 83.7 11.9 274 388

176 177 176 177 176 177 176 177 λt 176 177 Footnote: εHf(t) = 10,000{[( Hf/ Hf)S − ( Lu/ Hf)S ×(eλt − 1)]/[( Hf/ Hf)CHUR,0 − ( Lu/ Hf)CHUR ×(e − 1)] − 1}; TDM =1/λ ×ln{1+[( Hf/ Hf)S − 176 177 176 177 176 177 C 176 177 176 177 176 177 176 177 176 177 176 177 ( Hf/ Hf)DM]/[( Lu/ Hf)S − ( Lu/ Hf)DM]}; TDM =1/λ ×ln{1+[( Hf/ Hf)S, t − ( Hf/ Hf)DM, t]/[( Lu/ Hf)C − ( Lu/ Hf)DM]} + t. The Hf/ Hf and Lu/ Hf ratios of chondrite and depleted mantle at the present are 0.282772 and 0.0332, 0.28325 and 0.0384, respectively (Griffinetal.2006). λ = 1.867 × 10−11a−1 (Soderlund et al., 2004). 176 177 ( Lu/ Hf)C = 0.015, t = crystallization age of zircon.

The term “adakite” was first proposed by Defant and Drummond (1990) the adakite-like melts crystallized (Castillo et al., 1999; Defant and to represent silica-rich, high Sr/Y and La/Yb volcanic rocks of to Drummond, 1990; Macpherson et al., 2006; Peacock et al., 1994). sodic rhyolite compositions that originated by partial melting of hydrat- With the exception of the mafic enclaves and lamprophyre dyke that ed oceanic slab (Martin et al., 2005; Rapp et al., 1999). Adapting the occur in the Langxian suite, all other whole-rock samples documented geochemical criteria by Defant and Drummond (1990), various other in this study exhibit chemical features suggestive of adakite-like affinity, petrogenetic models have previously been proposed for the origin of characterized by relatively high Al2O3 (N15 wt.%), Sr (N400 ppm) adakite-like rocks in arc systems: fractional crystallization of basaltic contents, high Sr/Y (N40) ratios, and low HREE (e.g., Yb b 1.9 ppm) melts (a) under low-pressure conditions (Castillo et al., 1999) or (b) and Y (b18 ppm) contents, and positive Sr and Eu anomalies under high-pressure conditions (Macpherson et al., 2006); (c) melting (Figs. 4–7). These particular adakite-like samples show fractionated of foundering lower continental crust (Gao et al., 2004; Stern and REE patterns and relatively low HREE contents and high (La/Yb)N ratios. Hanson, 1991; Wang et al., 2007; Xu et al., 2002); and (d) melting of These rocks also have low concentrations of high field strength thickened mafic lower crust (Atherton and Petford, 1993; Chung et al., elements (HFSEs), show strong enrichment in large ion lithophile ele- 2003; Hou et al., 2004; Zheng et al., 2012a, b). The high Sr/Y and La/Yb ments (LILEs: Rb, Sr, and Ba), and exhibit pronounced negative Nb–Ta characteristics are interpreted to reflect both the presence of residual anomalies and positive K and Pb anomalies on primitive mantle- garnet and the absence of plagioclase in either the source region of normalized incompatible element plots (Fig. 7), which are typical dis- adakite magmas, or in the actual magma chamber itself from which criminating features of adakite-like magmas (Defant and Drummond, 246 Y. Zheng et al. / Lithos 190–191 (2014) 240–263

Table 2 Major and trace element analysis of rocks from the Langxian and the Nuri granitoids, southern Tibet.

Sample# Granodiorite Maficenclave Two-mica granite

BB-41 BB-43 BB-112 BB-44 BB-45 BB-46 BB-47 BB-48 BB-50 BB-51 BB-53 BB-56

SiO2(%) 62.48 61.22 59.19 47.25 47.13 49.50 50.36 47.48 48.99 49.61 72.61 73.03 TiO2 0.58 0.58 0.69 1.02 1.05 0.92 0.95 1.10 0.97 1.07 0.10 0.04 Al2O3 16.99 17.21 17.24 18.86 18.82 18.53 18.05 18.94 18.64 18.57 15.01 15.59 Fe2O3 5.05 5.67 7.66 10.87 10.76 9.87 10.18 11.36 10.69 10.10 0.87 0.33 MnO 0.10 0.10 0.12 0.18 0.16 0.19 0.20 0.20 0.18 0.18 0.06 0.02 MgO 2.33 2.56 3.34 6.15 6.30 5.62 5.55 6.05 5.70 5.40 0.08 0.07 CaO 5.40 5.57 5.49 9.54 10.08 8.81 8.49 8.63 9.05 7.61 1.77 1.85 Na2O 3.80 4.14 3.71 2.84 2.78 3.49 3.74 3.11 3.83 3.56 4.62 5.30 K2O 1.82 1.54 1.61 1.67 1.30 1.57 1.56 2.25 1.22 2.19 3.24 2.93 P2O5 0.22 0.24 0.20 0.22 0.17 0.18 0.17 0.25 0.21 0.29 0.06 0.04 H2O+ 0.96 0.82 0.72 1.72 1.38 1.38 1.28 1.04 1.30 1.38 0.74 0.32 CO2 0.26 0.61 0.26 0.44 0.52 0.56 0.26 0.17 0.09 0.35 0.26 0.09 LOI 0.92 1.01 1.18 1.54 1.17 1.07 0.97 0.94 1.10 1.04 1.02 0.39 Mg#50.449.849.055.556.355.654.554.054.054.116.831.8 A/CNK 0.94 0.93 0.97 0.79 0.77 0.79 0.78 0.81 0.77 0.84 1.05 1.02 ANK 2.07 2.03 2.20 2.91 3.15 2.49 2.30 2.51 2.45 2.26 1.35 1.31 Li(ppm)24.027.642.227.120.525.725.438.516.049.226.216.3 Sc 11.6 12.3 15.5 30.2 34.1 28.5 30.7 29.2 28.6 22.2 2.3 1.4 Rb 42.2 41.2 54.3 47.1 23.0 33.2 32.2 53.5 21.8 55.6 87.3 76.6 Sr 667 709 599 553 597 602 601 547 588 627 423 336 Y 11.90 11.70 12.60 17.90 17.70 20.30 26.00 20.35 22.40 15.60 6.46 4.08 Zr 41.4 51.4 58.0 27.6 37.2 42.8 24.1 22.7 28.8 39.1 56.5 36.8 Nb 5.13 4.67 5.19 3.47 3.18 4.68 6.08 4.83 4.42 4.94 5.31 2.94 Cs 1.65 2.02 2.90 2.43 1.19 1.64 1.53 2.84 0.83 2.45 1.49 1.45 Ba 404 232 216 264 212 276 278 395 184 447 417 262 La 17.8 35.0 5.5 10.4 13.4 7.9 19.0 14.3 16.2 11.6 11.8 4.8 Ce 37.4 67.0 13.6 24.4 29.9 23.6 46.9 34.8 37.7 27.8 20.4 8.0 Pr 4.35 7.09 2.36 3.21 3.65 3.65 6.25 4.54 4.85 3.74 2.52 1.00 Nd 17.00 25.10 11.40 14.20 15.10 17.30 27.20 19.45 20.40 16.90 8.81 3.48 Sm 3.30 3.86 3.13 3.33 3.21 4.21 5.98 4.31 4.41 4.09 1.57 0.67 Eu 0.95 1.03 1.07 1.18 1.08 1.34 1.61 1.43 1.34 1.33 0.37 0.28 Gd 3.04 3.24 3.17 3.48 3.37 4.14 5.68 4.17 4.30 3.89 1.37 0.69 Tb 0.38 0.42 0.43 0.50 0.47 0.57 0.78 0.59 0.63 0.50 0.18 0.11 Dy 2.14 2.20 2.47 3.11 3.04 3.54 4.65 3.55 3.89 2.86 1.10 0.64 Ho 0.43 0.41 0.50 0.64 0.63 0.71 0.90 0.72 0.78 0.55 0.20 0.12 Er 1.16 1.18 1.43 1.89 1.84 2.08 2.61 2.12 2.37 1.55 0.57 0.36 Tm 0.15 0.16 0.18 0.27 0.26 0.29 0.35 0.29 0.33 0.21 0.08 0.05 Yb 1.04 0.99 1.25 1.74 1.68 1.97 2.32 1.93 2.17 1.39 0.54 0.32 Lu 0.16 0.15 0.18 0.26 0.26 0.29 0.34 0.29 0.32 0.20 0.08 0.05 Hf 1.32 1.51 2.01 1.18 1.40 1.47 1.28 1.09 1.30 1.34 1.99 1.45 Ta 0.38 0.34 0.47 0.25 0.18 0.31 0.36 0.31 0.30 0.30 0.56 0.42 Pb 9.67 9.19 9.27 5.90 6.25 6.90 7.03 7.44 6.19 6.85 25.40 16.20 Th 5.74 10.50 1.48 1.85 1.59 0.80 0.94 1.45 1.79 1.24 3.19 1.08 U 1.18 1.29 1.30 0.59 0.58 0.56 0.32 0.61 0.60 0.41 1.10 1.37 V 105.00 119.00 179.00 327.00 351.00 294.00 291.00 331.50 316.00 281.00 5.19 2.19 Cr 36.70 23.20 51.40 29.50 22.00 23.80 18.20 23.45 9.71 65.20 6.46 0.59 Co 13.20 14.10 19.30 32.90 34.30 30.70 30.80 34.60 32.30 30.20 0.93 0.33 Ni 17.40 15.90 22.00 30.60 28.60 26.70 24.90 26.50 21.40 31.70 1.69 0.79 Sr/Y 56.1 60.6 47.5 30.9 33.7 29.7 23.1 26.9 26.3 40.2 65.5 82.4 δEu 0.87 0.83 0.99 1.01 0.96 0.93 0.80 0.98 0.90 0.97 0.73 1.20 (La/Yb)n 12.3 25.4 3.2 4.3 5.7 2.9 5.9 5.3 5.4 6.0 15.7 10.8

Sample# Two-mica granite

BB-59 BB-61 BB-62 BB-68 BB-70 BB-72 BB-74 BB-113 BB-114 BB-115 BB-116

SiO2(%) 72.01 71.39 74.34 66.44 68.23 67.80 69.98 73.33 72.36 66.65 67.70 TiO2 0.14 0.15 0.08 0.33 0.26 0.23 0.22 0.09 0.07 0.30 0.25 Al2O3 16.01 16.6 14.66 16.85 17.15 17.86 16.45 15.16 15.62 17.56 18.24 Fe2O3 1.20 1.44 0.92 3.08 2.43 2.13 1.77 1.00 0.74 3.06 2.36 MnO 0.05 0.08 0.05 0.10 0.07 0.06 0.05 0.07 0.04 0.09 0.06 MgO 0.12 0.19 0.04 0.94 0.67 0.52 0.40 0.39 0.21 0.84 0.48 CaO 2.27 2.53 1.56 3.77 3.15 3.22 3.10 1.70 1.69 3.93 3.54 Na2O 4.72 5.12 4.07 4.71 4.68 5.76 4.97 4.22 4.06 4.57 5.41 K2O 3.54 2.65 4.38 2.23 3.02 2.14 2.32 3.61 4.59 2.52 1.85 P2O5 0.04 0.05 0.03 0.15 0.12 0.08 0.07 0.04 0.03 0.15 0.08 H2O+ 0.48 0.40 0.16 0.94 0.34 0.32 0.22 0.24 0.12 0.14 0.34 CO2 0.09 0.09 0.09 0.44 0.09 0.17 0.17 0.17 0.26 0.35 0.17 LOI 0.38 0.37 0.23 1.00 0.29 0.48 0.23 0.5 0.33 0.80 0.36 Mg# 18.0 22.5 8.7 40.2 37.8 35.0 33.2 46.2 38.4 37.7 30.9 A/CNK 1.02 1.05 1.03 0.99 1.03 1.01 1.01 1.09 1.06 1.01 1.05 ANK 1.38 1.48 1.28 1.66 1.56 1.51 1.54 1.40 1.34 1.71 1.67 Li (ppm) 33.0 50.7 35.6 22.2 35.5 18.8 18.3 33.5 22.3 24.7 18.6 Sc 2.8 3.0 1.6 3.8 4.1 3.8 3.0 1.7 1.7 3.8 2.2 Rb 93.6 89.6 116.0 42.5 66.3 34.7 41.7 102.0 152.0 53.4 39.4 Y. Zheng et al. / Lithos 190–191 (2014) 240–263 247

TableTable 2 ( 2continued(continued) ) Sample# Two-mica granite

BB-59 BB-61 BB-62 BB-68 BB-70 BB-72 BB-74 BB-113 BB-114 BB-115 BB-116

Sr 556 512 278 674 627 743 711 296 265 594 675 Y 6.46 6.75 7.92 9.56 7.29 4.95 6.89 7.21 11.20 7.25 5.58 Zr 63.5 72.8 40.1 74.5 57.2 96.7 67.3 45.8 51.3 96.8 136.0 Nb 5.67 7.66 5.05 4.24 4.78 3.37 2.7 5.71 4.93 3.93 3.61 Cs 3.67 3.98 5.77 1.89 2.83 0.97 0.78 1.10 8.37 2.35 0.90 Ba 580 295 316 564 702 545 518 327 409 565 398 La 19.9 13.7 9.7 14.5 17.6 19.7 15.8 7.6 6.9 22.0 24.3 Ce 37.5 24.1 17.5 30.5 30.0 34.5 29.2 15.2 12.7 36.6 42.0 Pr 4.10 2.84 2.12 3.57 3.57 3.96 3.27 1.70 1.60 3.91 4.98 Nd 14.30 9.86 7.52 13.70 12.90 14.00 11.60 5.95 5.95 13.40 17.80 Sm 2.35 1.81 1.48 2.47 2.07 2.15 1.87 1.16 1.31 1.97 2.66 Eu 0.55 0.49 0.40 0.77 0.68 0.64 0.61 0.30 0.50 0.73 0.76 Gd 1.83 1.45 1.41 2.23 1.82 1.64 1.43 1.20 1.21 1.78 1.82 Tb 0.22 0.21 0.21 0.30 0.23 0.18 0.17 0.18 0.18 0.23 0.22 Dy 1.18 1.17 1.27 1.64 1.20 0.91 0.99 1.11 1.10 1.25 1.08 Ho 0.20 0.20 0.25 0.31 0.23 0.17 0.20 0.22 0.22 0.26 0.21 Er 0.59 0.61 0.72 0.97 0.69 0.51 0.69 0.72 0.66 0.81 0.67 Tm 0.08 0.09 0.11 0.15 0.10 0.07 0.08 0.10 0.09 0.12 0.09 Yb 0.59 0.65 0.74 0.98 0.72 0.48 0.56 0.69 0.58 0.90 0.70 Lu 0.09 0.10 0.12 0.15 0.11 0.08 0.10 0.10 0.09 0.16 0.12 Hf 1.99 2.28 1.53 2.17 1.73 2.61 1.89 1.73 2.00 3.33 4.13 Ta 0.49 0.81 0.54 0.29 0.37 0.14 0.20 0.51 0.57 0.32 0.19 Pb 21.10 20.00 25.30 13.20 18.30 15.8 16.20 18.00 31.00 15.50 15.60 Th 6.86 3.99 5.24 2.08 5.34 3.69 2.95 2.19 4.23 3.1 5.00 U 0.59 0.56 1.01 0.55 0.91 0.50 0.71 0.45 0.64 0.67 0.64 V 7.48 8.40 4.69 31.80 25.30 15.60 15.2 11.40 7.82 35.30 19.00 Cr 1.35 0.42 0.42 4.67 1.11 5.47 0.55 3.17 1.95 3.57 2.28 Co 1.16 1.43 0.65 4.47 3.60 2.33 2.30 0.67 0.58 4.28 2.52 Ni 1.29 1.08 0.80 2.38 2.07 2.20 1.45 1.97 1.41 2.36 2.01 Sr/Y 86.1 75.9 35.1 70.5 86.0 150.1 103.2 41.1 23.7 81.9 121.0 δEu 0.75 0.86 0.80 0.95 1.01 0.96 1.05 0.74 1.15 1.13 0.96 (La/Yb)n 24.2 15.1 9.4 10.6 17.5 29.4 20.2 7.9 8.6 17.5 24.9

Sample# Nuri diorite

cb-70 NR-15 NR-16 NR -17 ZK45-711 ZK45-765

SiO2(%) 64.12 59.31 61.24 57.70 64.39 57.77 TiO2 0.67 0.82 0.73 0.85 0.61 0.81 Al2O3 15.53 16.77 16.31 15.57 15.69 15.41 Fe2O3 4.78 6.26 4.90 6.06 5.21 6.60 MnO 0.11 0.10 0.09 0.12 0.09 0.15 MgO 2.33 4.06 3.59 4.95 2.16 5.09 CaO 6.16 5.27 5.44 6.21 7.17 7.17 Na2O 3.15 3.54 3.28 3.72 2.05 3.26 K2O 1.03 2.09 1.90 2.46 1.08 1.41 P2O5 0.24 0.33 0.33 0.33 0.22 0.34 H2O+ 1.34 1.94 2.04 1.06 1.56 CO2 0.44 0.17 0.17 0.35 0.35 LOI 1.24 1.60 1.70 1.58 1.10 1.37 Mg# 51.8 58.8 61.7 64.3 47.7 62.9 A/CNK 0.89 0.95 0.94 0.78 0.89 0.77 ANK 2.47 2.07 2.19 1.77 3.45 2.24 Li (ppm) 12.6 17.1 10.1 11.5 10.8 13.1 Sc 10.5 13.1 12 16.3 10.7 15.8 Rb 73.5 85.7 83.8 84.9 77.1 64.7 Sr 803 917 837 1004 753 942 Y 13.1 11.7 10.9 13.5 12.3 11.8 Zr 176 97.4 98.2 73.7 171 78.2 Nb 6.04 4.47 5.03 4.75 5.31 3.93 Cs 7.31 4.21 3.37 3.44 8.92 4.42 Ba 156 446 430 457 133 308 La 29.9 24.5 25.8 25.3 28.6 22.3 Ce 60.9 47.5 49.3 51.9 56.1 45.1 Pr 6.77 6.28 6.46 6.60 7.14 6.14 Nd 25.0 25 24.8 26.3 27 25 Sm 4.36 4.58 4.49 5.26 4.63 4.75 Eu 1.34 1.39 1.32 1.31 1.62 1.29 Gd 3.36 3.79 3.58 3.78 3.8 3.9 Tb 0.48 0.5 0.44 0.51 0.49 0.49 Dy 2.33 2.55 2.31 2.75 2.59 2.59 Ho 0.47 0.49 0.44 0.48 0.5 0.46 Er 1.27 1.36 1.29 1.38 1.5 1.41 Tm 0.17 0.17 0.17 0.17 0.19 0.17 Yb 1.05 1.12 1.04 1.13 1.34 1.11

(continued on next page) 248 Y. Zheng et al. / Lithos 190–191 (2014) 240–263

TableTable 2 ( 2continued(continued) ) Sample# Nuri diorite

cb-70 NR-15 NR-16 NR -17 ZK45-711 ZK45-765

Lu 0.16 0.18 0.15 0.17 0.18 0.16 Hf 3.98 3 2.95 2.12 5.18 2.69 Ta 0.42 0.31 0.34 0.73 0.43 0.28 Pb 10.5 9.21 9.37 0.98 9.95 9.84 Th 9.10 4.68 5.21 4.74 9.06 3.96 U 2.75 1.03 1.34 1.10 3.25 1.23 V 118 146 132 167 119 172 Cr 76.6 100 97.9 150 73.5 167 Co 14.4 20.2 9.63 25.7 11.6 22.7 Ni 37.2 82.7 47.3 87.9 29 73.2 Sr/Y 61.3 78.4 76.8 74.3 61.2 79.8 δEu 0.99 0.95 0.94 0.82 1.10 0.86 (La/Yb)n 20.4 15.7 17.8 16.0 15.3 14.4

1990; Martin, 1999; Martin et al., 2005). In addition, in the Sr/Y vs. Y and contents and Th/Ce, Th/Yb, and Th/Sm ratios of all rock samples

(La/Yb)N vs. YbN discrimination diagrams (Defant and Drummond, described in this study. However, the lamprophyre dyke sample has 1990), felsic rocks from the Langxian and Nuri suites plot in the adakite somewhat similar trace element contents as the enclaves (Fig. 7A). field, while the mafic rocks in the Langxian suite plot in the normal arc Isotopically, rock samples of the granodiorite have Nd isotopic com- 87 86 magma field (Fig. 6A, B). positions (εNd(t)) varying from 1.8 to 3.2, with Sr/ Sr(i) values ranging There are some other geochemical characteristics of samples report- from 0.7042 to 0.7044, and show whole-rock δ18O values of 7.0 to 8.1‰, 87 86 ed in this study, such as REE and LILE contents, LREE/HREE, LILE/LREE, whereas the mafic enclave sample BB–44 has an εNd(t) of 3.3, a Sr/ Sr(i) LREE/HFSE, and whole-rock Sr–Nd and zircon Hf isotopic compositions, value of 0.7047 and a δ18O value of 7.8‰ (Fig. 8A). The calculated initial which may actually reflect magma source compositions rather than εHf(t) values of zircons from the mafic enclaves are high at 9.5–10.9 source mineralogy, and vary significantly between individual magmas (Fig. 8). Sample BB–42 of the lamprophyre dyke has similar whole- (Figs. 5–9). Consequently, in presenting our geochemical data for rock Sr–Nd–O and zircon Hf isotopic compositions as the mafic enclaves 87 86 adakites in this study, we describe the results for each igneous rock (Fig. 8). This sample has an εNd(t) value of 2.8, an Sr/ Sr(i) value of type individually. 0.7039, and a δ18O value of 7.1‰ and zircons with calculated initial

εHf(t) values of 6.2 to 10.4 (Fig. 8). 3.2.1. The Langxian granodiorite The Langxian granodiorite is classified as metaluminous, and has 3.2.2. The Langxian two-mica granite

SiO2 values ranging from 59.2 to 62.5 wt.% (Fig. 4A, B). Rock samples Rock samples from the Langxian two-mica granite have significantly from this intrusion are CaO-rich and medium-K in nature (Fig. 4A), have higher SiO2,K2O, Rb, Ba, Cs, Pb, Th, and U concentrations, but consider- relatively high MgO contents (Fig. 5A, B), and have low concentrations ably lower MgO and compatible element contents, than either the of compatible elements (e.g., Cr, Ni; Fig. 5C, D). In addition, these granodi- granodiorite host and mafic enclaves of the Langxian granodiorite, or orite samples exhibit moderately fractionated REE patterns (Fig. 7A). the Nuri quartz–diorite. They have similar LREE but significantly lower Mafic enclaves hosted by the Langxian granodiorite are also MREE and HREE contents than the host and maficenclavesofthe metaluminous (Fig. 5B), and have relatively high MgO (Fig. 5A, B) and Langxian granodiorite, indicating pronounced REE fractionation. In compatible element contents compared with their host rocks (Fig. 5C, addition, the LILE contents, and Th/Ce, Th/Yb, and Th/Sm ratios of the D). These enclaves exhibit weakly fractionated REE patterns and no Langxian two-mica granite samples are higher than in all other rock significant Eu anomalies (Fig. 7A), and they have the lowest LILE samples described in this study.

Table 3 Sr–Nd isotopic data for rocks from the Langxian and the Nuri granitoids, southern Tibet.

#8786 87 86 87 86 143 144 147 144 143 144 Sample Rb/ Sr Sr/ Sr ±2σ ( Sr/ Sr)i Nd/ Nd Sm/ Nd ±2σ ( Nd/ Nd)i εNd(t) TDM2 (Ma) fSm/Nd Granodiorite BB-41 0.183 0.704566 9 0.704400 0.512747 0.117344 12 0.512669 3.16 641 -0.40 BB-112 0.262 0.704633 15 0.704249 0.512785 0.274561 9 0.512600 1.84 1322 0.40

Lamprophyre BB-42 0.415 0.704846 11 0.703880 0.512735 0.127471 15 0.512654 2.75 708 -0.35

Mafic enclave BB-44 0.247 0.704683 9 0.704460 0.512767 0.141760 11 0.512670 3.25 724 -0.28

Two-mica granite BB-59 0.487 0.705405 20 0.704964 0.512629 0.099341 10 0.512578 0.79 744 -0.49 BB-66 0.459 0.704999 10 0.704584 0.512726 0.124310 10 0.512659 2.47 707 -0.37 BB-68 0.183 0.705427 24 0.705262 0.512708 0.108986 12 0.512650 2.28 664 -0.45 BB-113 0.998 0.705783 9 0.704677 0.512658 0.194958 7 0.512559 0.41 1143 -0.01 BB-114 1.661 0.706185 15 0.704345 0.512699 0.220168 12 0.512587 0.96 1217 0.12 BB-115 0.260 0.704720 14 0.704432 0.512683 0.147015 5 0.512608 1.37 882 -0.25 BB-116 0.169 0.704561 9 0.704374 0.512707 0.149438 10 0.512631 1.82 855 -0.24

Nuri diorite NR-15 0.271 0.704344 15 0.703979 0.512771 0.183200 12 0.512657 2.76 911 -0.07 NR-16 0.290 0.704417 14 0.704026 0.512765 0.181048 9 0.512652 2.67 911 -0.08

87 86 143 144 The initial εNd values and ( Sr/ Sr)i ratios were calculated at the time of crystallization ages of the igneous rocks. εNd was calculated relative to Nd/ Nd(CHUR) = 0.512638. In the TDM 143 144 147 144 calculation, Nd/ NdDM = 0.51315, and Sm/ NdDM = 0.21357. Y. Zheng et al. / Lithos 190–191 (2014) 240–263 249

10 Langxian granodiorite 3.6 Langxian mafic enclave this A B Langxian two-mica granite study 3.2 Adakites related to 8 Lamprophyre dyke slab melting Nuri quartz-diorite Mamen adakitic rock (137Ma) 2.8 Lilong charnockite (86-90Ma) 6 Lilong mafic rock (88-98Ma) 2.4 Post-collisional adakite-like rock Metaluminous Peraluminous shoshonitic A/NK 2.0 2 4 alkaline KO (wt.%) KO K calc- 1.6 h- hig 2 1.2 Adakite-like rocks related kaline to lower crust melting calc-al 0.8 low-k calc-alkaline 0 45 50 55 60 65 70 75 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3

SiO2 (wt.%) A/CNK

Fig. 4. (A) Rock suite classification in the K2O–SiO2 plot according to Peccerillo and Taylor (1976) and (B) A/CNK (Al2O3/(CaO + Na2O+K2O)) vs. A/NK (Al2O3/(Na2O+K2O)) diagram for rocks from the Langxian and the Nuri suites, southern Tibet. The field of adakites related to slab melting and lower crustal melting are from Condie (2005) and Zhu et al. (2009a).Data for the Mamen adakite are from Zhu et al. (2009a). Data for the post-collisional adakite-like rocks (26–10 Ma) are from Hou et al. (2004); Guo et al. (2007), Gao et al. (2007, 2010), Xu et al. (2010),andZheng et al. (2012a, b). Data for the Lilong charnockite and the Lilong mafic rocks are from Zhang et al. (2010b) and Guan et al. (2011), respectively.

87 86 18 206 238 The whole-rock Sr/ Sr(i) (0.7043–0.7053) and δ O (8.2–9.1‰) closed Pb/ U ages with relatively low MSWD values. Thus we isotopic ratios of these two-mica granites are higher, but whole-rock regard these mean 206Pb/238U ages as representing the timing of emplace-

εNd(t) (0.4–2.3) values are lower than the mafic enclaves and host ment and crystallization of these granitoids and their maficenclaves. rocks of the Langxian granodiorite (Fig. 8A). Meanwhile, zircons from A large population of relatively old zircons has been identified in the two samples of the two mica-granite show different initial εHf(t) values. samples from the Langxian two-mica granite, and these zircons are Sample BB–114 have very similar initial εHf(t) values (7.1–10.1) to commonly interpreted as evidence for inherited xenocrystic zircon zircons from the mafic enclaves and lamprophyre dyke that occur in from the magma source regions (Zheng et al., 2012a), or xenocrystic the Langxian granodiorite, while sample BB–55 have relatively lower zircon inherited from the wall rocks as the magma ascended upwards initial εHf(t) values (3.6–9.2) (Fig. 8). These zircons from BB–55 are char- through the crust (Hou et al., 2012; Zeng et al., 2011). Some authors, acterized by a narrow range of δ18O values varying from 6.1 to 7.0‰. however, have argued instead that the old zircons represent the initial Five spot analyses were made on the zircon cores from the two samples. stages of crystallization of magma in a long lived magmatic system, The analyses spot with 238U/206Pb age at 351 Ma is characterized while the younger zircons represent the timing of emplacement and 18 negative initial εHf(t) value at −5.2 and clearly high δ O value at 10‰, final crystallization of the magma (Yang et al., 2011; Zheng et al., while the other inherited zircons (89.4–111.8 Ma) have initial εHf(t) 2012b). In such a case, the zircons from both age groups should have values ranging from 7.4 to 11.7, which are also very similar to zircons similar textures in CL images, and also similar Hf isotopic compositions, from the Langxian granodiorite suites, with δ18O value from 6.0 to 6.4‰. but this is contradicted by the characteristics of old zircons in the two- mica granite, which have significantly different core–rim textures, as observed in CL images, and variable core-to-rim Hf–Oisotopic 3.2.3. The Nuri quartz–diorite compositions (cf., Chu et al., 2011; Chung et al., 2009). Additionally, Rock samples of the Nuri quartz–diorite are characterized by a limit- the 206Pb/238U ages of the old zircons in samples BB–113 and BB–114 ed range in SiO contents (57.7–64.4 wt.%), and are predominantly 2 are ~30 Ma and ~10 Ma older than the younger zircon age populations, classified as metaluminous, medium-K calc-alkaline rocks (Fig. 4). respectively. Thus, it is difficult to envision how the two-mica granites These samples have lower K O and LILE contents and Th/Ce, Th/Yb, 2 can be ‘evolved melts’ that were derived from relatively mafic melts, and Th/Sm ratios, but higher MREE and HREE contents than rocks persisting for ~30 Ma after crystallization of the initial melts. On these from the Langxian two-mica granite, and overall, they are geochemical- grounds, therefore, we support the concept that the zircons with older ly quite similar to the Langxian granodiorite (Figs. 4–7). However, the 206Pb/238U ages in the two-mica granites are xenocrysts. Most of the CaO, MgO (Fig. 4), and compatible element contents of the Nuri magmatic domains (rims) of the zircon crystals from the four Linzhi quartz–diorite are significantly higher than in the Langxian igneous granitoid samples yield similar 206Pb/238U age of around 78 ± 2 Ma, suite, even the mafic enclaves. The two Nuri quartz–diorites have and we interpret this age as representing the time of crystallization of whole-rock ε values of 2.7 and 2.8, respectively, with 87Sr/86Sr Nd(t) (i) the Langxian two-mica granites. values at 0.7040 and δ18O values at 8.9 and 9.2‰ (Fig. 8A).

4. Discussion 4.2. Origin of the Langxian and Nuri suites

4.1. Interpretation of the U–Pb ages for the Langxian and Nuri granitoids As outlined above, all felsic rocks sampled from the Langxian and Nuri igneous suites exhibit geochemical affinities with adakite-like The relatively high Th/U ratios and the well-developed oscillatory rocks (Fig. 5E, F). However, they also show considerable geochemical growth zoning in the zircons (both cores and rims) from the Langxian differences with regards to LILE concentrations, LREE/HREE, Th/Ce, and Nuri specimens indicate that they are all of magmatic origin. Th/Yb, and Th/Sm ratios, and whole rock Sr–Nd–O and zircon Hf iso- Apart from two analysis spots for sample BB–45 (mafic enclave), zircons topic compositions, suggesting contributions from variable proportions of the granodiorite and its mafic enclaves, the lamprophyre from the of a number of different source materials in the generation of these Langxian district, and the quartz–diorite from the Nuri area, all have adakite-like igneous intrusions of southern Tibet. 250 Y. Zheng et al. / Lithos 190–191 (2014) 240–263

15 1.2 20 A B C 1.0

18 10 0.8

0.6 16 5 0.4 14 0.2

Al23 O (wt.%) Fe23 O (wt.%) TiO2 (wt.%) 12 0 45 50 55 60 65 70 75 45 50 55 60 65 70 75 45 50 55 60 65 70 75

SiO2 (wt.%) SiO2 (wt.%) SiO2 (wt.%) 12 CaO (wt.%) 7 Na O (wt.%) 40 Zr/Nb D 2 E F 10 6 35 30 8 5 25 6 4 20 4 3 15

2 2 10 5 0 1 45 50 55 60 65 70 75 45 50 55 60 65 70 75 20 40 60 80 100 120 140 160 180

SiO2 (wt.%) SiO2 (wt.%) Zr (ppm) 2.5 14 Th/Nb Nb/Yb 2.2 (Dy/Yb) G H N 12 I 2.0 10 1.8

1.5 8 1.4 1.0 6 4 1.0 0.5 2

0.0 0 0.6 024 6810120 4 81245 50 55 60 65 70 75

Th (ppm) Th (ppm) SiO2 (wt.%) 16 1000 650 (La/Sm) Sr Ba N J K L 530 12 800

410 8 600 290

4 400 170

0 200 45 50 55 60 65 70 75 45 50 55 60 65 70 75 50 45 50 55 60 65 70 75

SiO2 (wt.%) SiO2 (wt.%) SiO2 (wt.%) Langxian granodiorite Mamen adakitic rock (137 Ma) Langxian mafic enclave Lilong adakite-like charnockite (86-90 Ma) Langxian two-mica granite this study Lilong mafic rock (88-98 Ma) Lamprophyre dyke Post-collisional adakite-like rock (30-10 Ma) Nuri quartz-diorite

Fig. 5. (A–E) Harker variation diagrams showing the major element variations, (F) Zr/Nb vs. Zr contents, (G–H) Th/Nb and Nb/Yb vs. Th contents, (I–L) (Dy/Yb)N,(La/Sm)N, Sr and Ba vs.

SiO2 contents of rocks from the Langxian and the Nuri granitoids, southern Tibet. Chondrite normalizing values are from Sun and McDonough (1989). Data are from the same source as Fig. 4.

4.2.1. Langxian granodiorite chambers by magma mixing (Didier, 1987; Holden et al., 1987; Yang We discuss the origin of the Langxian mafic enclaves first because it et al., 2004; Zheng et al., 2012b), or primitive early stage magmas of is essential to understanding the petrogenesis of the entire Langxian the host granite (Li et al., 2010). In the restite model, mafic enclaves igneous suite. Mafic enclaves within intermediate to felsic igneous are commonly thought to represent residual material that progressively rocks worldwide are generally interpreted as residual material from unmixes from the melt during the rise of a crystal mush from its source the site of melting (restite) (Chappell and White, 1992), relics of a region (Chappell et al., 1987). However, mafic enclaves in the Langxian mafic igneous component added to intermediate to felsic magma granodiorite have typical igneous textures that are distinct from the Y. Zheng et al. / Lithos 190–191 (2014) 240–263 251

250 Langxian granodiorite MORB Langxian mafic enclave this Adakite or high-Al TTD Langxian two-mica granite study 100 25% Garnet- Lower-Al TTD 200 Lamprophyre dyke amphibolite Partial melting lines Nuri quartz-diorite Mamen adakitic rock (137 Ma) Lilong adakite-like 10% Garnet- 150 N amphibolite Adakite charnockite (86-90 Ma) Lilong mafic rock (88-98 Ma) 10

Sr/Y Post-collisional adakite-like

(La/Yb) 50 100 rock (30-10 Ma) 10

Amphibolite 50 25 10 Arc magmatic rocks 50 A 50 MORB B 0 0 50 0 5 10 15 20 25 30 35 40 45 50 0 5 10 15 20

Y (ppm) YbN 80 7 Adakites related Mantle Adakites related to slab melting melts 6 to slab melting 60 5 Crustal AFC 4 40

3 Mg# MgO (wt.%) MgO 2 20

1 Adakite-like rocks related Adakite-like rocks related to lower crust melting to lower crust melting CC DD 0 0 45 50 55 60 65 70 75 45 50 55 60 65 70 75

SiO2 (wt.%) SiO2 (wt.%) 150 1.2

Adakites related to slab melting 1.0 Post-collisional adakite-like rocks in southern Tibet 100 0.8 m)

Ce 0.6 Th/ Ni ( pp Ni Adakite-like rocks related 50 0.4 to lower crust melting

0.2 Cenozoic adakites related to ECE slab melting (arc setting) FF 0 0.0 10 20 30 40 50 60 70 80 0 102030405060 Mg# Th (ppm) 1400 1000 fluid-related enrichment Sediment melt

1200 Post-collisional adakite-like 100 rocks in southern Tibet 1000

10 800

600 Th /Yb

Ba (ppm) 1

400

0.1 Tethyan 200 Melt-related enrichment GG HF 0 0.01 0 0.5 1 1.5 2 2.5 0.01 0.1 1 10 100 Nb/Y Th/Sm

Fig. 6. (A)Sr/Yvs.Yand(B)(La/Yb)N vs. Ybn discrimination diagrams showing data for adakites and normal calc–alkaline rocks (Defant and Drummond, 1990) (C) MgO contents and (D)

Mg# vs. SiO2 contents, (E) Ni contents vs. Mg#, (F) Th/Ce vs. Th, (G) Ba vs. Nb/Y, and (H) Th/Yb vs. Th/Sm diagrams for rocks from the Langxian and Nuri granitoids, southern Tibet. The mantle AFC curves are after Stern and Kilian (1996).Thefield of adakites related to slab melting and lower crustal melting are from Condie (2005).Thefield of Cenozoic slab-derived adakities (arc setting) and Tethyan basalt are from Wang et al. (2008) and Zhu et al. (2009a), respectively. Chondrite normalizing values are from Sun and McDonough (1989).Data are from the same source as Fig. 4. 252 Y. Zheng et al. / Lithos 190–191 (2014) 240–263

200 1000 Nuri quartz-diorite 100 Lamprophyre dyke tl e 100

te Lilong adakite-like Lamprophyre charnockite (86-90Ma) dyke Langxian mafic enclave ve man Lilong mafic rock

ho ndri (88-98Ma)

rimiti 10

10 P le s/ C Langxian p granodiorite Langxian am mafic enclave S Langxian 1 two-mica granite Samples/ AA BAB 1 .2 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Rb BaTh U Nb Ta K LaCePbNd Sr SmHf Zr Ti Eu GdTb Y Yb Lu 1000 1000 Post-collisional adakite-like rocks (30-10Ma) Post-collisional adakite-like rocks (30-10Ma) le t 100 100 Mamen adakitic rock (137Ma) Mamen adakitic rock Langxian (137Ma) mafic enclave ve m antle imiti ve 10 10 s/Primitive man s/Primitive s/P r e Langxian granodiorite Nuri quartz-diorite 1 1 Sampl e Sampl Langxian two-mica granite CAC DABD .2 .2 Rb BaTh U Nb Ta K LaCePbNd Sr SmHf Zr Ti Eu GdTb Y Yb Lu Rb BaTh U Nb Ta K LaCePbNd Sr SmHf Zr Ti Eu GdTb Y Yb Lu

Fig. 7. (A) Chondrite-normalized REE patterns, and (B–D) Primitive-mantle-normalized trace element patterns for the Langxian and the Nuri granitoids, southern Tibet. Chondrite and primitive mantle normalizing values are from Sun and McDonough (1989). Data are from the same source as Fig. 4.

metamorphic textures expected for restite enclaves (cf. Eichelberger, of MgO (Fig. 6C), Fe2O3,Al2O3,TiO2,andCaO(Fig. 5A–D) decrease with 1980; Vernon, 1984). Moreover, U–Pb zircon ages of restite enclaves increasing SiO2 content, while K2O(Fig. 4A) and Na2O(Fig. 5E) increase. are commonly much older than the age of crystallization of their host The Langxian mafic enclaves and host granodiorites also share similari- granites (Chappell et al., 1987), whereas the Langxian enclaves and ties in terms of their whole-rock Sr–Nd–O isotopic compositions, host granodiorite in this study have nearly identical (i.e., overlap within indicating a close genetic relationship. Therefore, on the basis of all of error) U–Pb zircon ages of 106.4 ± 2.6 Ma and 103.1 ± 1.8 Ma, respec- these observed geochemical and isotopic relationships, it seems reason- tively. Thus, it is unlikely that these Langxian mafic enclaves are restitic able to suggest that these mafic enclaves and host granodiorites could in origin. have been genetically linked through protracted igneous fractionation In recent studies, magma mixing has been the mechanism that is processes. Certainly, this fractional crystallization model seems to be most frequently invoked to account for the presence of mafic enclaves consistent with the trace element characteristics of the mafic enclaves within intermediate to felsic igneous intrusions (Guan et al., 2012; and granodiorite. Fractional crystallization of basaltic magmas will Yang et al., 2004; Zheng et al., 2012b). Although linear geochemical result in continuously evolving trends in the abundances and ratios of trends exhibited by pairs of elements on variation diagrams (e.g., incompatible trace elements for the associated basaltic and felsic rocks Fig. 5) could result from mixing between two magmas, the absence of (Fig. 5F–H) (Li et al., 2010; Peccerillo et al., 2003). acicular apatites, quenched margins, and K-feldspar and quartz Major element contents determined on the mafic enclaves, megacrysts within the mafic enclaves from this study, indicates that a lamprophyre and granodiorites in this study collectively define a nor- magma mixing model is not a very suitable mechanism to explain the mal fractionation trend that indicates a common fractional process presence and evolution of enclaves of the Langxian suite (Guan et al., characterized by the sequential appearance of olivine, clinopyroxene, 2012; Yang et al., 2004). This deduction is also supported by the and hornblende fractional crystallization (Chappell, 1996; Richards whole-rock Sr–Nd–O and zircon Hf isotopic data presented here and Kerrich, 2007). Although clinopyroxene and hornblende are both (Fig. 8), because according to the magma mixing model, the host known to preferentially incorporate HREEs over LREEs, hornblende is granites and entrained enclaves commonly exhibit large differences in more effective in this process (Bachman et al., 2005; Prowatke and isotopic compositions (Yang et al., 2004; Zheng et al., 2012b), which is Klemme, 2006; Richards and Kerrich, 2007). Thus, the fractionation of something that we do not observe here between the Langxian enclaves hornblende from original magmas–in the absence of plagioclase as a and host granodiorite (Fig. 8). crystallizing phase–should yield increasing LREE/HREE ratios and Final possibility is that the Langxian enclaves represent relict blebs of progressively chondrite-normalized REE patterns, systematically, the primary melts present at the earliest stages of magmatic evolution during sequential magma evolution and emplacement of these igneous of their host granodiorite. To some degree, the mafic enclave and host rocks—similar in nature to the geochemical relationships observed for granodiorite samples exhibit the same differentiation trends on Harker typical adakites reported by Defant and Drummond (1990). However, diagrams (Figs. 4–6). For example, in both rock types the concentrations in high-pressure environments, garnet will appear as a crystallizing Y. Zheng et al. / Lithos 190–191 (2014) 240–263 253

Langxian granodiorite 12 Tethyan basalt Langxian mafic enclave Langxian two-mica granite S-type granites 8 (150 Ma) Langxian two-mica granite this Inherited zircon in Lamprophyre dyke study Langxian two-mica granite Nuri quartz-diorite Post-collisionall adakite-like 351 Ma Mamen adakitic Lilong mafic rock (88-98Ma) Ancient rock (30-10Ma) Post-collisional adakite-like rock 10 4 90 rock (137 Ma) lower crust (30-10Ma) Hf /Hf = 1.5 pm 90 c 0.7 30 0 0.2 90 Nd(t) 8 0.05 ε 70 O (‰) 50 18

Post-collisional adakite-like δ 112 Ma -4 rock in southern Tibet 70 50 70 103 Ma Indian Ocean Lower crust 30 6 90 pelagic clay 89 Ma -8 50 10 Upper crust Enriched Depleted 70 30 Mantle zircon A mantle mantle -12 4 0.702 0.704 0.706 0.708 0.710 0.712 0.714 0.716 0.718 -15 -10 -5 0 5 10 15 87Sr/86Sr(i) εHf(t) 20 Depleted mantle Fig. 9. Plot of εHf(t) isotope versus O isotopes of zircons from the Langxian two-mica gran- Dep ite, southern Tibet. The dotted lines denote the two-component mixing trends between leted mantle 0.5Ga Ga the mantle- and supercrust-derived magmas, while the solid lines denote the two- 0.5 Ga 10 1.0 component mixing trends between the mantle- and lower-crust-derived magmas. Mamen a Hf /Hf is the ratio of Hf concentration in the parental mantle magma (pm) over crustal adakitic pm c .0G rock 2 (c) melt indicated for each curves, and ticks on the curves represent 10% mixing incre- 18 1.0 Ga ments by assuming the mantle zircon has εHf = 12 and δ O = 5.3‰; the supercrustal

Nf(t) Chondrite ε − δ18 ‰

ε zircon has = 12 and O=10 (Li et al., 2009, 2010); and the lower crust zircon 0 Hf Hf=0.015 has ε = −5.2 and δ18O = 9.4‰ (Zheng et al., 2012a). The ratio of Hf concentrations in Cretaceous-Paleocene Hf Gangdese the parental magma (pm) and crustal (c) end members (Hfpm/Hfc) is indicated for each. 6Lu/177 The post-collisional adakite-like rocks in southern Tibetan is from Zheng et al. (2012a). 17

-10

2.0Ga phase, and therefore fractionation of garnet may also yield REE distribu- B tion patterns that are similar in nature to those observed for typical 0 20 40 60 80 100 120 140 500 1000 1500 2000 2500 adakites (Macpherson et al., 2006). Compared with the strong affinity U-Pb age (Ma) of garnet for the HREEs, hornblende has only a moderate affinity for Cretaceous MREEs. Consequently, hornblende fractionation should in principle 20 Gangdese yield increasing (La/Sm)N ratios, but steady or decreasing (Dy/Yb)N 99 MORB 99 ratios (i.e., listric patterns; Fig. 7A), whereas garnet fractionation 98 98 10 96 96 should in principle yield both increasing (La/Sm)N ratios and increasing 99 90 (Dy/Yb)N ratios. As shown in Fig. 5, (La/Sm)N increases from ~2 in the 80 90 70 98OIB+AV Mamen mafic enclaves to ~6 in the granodiorite (Fig. 5J), whereas (Dy/Yb) adakitic N

Nf(t) 0

ε 96 rock shows a steady value at ~1.4 throughout the fractionation range 80 Ocean 70 pelagic clay (Fig. 5I). These geochemical trends are inconsistent with significant 60 -10 50 90 involvement of garnet fractionation in the magmatic history of these Post-collisional adakite-like 80 rock in southern Tibet rocks (cf. Macpherson et al., 2006), but are instead consistent with 70 60 hornblende fractionation during magma evolution from a basaltic -20 Tibetan Tibetan Lower Crust Upper Crust C andesite composition through to a dacite (Chappell, 1996; Richards -25 -20 -15 -10 -5 0 5 10 15 and Kerrich, 2007). Therefore, on the basis of all of this geochemical ε Nd(t) reasoning, we suggest that the Langxian adakite-like granodiorite were generated by low-pressure fractional crystallization from the ε 87 86 ε – Fig. 8. (A) Whole-rock Nd(t) vs. initial Sr/ Sr isotopic compositions, (B) Hf(t) vs. U Pb parental basaltic magmas, which are now preserved as relict mafic ε ε ages of zircons, and (C) Hf(t) vs. Nd(t) isotopic compositions diagrams of the Langxian enclaves found in the granodiorite. The high Sr contents of the mafic and Nuri granitoids, southern Tibet. The field of the Tethyan basalt is from Mahoney enclaves and Langxian granodiorite samples rule out the possibility of et al. (1998). The field of post-collisional adakite-like rocks in southern Tibetan is from fi Hou et al. (2004), Guo et al. (2007), Gao et al. (2007, 2010), Xu et al. (2010),andZheng any signi cant plagioclase fractionation in their magmatic history, et al. (2012a, b).Thefield of the Cretaceous–Paleocene Gangdese is from Chu et al. which is also supported by the absence of any discernible Eu anomalies (2011). Data for the Mamen adakite and the Late Cretaceous Lilong mafic rocks are from in any of these rocks (Chappell, 1996; Richards and Kerrich, 2007). ε Zhu et al. (2009a) and Guan et al. (2011), respectively. Zircon Hf(t) isotopic data for the As outlined above, various petrogenetic models have previously Nuri quartz–diorite are from Chen (2011). Whole-rock εHf(t) and εNd(t) isotopic data for the Cretaceous Gangdese batholiths are from Chu et al. (2011). The compositions of been proposed for the origin of adakite-like rocks; therefore, to address end-members or components assumed for mixing calculations are: (1) The old lower the origin of adakite-like rocks in this study, we will focus on alternative ε − crust in Tibet, used in the modeling, has Hf = 14, Hf = 1.8 ppm (Vervoort et al., models arising from the specific characteristics of the Langxian adakite- ε − 87 86 1999), Nd = 22, Nd = 26 ppm, Sr/ Sr(i) = 0.7100, Sr = 300 ppm (Miller et al., like rocks described in this paper. Collectively, several geochemical 1999). (2) The upper crust in Tibet has εHf = −14, Hf = 1.8 ppm (Vervoort et al., 1999), 87 86 features of the Langxian granodiorite suite can be used to rule out the εNd = −9.24, Nd = 33.7 ppm, Sr/ Sr(i) = 0.7326, Sr = 100 ppm (Miller et al., 1999). possibility that they were derived from partial melting of either a (3) The Tethyan basalt or Inidan MORB has εHf = 16.55, Hf = 0.199 ppm (Griffinetal., 87 86 2000), εNd = 8.9, Nd = 6.66 ppm, Sr/ Sr(i) = 0.7035, Sr = 187.0 ppm (Zhang et al., subducted oceanic slab or a foundering region of lower crust. If such ε 2005). (4) Indian Ocean pelagic sediment has Hf = 0.4, Hf = 5.73 ppm (Chu et al., melts were to form in a subducting slab or foundering region of lower ε − 87 86 2011), Nd = 9.3, Nd = 23.05 ppm, Sr/ Sr(i) = 0.7168, Sr = 119 ppm (Ben Othman crust, they would likely increase their Mg# and compatible element et al., 1989). contents (eg. Cr, Ni) significantly through interactions with peridotite as they rise upward through the overlying mantle wedge (Stern and Kilian, 1996; Wang et al., 2007; Zhu et al., 2009a),andthisisnot 254 Y. Zheng et al. / Lithos 190–191 (2014) 240–263

consistent with the observations of Mg# and compatible element high above these lower crustal adakite-like rocks’ on the SiO2 versus contents documented here (Fig. 6A–E). The post-collisional adakites in Mg# diagram (Fig. 6D; Fig. 8A of Condie, 2005), but higher than the southern Tibet are generally regarded as having been derived from the geochemical fields for coeval mafic rocks on both types of diagrams Tibetan lower crust (Chung et al., 2003; Guo et al., 2007; Hou et al., (Guan et al., 2011; Zhang et al., 2010b). As shown in Fig. 7D, the post- 2004; Zheng et al., 2012a, b). Although low MgO and compatible collisional lower crustal-derived adakite-like rocks generally tend to be element contents of the lower-crust-derived post-collisional adakite- K-rich and characterized by high contents of strongly incompatible like rocks are comparable to rocks from the Langxian granodiorite elements such as Rb, Ba, Th, and U. In the present case, however, those

(Fig. 6), the K2O, LILE, and Pb contents, and Th/Ce, Th/Yb, and Th/Sm two geochemical characteristics were not observed in the geochemistry ratios of the post-collisional adakite-like rocks are considerably higher of Nuri adakite-like rocks analyzed in this study. Moreover, compared than in the Langxian granodiorite (Chung et al., 2003, 2009; Hou et al., with the post-collisional adakite-like rocks, the Nuri adakite-like rocks 2004; Qu et al., 2004; Zheng et al., 2012a, b). Meanwhile, the Langxian have substantially lower Th contents and Th/Ce ratios (Fig. 6F), similar granodiorite are characterized by metaluminous, while the post- to Cenozoic slab-derived adakites in arc settings (Wang et al., 2008). collisional adakite-like rocks are dominated by peraluminous. Addition- Therefore, on the basis of all of these observed geochemical discrepan- ally, the whole-rock Sr–Nd and zircon Hf isotopic compositions of the cies, we suggest that the Nuri igneous suite was not likely derived from post-collisional adakite-like rocks are also distinct from the Langxian the southern Tibetan lower-crust. This idea is also supported by our granodiorite (Fig. 8). Thus, unlike the post-collisional adakite-like isotopic data, which indicate that both whole rock Sr–Nd and zircon Hf rocks, these geochemical and isotopic features indicate that the isotopes of the Nuri adakite-like rocks are significantly different from Langxian granodiorite was probably not generated by partial melting the previously studied post-collisional adakite-like rocks (Fig. 8). of the southern Tibetan lower crust. The Nuri quartz–diorites generally show different differentiation trends (Fig. 5)thanthemafic enclaves, granodiorite, and two-mica 4.2.2. Langxian two-mica granite granite from the Langxian suite. In addition, the Nuri rocks have higher Although the rocks from the Langxian two-mica granites, the mafic Mg# and compatible element abundances than mafic enclaves of the enclaves and the host granodiorite have similar differentiation trends Langxian suite. These geochemical features combined, indicate that in terms of their concentrations of MgO, Fe2O3,TiO2,Na2O, and CaO the Nuri adakite-like rocks were also not likely to have been derived (Fig. 6A–E), the trends in concentration of Al2O3 exhibit differences, from coeval mafic melts by fractional crystallization processes. Thus, which have also been observed in the trace element diagrams (e.g., Th the only remaining candidate geological source region for Nuri vs. Nb/Yb, SiO2 vs. (Dy/Yb)N, and SiO2 vs. (La/Sm)N). The concentrations adakite-like magmas is the subducted Neo-Tethyan oceanic slab. This of Al2O3 in the two-mica granite samples are distinctively higher than inference is supported by the observation that the whole-rock geo- those in the granodiorite. Meanwhile, the two-mica granites are chemical and zircon Hf isotopic compositions of the Nuri adakite-like dominantly peraluminous, while the granodiorites are characteristically rocks are comparable to that of the Cretaceous (137 Ma) Mamen metaluminous. All these features indicate that the two-mica granites adakitic rocks from the same area, which are generally interpreted as cannot be the simple result of fractional crystallization of mafic enclaves being the products of partial melting of the subducted Neo-Tethyan in the granodiorite, and this is also supported by our new U–Pb zircon oceanic slab (Zhu et al., 2009a). Both the Nuri and Mamen adakite-like data for the two-mica granites, already discussed in the U–Pb age rocks are characterized by high Mg# and high compatible element section. Additionally, the lack of mafic enclaves may also be used to contents, high εNd(t) and elevated zircon εHf(t) values, low K2O, LILE, 87 86 rule out the possibility of low-pressure or high-pressure fractional and Pb contents, and low Th/Ce, Th/Yb, Th/Sm, and Sr/ Sr(i) ratios crystallization of basaltic melts for the origin of the adakite-like two- (Figs. 4–8). Collectively, these geochemical and isotopic similarities mica granite. indicate that–like the Mamen adakitic rocks–the Nuri quartz–diorite As shown in Fig. 6, the two-mica granites have significantly lower suite was probably derived from partial melting of the subducted Neo- values of Mg#, MgO, and compatible element contents when compared Tethyan oceanic slab. with the Langxian granodiorite. As deduced above, therefore, it is not possible that melting of subducted Neo-Tethyan oceanic crust or a 4.3. Nature of the magma source regions foundered lower crust of southern Tibet generated the Langxian two- mica granite. The pronounced negative Ti anomalies and positive K As with more typical subduction-related tectonic settings (Elburg and Pb anomalies that are apparent in primitive mantle-normalized et al., 2002; Guo et al., 2005; Kay et al., 1987), to varying degrees trace element plots of the Langxian two-granites are very similar to the subducted oceanic crust (mid-ocean ridge basalt (MORB) + those of lower-crust-derived post-collisional adakite-like rocks, but sediment + slab-derived fluids), the mantle wedge, and the overlying distinct from those of the Langxian granodiorites and the ocean-crust- crust may all have contributed as source regions in the generation of derived Mamen adakite (Fig. 7C), indicating that the two-mica granites Langxian and Nuri suite magmas. are possibly derived from melting of the southern Tibetan maficlower crust. Post-collisional adakite-like rocks commonly have relatively low 4.3.1. Langxian granodiorite values of Mg#, and low MgO, Cr and Ni contents, but high Al2O3 The enclaves and the lamprophyre dyke sampled in the Langxian contents, Th/Yb and Th/Sm ratios, and large numbers of ancient suite exhibit similar major element fractionation trends as coeval inherited zircons (Chung et al., 2003; Gao et al., 2007, 2010; Hou et al., mafic intrusions (88–98 Ma) from the same area (Figs. 4–6). All these 2004; Mo et al., 2005; Xu et al., 2010; Zheng et al., 2012a). These rocks have very similar trace element contents and distribution patterns features are all observed in the two-mica granites described here, (Figs. 5–7), noticeably strong enrichments of LILEs relative to HFSEs, and further supporting our hypothesis. Relatively low (La/Yb)N and pronounced negative Nb–Ta anomalies and positive K and Pb anomalies (Dy/Yb)N values indicate that the two-mica granite was derived from on primitive mantle-normalized trace element plots (Fig. 7B–C), the mafic lower crust in the absence of garnet. indicating that all these igneous rocks probably share a similar source region. This conclusion is further supported by the fact that these differ- 4.2.3. The Nuri quartz–diorite ent intrusive rocks are similar in terms of their whole-rock Sr–Nd–Oand The geochemical fields defined by Nuri adakite-like rocks on zircon Hf isotopic compositions (Fig. 8), all of which may also reflect the diagrams of Mg# versus compatible element abundances plot consider- composition of the source for these magmas. ably higher than fields for the Langxian granodiorite and two-mica The low SiO2 contents of 47–50 wt.% in whole-rock samples of the granite, and the lower-crust derived post-collisional adakite-like rocks Langxian mafic enclaves suggest that their precursor could be basaltic. in southern Tibet (Fig. 6E), and by comparison the Nuri field plots ‘less However, the relatively low MgO (b6.3 wt.%), Ni (21–31 ppm) and Cr Y. Zheng et al. / Lithos 190–191 (2014) 240–263 255

(10–65 ppm) concentrations demonstrate that the maficenclavesand and 3). However, rocks from the two-mica granite suite also have lamprophyre dyke do not represent primary mantle derived melts, but lower whole-rock εNd(t) and in-situ zircon εHf(t) values, and higher 87 86 are significantly evolved. Primitive arc are commonly character- whole-rock Sr/ Sr(i) values than those of juvenile melts represented ized by MgO N8 wt.%, Ni N200 ppm, and Cr N400 ppm (Tatsumi and by the Langxian mafic enclaves, indicating the involvement of ancient Eggins, 1995). lower crustal materials in the genesis of the two-mica granite. On the Oceanic basalts [i.e. mid-ocean ridge basalts (MORB) and ocean basis of previous studies, it has been suggested that the Lhasa terrane island basalts (OIB)] typically exhibit positive Nb–Ta–Ti anomalies and has an ancient Proterozoic to Archean basement (Zheng et al., 2012a; negative Pb anomalies in primitive mantle-normalized trace element Zhu et al., 2011), which has also been taken as the ancient crustal diagrams (e.g., Hofmann et al., 1986). However, the maficenclavesare end-member for the two-mica granite here. This suggests, therefore, characterized by significantly negative Nb–Ta–Ti anomalies and positive that most zircons in the two-mica granite precipitated from melts that Pb anomalies (Fig. 6B), similar to those of arc magmas (Rudnick and contained less than 30% ancient crustal materials (Fig. 9), which is Fountain, 1995), but distinctly different from those of mid-ocean ridge significantly higher than the amount estimated using whole-rock Sr– and oceanic island basalts (Hofmann et al., 1986), indicating that the Nd isotopes (Fig. 8). 18 18 mafic enclaves are unlikely to have originated from normal MORB or As shown by Valley (2003), the values of whole-rock δ O(δ OWR) 18 18 OIB source mantle material, but from a metasomatized mantle wedge. in equilibrium with zircon δ O(δ OZRN) can be calculated. This calcu- The mantle wedge as the source region of subduction-related lation has been carried out according to the relation given by Valley et al. 18 18 magmas can often become metasomatized and geochemically enriched (1994) where δ OWR = 0.06(wt.% SiO2) – 2.25 + δ OZRN. If we as- 18 during interactions with slab-derived fluids and the partial melts of sume zircons crystallized from a melt with 72 wt.% SiO2, the δ OZRN subducted sediments (Elburg et al., 2002; Guo et al., 2005). The latter values of the Langxian two-mica granite are similar to the calculated 18 commonly have high abundances of Th and LREE, whereas the former δ OWR values (ranging from 8.1 to 9.1). Thus it appears that the two- are characterized by high concentrations of Ba, Rb, Sr, U, and Pb (Guo mica granite was not significantly contaminated by wallrocks after et al., 2007; Hawkesworth et al., 1997). Rocks from the Langxian mafic crystallization of zircon, nor was the source region for the two-mica enclaves exhibit variable Ba concentrations coupled with a narrow granite significantly altered by low-temperature fluids (Zheng et al., range of Nb/Y ratios, which is consistent with fluid-related enrichment 2011a, b). Therefore, the Langxian two-mica granite originated in a (Fig. 6G). Present-day fluid-dominated arc settings show Th/Yb ratios rejuvenated mafic lower crust below southern Tibet, which was not of b1, whereas arc environments in which significant amounts of significantly altered by regional fluids. sediments are subducted, typically exhibit Th/Yb ratios of N2(Nebel A rapid drop in Sr content with increasing SiO2,observedinthetwo- et al., 2007; Woodhead et al., 2001). The Langxian mafic enclaves mica granites (Fig. 5K), could be an indication of significant fractional commonly have Th/Yb ratios of b1, indicating limited contributions crystallization of plagioclase in the magmatic history of these two- from subducted sediment to the sources of the Langxian mafic enclaves. mica granites (Chappell, 1996). A correspondingly abrupt drop in Ba

Both the lamprophyre dyke and the granodiorite have lower whole- abundance with increasing SiO2 content (Fig. 5L) may have resulted 87 86 rock εNd(t) and zircon εHf(t) values, and higher whole-rock Sr/ Sr(i) from the appearance of biotite as a crystallizing phase (Chappell and and δ18O values than the mafic enclaves, indicating that relatively White, 1992). These features indicate that the two-mica granite ancient Tibetan lower crustal and/or upper crustal source materials underwent further significant AFC (assimilation, fractional, and crystal- may have played important roles in the generation of the Langxian lization) processes during magma ascent and emplacement. This suites. However, two relatively young inherited zircons (206Pb/238U suggestion is supported by the fact that there is a large population of ages of 123 Ma and 126 Ma) with very positive zircon εHf(t) values, inherited zircons in the two-mica granite. together with the lack of ancient inherited zircons in both the lamprophyre and the granodiorite suggest that only the Tibetan lower 4.3.3. The Nuri quartz–diorite crust has been involved, without significant involvement of upper As discussed above, the Nuri quartz–diorite is most likely to have crustal materials. This means that these mantle-derived mafic melts been derived from partial melting of the subducted Neo-Tethyan ocean- seem to have stalled at the base of the crust, presumably due to density ic slab. The Nuri adakitic rocks, that we have investigated, have higher filtering, and have undergone MASH (melting, assimilation, storage, and Mg#, and MgO, Cr, and Ni contents than the typical adakites described homogenization, Hildreth and Moorbath, 1988) processes at the base of by Defant and Drummond (1990), and they are also enriched in both the lower crust. Zr and HREEs. These geochemical differences have been ascribed to the interaction of hydrous slab-derived melts with the overlying 4.3.2. Langxian two-mica granite mantle wedge (Martin et al., 2005; Stern and Kilian, 1996), including The inherited zircon U–Pb ages at 351 Ma and from 112 to 89 Ma slab-melt/peridotite reactions and hybridization processes (Keleman, coincide with periods of intense igneous activity in the Lhasa 1990, 1995; Keleman et al., 1993), and the triggering of both diapiric block (Ji et al., 2012; Mo et al., 2005; Zhu et al., 2011). The elevated uprise and decompression melting of the mantle wedge that ultimately oxygen isotope and negative εHf(t) values for the inherited zircons results in the mixing of slab- and mantle-derived melts (Yogodzinski aged 351 Ma indicate that they were probably derived from crust- et al., 1995). As outlined above, the Nuri adakite-like rocks have higher dominated felsic magmas or crustal sediments (Zheng et al., 2012a). Mg#, and Cr and Ni contents than other synchronous Lilong maficrocks

However, the low oxygen isotope and elevated εHf(t) values for the (Fig. 6D, E), and thus mixing between the two types of magma would inherited zircons with ages in the range 112 to 89 Ma indicate that not have produced an increase in the Mg#, and Cr and Ni contents of they are likely to have been derived from a source region that contains the Nuri adakite-like rocks. The mixing of slab- and mantle-derived more depleted mantle materials than zircons of the two-mica granite. melts is therefore not applicable in the case of the Nuri adakite-like As deduced above, the Langxian two-mica granite is derived from rocks, and instead we propose a slab-melt/peridotite reaction model partial melting of the southern Tibetan mafic lower crust. The relatively to explain the geochemical features of the Nuri quartz–diorite. Hydrous positive whole-rock εNd(t) and in-situ zircon εHf(t) values, low whole- slab-derived melts that are ascending through the mantle wedge would 87 86 18 rock Sr/ Sr(i) and in-situ zircon δ Ovaluesindicatethatlarge provide substantial amounts of thermal energy for peridotite dissolu- volumes of juvenile materials were present in the southern Tibetan tion. In turn, the dissolution of the peridotite has the potential to signif- mafic lower crust (N70%, N80%, and N96% of depleted mantle material icantly modify the SiO2, MgO, Ni, and Cr contents in the hybridized slab asdeducedbyzirconHf–O, Sr–Nd, and Nd–Hf isotopic modeling, melts, and increase their trace element abundances. The process respectively; Fig. 8A, C). Similarly, the whole-rock Nd and zircon Hf involves the reaction of mantle olivine and pyroxene grains with crustal model age spectra yield relatively young model ages (Tables 1 relatively silica-rich hydrous melts to produce new minerals (including 256 Y. Zheng et al. / Lithos 190–191 (2014) 240–263 amphiboles, pyroxenes, olivine, and/or garnet), and the formation of Much like the Gangdese Belt magmatism, the development and hybridized melts (Carroll and Wyllie, 1989; Keleman, 1990, 1995; evolution of the Neo-Tethyan Ocean took place over a very long period Sekine and Wyllie, 1983). of geological time (Middle Jurassic or Early Cretaceous to the Eocene) Rocks from the Nuri quartz–diorite have Th/Yb ratios of N3and (e.g., Searle et al., 2007; Zhang et al., 2005; Zhu et al., 2011), and thus variable Ba concentrations coupled with a narrow range of Nb/Y ratios the aforementioned ‘subduction of young oceanic crust’ model cannot (Fig. 6G–H), indicating contributions of enriched materials from both readily be applied to the formation of the Nuri slab-derived adakitic slab-derived fluids and subducted sediments. This conclusion is also rocks. The main active stage of Gangdese magmatism is recorded by supported by the fact that the Nuri quartz–diorite has lower whole- the presence of HT charnockitic magmatism (98–88 Ma adakite-like 87 86 rock εNd(t) values and higher whole-rock Sr/ Sr(i) values than Tethy- charnockites; Zhang et al., 2010b, 2010c) and HT metamorphism an basalt (150 Ma). Additionally, rocks from the Nuri suite have (90–85 Ma HT granulite-facies metamorphism, T = 750–850 °C; significantly higher whole-rock δ18O values than in fresh oceanic crust P = 1.0 GPa; Wang et al., 2009), both of which indicate a significantly (Aggarwal and Longstaffe, 1987), indicating that these rocks were either high heat flux in the Gangdese Belt during the Late Cretaceous derived from low-temperature altered oceanic crust (Zheng et al., (ca. 105–80 Ma). This means that the coeval Nuri adakite-like suite 2011a, b) or contain large amounts of recycled low-temperature altered (95 Ma) was probably also developed under conditions of extremely young crust (Lackey et al., 2005, 2006, 2008; Nelson et al., 2013). Since high heat flow. This high heat flux could be explained as a result of we have been unable to identify any such altered young crust, the either oceanic slab break-off or subduction of a mid-ocean ridge system former hypothesis seems more likely. (DeLong et al., 1979; Iwamori, 2000), because in both situations hot asthenosphere will ascend upwards through the slab window and 5. Implications come into direct contact with the crust of the subducting oceanic slab. Also, in both of these scenarios, high heat flow through a slab window 5.1. Implications for mid-ocean ridge subduction can induce partial melting of the basaltic oceanic crust along the edges of the slab window (Stern and Kilian, 1996; Yogodzinski et al., 2001). The Gangdese Plutonic Belt is generally regarded as a typical conti- However, we favor the ‘subducting mid-ocean ridge’ model over the nental magmatic arc, and it is one of the most prominent features of other because northwards subduction of the Neo-Tethyan oceanic slab the southern Tibetan Plateau (Allègre and Rousseau, 1984; Coulon under the Lhasa terrane initially began sometime in the Middle Jurassic et al., 1986; Pan et al., 2006; Yin and Harrison, 2000). Although new ev- or Early Cretaceous (Chu et al., 2006; Searle et al., 2007; Zhu et al., 2011), idence indicates that the development and evolution of Gangdese and was then terminated at ~55 Ma when the Indian plate collided with magmatism took place over a long, protracted period of geological the Lhasa terrane (Leech et al., 2005; Mo et al., 2005). Therefore, it is time (from ca. 210 to 40 Ma), the belt is marked by two peak episodes likely that the possible subduction of a mid-ocean ridge system may of activity at 109–80 Ma and 55–45 Ma (Chu et al., 2006; Mo et al., have transpired at about the mid-point (i.e., ca. 105 ± 10 Ma) in the 2005, 2007; Zhu et al., 2009b, 2011). These two epochs of granitoid subduction history of the Neo-Tethyan oceanic plate, which would plutonism are generally ascribed to the normal northwards subduction correlate well with the timing of magmatic activity associated with of the Neo-Tethyan oceanic plate (for the older plutonic stage) (Chu emplacement of the Nuri quartz–diorite (Fig. 10). et al., 2006, 2011; Honegger et al., 1982; Ji et al., 2009; Searle et al., Recently, some authors have suggested that the adakitic rocks in the 2007), and then the break-off of the Neo-Tethyan oceanic slab from Kelu (93–90 Ma) and Milin (100–89 Ma) areas, similar in age to the the subducted Indian lithosphere during the Eocene (for the younger Nuri quartz–diorite, were also derived from partial melting of the plutonic stage) (Chung et al., 2005; Gao et al., 2008; Mo et al., 2007). subducted Neo-Tethyan oceanic slab (Jiang et al., 2012; Ma et al., However, some researchers have recently suggested that some of the 2013). Thus, the lack of any age trend in the distribution of these Late Cretaceous Gangdese magmatism may have been induced by the adakitic rocks would suggest a spreading mid-ocean ridge sub-parallel subduction of a Neo-Tethyan mid-ocean ridge system (Guan et al., to the trench (Ma et al., 2013). However, Ma et al. (2013) suggested 2010; Zhang et al., 2010b, 2011; Zhu et al., 2011). Using our new U–Pb that the Late Cretaceous adakitic magmatisms in the Milin area could zircon geochronological, whole-rock geochemical, and in-situ zircon be ascribed to the upwelling of asthenosphere, triggered by the roll- Hf–O isotopic data, we have evaluated the origin and petrogenesis of back of the subducted Neo-Tethyan oceanic slab. Indeed, if the slab the Nuri and the Langxian adakite-like igneous intrusive suites. In the roll-back model is correct, magmatism in the Lhasa terrane should paragraphs that follow, we will show how this new information pro- have begun in the north and been propagated southwards, which is in- vides ways of testing the validity of the aforementioned hypotheses consistent with the facts. We suggest, therefore, that the Late Cretaceous for the origin of the Gangdese Plutonic Belt. adakitic magmatism (100–89 Ma) was controlled by the subduction of As discussed above in Section 4.2.3, the Nuri adakitic quartz–diorite the Neo-Tethyan mid-ocean ridge. is derived from the partial melting of the subducted Neo-Tethyan Except for the high heat flow through the slab window, the young oceanic slab. However, as deduced by studies in experimental petrology and hot subducted crust of the mid-ocean ridge will also dramatically (Rapp et al., 1991), the temperature of oceanic crust undergoing increase temperatures in the subduction zone (DeLong et al., 1979; subduction does not rise high enough for melting to occur before Iwamori, 2000). The amount of heat not only be capable of inducing prograde metamorphic dehydration reactions render the source rocks melting of the slab in association with low-P/high-T metamorphism, infusible under normal situations (Peacock, 1996; Richards and but would also be capable of producing extensive melting of the overly- Kerrich, 2007; Ringwood, 1977). Thus, adakitic rocks derived from ing mantle wedge and lithosphere (DeLong et al., 1979; Iwamori, 2000; basaltic oceanic crust should occur in various hot subduction settings, Tang et al., 2012a). Therefore, parts of the Late Cretaceous magmatic including regions where subduction of very young oceanic crust is tak- activity across the eastern Gangdese Plutonic Belt might also have ing place (b5Maold;Defant and Drummond, 1990; Sajona et al., 1993), been induced by the subduction of a Neo-Tethyan mid-ocean ridge in situations where the slab is abnormally heated at shallow depths— system. such as regions where the rates of subduction are anomalously slow The young and hot crust at a mid-ocean ridge has a relatively low (Peacock et al., 1994; Stern and Kilian, 1996; Yogodzinski et al., 1995; density, and thus potentially a positive buoyancy (Cloos, 1993; Delong Zhu et al., 2009a), or in situations where extraneous heat derived et al., 1979). Subduction of a buoyant mid-ocean ridge will commonly from hot asthenosphere has impinged on a slab (Martin, 1987; Wyllie, result in a reduction in the angle of subduction (Bourdon and Eissen, 1978), such as where a mid-ocean ridge is being actively subducted 2003; Pennington, 1984), and the relatively flat subducted of oceanic (Guivel et al., 1999; Kay et al., 1993) or where slab break-off is taking slab may then squeeze out the mantle wedge, terminating the place (Stern and Kilian, 1996; Yogodzinski et al., 2001). arc magmatism and inducing a contractional tectonic regime in the Y. Zheng et al. / Lithos 190–191 (2014) 240–263 257

(A) 106-86 Ma Adakitic rocks Lhasa terrane

Low P/High T Magma differentiation Crust Metamorphism and ascent ic Neo-Tethyan slab Ridge sub MASH Basaltic underplating zone Partial melting of mantle wedge duction Zone where melt-peridotite interaction antle Lithospher Ridge m Partial melting of ocean crust

Asthenosphere Asthenosphere upwelling Asthenosphere

(B) 86-65 Ma Lhasa terrane

Deep-seated intrusions Crust

Neo-Tethyan slab Partial melting of lower crust (83-76 Ma) Lithospheric Fla mantle t subduction

Asthenosphere Asthenosphere

Fig. 10. Illustrations of (A) the proposed tectonic context for the development of the Nuri and Langxian suites, and the main period of magmatic activity induced by the Neo-Tethyan mid- oceanic ridge subduction (106–86 Ma), and (B) the mantle-derived magmatic gap between 86 and 76 Ma and magmatic gap in the interval 76 to 65 Ma induced by flat subduction of the Neo-Tethyan oceanic slab, southern Tibet. arc (Bourdon and Eissen, 2003; Kay and Mpodozis, 2001). Scenarios Zhu et al., 2009b, 2011), indicating that southern Tibet may have the have also been observed in the eastern Gangdese Plutonic Belt in thickest juvenile continental crust on Earth. The addition of the juvenile which intense mantle-derived arc magmatism from 106 to 86 Ma has materials into the ancient crust has generally been ascribed to the been reported, but none at all after 86 Ma until the Eocene when the northwards subduction of the Neo-Tethyan oceanic slab. Neo-Tethyan oceanic slab break-off from the subducted Indian litho- Two major processes for generating juvenile crust have been sphere (Guan et al., 2011; Ma et al., 2013; Zhang et al., 2010a). Flat sub- recognized. The first involves horizontal accretion of an intra-oceanic duction of the Neo-Tethyan oceanic slab is also consistent with the arc system to the Lhasa terrane, which is represented by the Ladakh– crustal deformation and subsidence history of the southern Lhasa ter- Kohistan batholiths in the western Himalayan (Jagoutz, 2010; Jagoutz rane, as evidenced by the Cretaceous Takena Formation in the Penbo et al., 2011; Ravikant et al., 2009), and also by the Jurassic intra- and Maqu areas (Leier et al., 2002, 2007). Subsidence rates were con- oceanic arc system in the Gangdese batholiths (Chu et al., 2006). It has stant during the latest Jurassic and earliest Cretaceous (ca. commonly been suggested that these volcanoplutonic arcs were de- 160–105 Ma), but at the start of the Late Cretaceous, subsidence rates rived from the partial melting of a subduction-modified mantle wedge increased and continued at high levels from 105 to 90 Ma, which is in (Chu et al., 2006, 2011; Ravikant et al., 2009). The second process in- agreement with the time scale of the evolution of mantle-derived volves is vertical accretion of mantle-derived arc magmas by under- magmatism in southeastern Tibet (ca. 106–86 Ma) (Guan et al., 2011; plating and intrusion (Chung et al., 2009; Hou et al., 2013; Mo et al., Ma et al., 2013; Zhang et al., 2010a). Deposition of the Cretaceous 2007). These arc magmas are either derived from melting of the anhy- Takena Formation came to a halt at ca. 90–80 Ma, and it then drous mantle wedge caused by the subduction of the Neo-Tethyan oce- underwent folding and erosion from ca. 80 to 69 Ma, indicating a con- anic slab (Chu et al., 2006; Ravikant et al., 2009), or derived from tractional tectonic regime in the Lhasa terrane during that time. There- lithospheric mantle and/or asthenospheric mantle as a result of fore, the Neo-Tethyan mid-ocean ridge subduction may not only have rolling-back or breaking-off of the subducted Neo-Tethyan oceanic induced the intense arc magmatism that existed between 106 and slab (Chung et al., 2005, 2009; Hou et al., 2006; Mo et al., 2005; Zhu 86 Ma, but also controlled the mantle-derived magmatic gap in the in- et al., 2011). As shown in this paper, magmatism induced by the sub- terval from 86 to 76 Ma and the magmatic gap in the interval 76 to duction of a Neo-Tethyan mid-ocean ridge may be another important 65 Ma in southern Tibet (Fig. 10). way by which growth of the juvenile crust took place in southern Tibet. These magmas are derived from partial melting of the mantle 5.2. Implications for the crustal growth of southern Tibet wedge and the subducted slab. The Nuri quartz–diorite, formed by direct melting of subducted Geophysical studies indicate that the Tibetan crust beneath the oceanic crust, has a very similar geochemical composition to the aver- Gangdese belt is twice as thick as the average continental crust age andesitic continental crust (Taylor, 1977; Taylor and McLennan, (~65–80 km) (Murphy et al., 1997; Nelson et al., 1996; Priestley et al., 1995; Rudnick, 1995; Rudnick and Gao, 2003). Melting of subducted 2006). Geochemical studies indicate that the southern Tibetan crust is oceanic crust may have been the most important crust-forming process dominated by juvenile materials with an ancient Proterozoic to Archean prior to Archean (Taylor, 1967), but it rarely occurred after the Archean, crystalline basement (Chu et al., 2006; Ji et al., 2012; Zheng et al., 2012a; having only been documented in special situations (Defant and 258 Y. Zheng et al. / Lithos 190–191 (2014) 240–263

Drummond, 1990; Peacock et al., 1994; Sajona et al., 1993; Stern and like rocks in the Gangdese Belt appear to have been derived from an Kilian, 1996; Yogodzinski et al., 1995; 2001; Zhu et al., 2009a), such as anomalously thickened southern Tibetan lower crust, induced by the active subduction of a mid-ocean ridge, as reported here. All this is tectonic contraction owing to the reduction in the angle of Neo- consistent with the fact that only three sites of oceanic-slab-derived Tethyan subduction (Wen et al., 2008). However, our new observations, adakite have been documented along the Gangdese belt (Jiang et al., presented in this paper, indicate that the Cretaceous Langxian adakite- 2012; Zhu et al., 2009a), while the mantle-wedge-derived arc magmas like granodiorite and two-mica granite were produced by low- are extensive. pressure fractional crystallization of a basaltic melt and partial melting It has commonly been suggested that magmas derived from the of a mafic lower crust in the absence of garnet, indicating that the partial melting of mantle peridotite are basaltic (Baker et al., 1995; crust beneath southern Tibet may still have been normal in thickness, Tatsumi, 2006), and not andesitic like the bulk composition of the and that subduction of the Neo-Tethyan oceanic slab may still have continental crust (Rudnick and Gao, 2003; Taylor, 1967). This is consis- been at a normal angle prior to the emplacement of these intrusions tent with the observations of high-pressure experimental melting stud- at ca. 76 Ma (Fig. 10). Additionally, the La/Yb ratios in rocks derived ies of mantle peridotite that only basaltic or more magnesian magmas from the crust can be used to estimate crustal thicknesses, and most are commonly produced (Baker et al., 1994; Tatsumi, 1982; Wilson, of our new geochemical data has the La/Yb ratios at less than 41, 1991). The basaltic magmas are more mafic than the felsic magmas suggesting that the Langxian two-mica granite was generated from a that dominate the Gangdese batholiths (Guan et al., 2011; Mo et al., crust without significant thickening. (Chung et al., 2009; Guan et al., 2005; Zhang et al., 2010a, 2010b, 2010c; Zhu et al., 2011). Similar to 2012; Kay and Kay, 2002). It seems, therefore, that we can now rule the Langxian granodiorite suite reported here, these basaltic melts out the possibility of large parts of southern Tibet being elevated and should undergo MASH and/or AFC processes to produce more silicic the Tibetan crust being thickened prior to the Indian–Asian collision compositions. Meanwhile, as in the Langxian two-mica granite suite, (Kapp et al., 2005, 2007; Wen et al., 2008). Thus, crustal compression juvenile mafic magmas with a large amount of ancient crustal material and shortening as a result of the Indian–Asian collision is likely to be incorporated, will make intermediate to felsic magmas (Appleby et al., the primary cause of crustal thickening and plateau uplift in southern 2008; DePaolo, 1981; Kemp et al., 2007; Lackey et al., 2005, 2006, Tibet. The presence of magmatic epidote in several plutons of the east- 2008; Nelson et al., 2013). Both mechanisms will leave behind an ultra- ern Lhasa terrane, and also observed in our study, indicates depths of mafic lower crust, either as an olivine-rich cumulate or an eclogite, emplacement in excess of 25 km (Wen et al., 2008; Zhang et al., which will not change the bulk basaltic compositions of the juvenile 2010a), implying a regional deepening in the level of exposure over crust, although such mechanisms can effectively transform the domi- the area as a whole (Fig. 10). nantly mafic mantle-derived melts into the felsic upper crust composi- Supplementary data to this article can be found online at http://dx. tions found in the Gangdese belt. doi.org/10.1016/j.lithos.2013.12.013. The existence of an ultramafic juvenile lower crust is supported by the available geophysical data. Seismic data show that there exists a Acknowledgments ~14–20 km thick high-velocity layer with Vp = 7.2–7.5 km/s and Vs = 6 km/s (Kind et al., 1996) below 60 km depth within the south- This study is supported by grants from the Ministry of Science and ern Tibetan lower crust, and it has been interpreted as a high-density Technology of China (973 Project 2011CB4031006), IGCP/SIDA-600, (N3.0 g/cm3), eclogitic mafic rock layer (Nábělek et al., 2009; Owens NSFC (41102033), the Program of the China Geological Survey and Zandt, 1997). Delamination of the lower continental crust has (1212011121255, 12120113037900), and the Fundamental Research generally been proposed to account for the intermediate composition Funds for the Central Universities (53200959586). We are most grateful of the continental crust (Gao et al., 2004; Jagoutz, 2010; Jagoutz et al., to the two anonymous referees for their critical and constructive 2011; Rudnick, 1995). However, as shown by the seismic studies, both reviews of this manuscript. the lower crust and a large volume of continental lithospheric mantle are still preserved below southern Tibet (Nábělek et al., 2009; Schulte- Appendix I. Descriptions of analytical methods Pelkum et al., 2005), which is inconsistent with the idea of delamination of the lower continental crust. Thus, delamination was not the cause of 1. LA-ICP-MS and SIMS Zircon U–Pb dating the shift in the composition of the juvenile southern Tibetan crust from basaltic to andesitic. The possible explanation is that, over time, an ul- In this study, eight samples were subjected to zircon separations by tramafic lower crust gradually develops with high densities and acous- standard density and magnetic separation techniques and then, togeth- tic velocities that become indistinguishable from those of mantle er with zircon standard TEMORA 2 and 91500, mounted on epoxy peridotite by geophysical methods (Griffin and O'Reilly, 1987; Kay and mounts that were then polished in order to section the crystals in half Kay, 1985; Kushiro, 1990), and at the same time the rest of the crust is for analysis. Zircons were documented with transmitted and reflected shifting gradually towards bulk intermediate compositions. light micrographs as well as cathodoluminescence (CL) images to reveal their internal structures. The mount was vacuum-coated with an 5.3. Implication for crustal thickening in southern Tibet ~500 nm high-purity . Measurements of U, Th and Pb isotopes were conducted using As discussed above, southern Tibet has been thickened to approxi- the Laser Ablation-Inductively Coupled Plasma-Mass Spectrometry mately twice the average thickness of the Earth's continental crust. (LA-ICP-MS) at the Institute of Geology and Geophysics, Chinese However, the period over which this thickness was attained in southern Academy of Sciences (Beijing), with technical specifications essentially Tibet remains poorly constrained. In most previous studies addressing identical to those described by Xie et al. (2008). An Agilent 7500a this topic it has been suggested that the Indian–Asian continental quadruple (Q)-ICP-MS was used for determinations of zircon U–Pb collision, taking place since the Paleocene, could be the primary cause ages were measured with an attached GeoLas Plus 193-nm excimer of the thickening (Chung et al., 2003; Gao et al., 2007, 2010; Hou et al., ArF laser-ablation system. All of the gas lines were purged for over 1 h 204 2004; Mo et al., 2005; Xu et al., 2010), while other researchers have prior to each analytical session to reduce the Pb on the surface to Pb - suggested that the Tibetan crust had already been thickened up to 50 cps in the gas blank. The measurements were carried out using ~50 km in the Cretaceous, prior to the Indian–Asian collision (Ding time-resolved analysis operating in a fast, peak-hopping sequence in 29 and Lai, 2003; Ding et al., 2003; Kapp et al., 2005, 2007; Murphy et al., the DUAL detector mode. Raw count rates of Si, 204Pb, 206Pb, 207Pb, 1997; Wen et al., 2008). One major piece of evidence for the latter 208Pb, 232Th and 238U were collected for age determination. The 202Hg hypothesis comes from the deduction that Late Cretaceous adakite- resistance is usually b10 cps in the gas blank; therefore, the Y. Zheng et al. / Lithos 190–191 (2014) 240–263 259 contribution of 204Hg to 204Pb is negligible and was not considered fur- 3. Zircon oxygen isotopes ther. The integration time for the four Pb isotopes was 62.76 ms, and it was 30 ms for the other isotopes (including 29Si, 232Th and 238U). Data Zircon oxygen isotopes were measured using the same Cameca IMS were acquired over 30 s with the laser off and 40 s with the laser on, 1280 SIMS, with technical specifications essentially identical to those giving ~340 (=170 readings/replicate × 2 sweeps) mass scans for a described by Li et al. (2009). The Cs + primary ion beam was accelerat- penetration depth of ~20 μm. The U, Th and Pb concentrations were ed at 10 kV, with an intensity of ~2 nA (Gaussian mode with a primary calibrated by using 29Si as an internal standard and NIST SRM 610 beam aperture of 200 μm to reduce aberrations) and rastered over a as an external standard. The 207Pb/206Pb and 206Pb/238U ratios were cal- 10 μm area. The spot is about 20 μmindiameter(10μm beam diameter culated using GLITTER 4.0 (Jackson et al., 2004) and were then corrected +10 μm raster). The normal-incidence electron flood gun was used to using the Harvard zircon 91500 as an external standard. The 207Pb/235U compensate for sample charging during analysis with homogeneous ratio was calculated from the values of 207Pb/206Pb and 206Pb/238U. electron density over a 100 μm oval area. Negative secondary ions CommonPbwascorrectedaccordingtothemethodproposedby were extracted with a −10 kV potential. The field aperture was set to Andersen (2002). 5000 μm, and the transfer-optics magnification was adjusted to give a Only one sample of the two-mica granite (BB–55) was conducted fieldofviewof125μm (FA = 8000). Energy slit width was 30 eV, the using the CAMECA IMS 1280 ion microprobe (CASIMS) at the Institute mechanical position of the energy slit was controlled before starting of Geology and Geophysics, Chinese Academy of Sciences (Beijing), the analysis (5 eV gap, −500 digits with respect to the maximum). with technical specifications essentially identical to those described by The entrance slit width is ~120 μm and exit slit width for multicollector Li et al. (2009). The ellipsoidal spot is about 20 × 30 μminsize.U–Th– Farady cups (FCs) for 16Oand18Ois500μm (MRP = 2500). The inten- Pb isotope ratios were determined relative to the 1065 Ma standard sity of 16O was typically 1 × 109 cps. Oxygen isotopes were measured in zircon 91500 with Th and U concentrations of ~29 and 81 ppm respec- multi-collector mode using two off-axis Faraday cups. The NMR tively (Wiedenbeck et al., 1995). Measured compositions were (Nuclear Magnetic Resonance) probe was used for magnetic field corrected for common Pb using the 204Pb-correction method. An control with stability better than 2.5 ppm over 16 h on mass 17. One average Pb of present-day crustal composition (Stacey and Kramers, analysis takes ~4 min consisting of pre-sputtering (~120 s), automatic 1975) is used for the common Pb assuming that it is largely due to beam centering (~60 s) and integration of oxygen isotopes (10 cycles surface contamination introduced during sample preparation. Because × 4 s, total 40 s). Uncertainties on individual analyses are reported at our measured zircon 206Pb/204Pb ratios are mostly higher than 10,000 1σ level. With low noise on the two FC amplifiers, the internal precision the common-Pb corrections were insensitive to the choice of common- of a single analysis is generally better than 0.2‰ for 18O/16O ratio. Values Pb compositions. The weighted mean U–Pb ages and concordia plots of δ18O are standardized to VSMOW and reported in standard per mil were processed using ISOPLOT 3.0. notation. The instrumental mass fractionation factor (IMF) is corrected 18 using 91500 zircon standard with (δ O)VSMOW =9.9‰ (Wiedenbeck 2. LA-MC-ICPMS Zircon Lu–Hf isotope measurements et al., 2004). Measured 18O/16O is normalized by using Vienna Standard Mean Ocean Water compositions (VSMOW), and then corrected for the In situ Lu–Hf isotopic measurements were carried out on zircon instrumental mass fractionation factor (IMF) as follows: grains with concordant ages within samples that were previously 01 dated by LA-ICP-MS. In situ Lu–Hf analyses were performed by the LA- 18 16 O= O MC-ICPMS method using a Thermo Finnigan Neptune multicollector- δ18 ¼ @ M − A ðÞ‰ ; O : 1 1000 ICPMS and a Geolas CQ 193-nm laser ablation system housed at the M 0 0020052 Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing. The Lu–Hf isotopic analyses were performed on the same zircon IMF ¼ δ18O – δ18O ; ðÞ – Mstandard VSMOW grains that were previously analyzed for U Pb ages, with ablation pits of 63 μm in diameter and an ablation time of 26 s. Detailed descriptions of δ18 ¼ δ18 þ : Osample O IMF the analytical procedures can be found in Wu et al. (2006). M The results of 176Hf/177Hf for JMC475 Hf standard solution give an average 176Hf/177Hf ratio of 0.282158 ± 16 (n = 140, 2SD) normalized to 179Hf/177Hf = 0.7325 using an exponential law for mass bias cor- 4. Major and trace elements rection (Wu et al., 2006). During laser ablation analyses, the isobaric interference of 176Lu on 176Hf is negligible due to the extremely low The major sample locations analyzed are shown in Fig. 1C. Samples 176Lu/177Hf in zircon (normally b0.002). However, the interference of for geochemical analyses were ground to pass through a 200 mesh 176Yb on 176Hf must be carefully corrected since the contribution of and further ground and homogenized in an agate mortar under alcohol. 176Yb to 176Hf could profoundly affect the accuracy of the measured Major element oxides, trace elements and rare earth elements (REEs) of 176Hf/177Hf ratio. In this project, the mean 173Yb/171Yb ratio of the indi- those samples were analyzed by X–ray fluorescence (XRF) and by vidual spots was used to calculate the fractionation coefficient (βYb), inductively coupled plasma mass spectrometry (ICP-MS), respectively, and then to calculate the contribution of 176Yb to 176Hf. It is shown at the National Research Centre for Geoanalysis, Chinese Academy of that this method can provide an accurate correction of the 176Yb inter- Geological Science (Beijing). The analytical uncertainty of XRF analyses ference on 176Hf (Kemp et al., 2009; Woodhead et al., 2004; Wu et al., for major elements was within 5%, and the uncertainty of the elements 2006). During analysis, an isotopic ratio of 176Yb/172Yb = 0.5887 was examined here was also less than 5% for the ICP-MS analyses. applied (Wu et al., 2006). Standard zircon 91500 was used for external correction. During analytical sessions, the obtained 176Hf/177Hf value of 5. Whole-rock Sr–Nd–Oisotopes 91500 was 0.282301 ± 8 (2σ), which was adjusted to 0.282305 (correction of 0.000004), a standard value recommended for 91500 Sr–Nd isotopic analysis was done by a Triton mass spectrometer (Wu et al., 2006), although it is similar to the values obtained by the (TIMS) at the Isotope Geology Lab, Chinese Academy of Geological solution method, within error (Davis et al., 2005; Wiedenbeck et al., Science (Beijing). For the NBS987 standard, the ratio of 87Sr/86Sr = 1995; Woodhead et al., 2004). During data acquisition, analyses of 0.71025 ± 2 (2σ). The measurement accuracy of the Rb/Sr ratio was TEMORA-2 as an unknown yielded a weighted 176Hf/177Hf ratio of better than 0.1% and the mass fractionation of Sr isotopes was corrected 0.282673 ± 5 (2σ, n = 50), identical to the recommended value with- by using 88Sr/86Sr = 8.37521; the J&M ratio of 143Nd/144Nd = in error (Wu et al., 2006). 0.511125 ± 8 (2σ), while the measurement accuracy of the Sm/Nd 260 Y. Zheng et al. / Lithos 190–191 (2014) 240–263 ratio was better than 0.1% and the mass fractionation of Nd isotopes was Carroll, M.R., Wyllie, P.J., 1989. Experimental phase relations in the system tonalite–peri- 146 144 87 86 147 dotite-H2O at 15 kb; implications for assimilation and differentiation processes near corrected by using Nd/ Nd = 0.7219. The Rb/ Sr and Sm/ the crust–mantle boundary. Journal of Petrology 30, 1351–1382. 144 Nd ratios were calculated using the Rb, Sr, Sm and Nd concentrations Castillo, P.R., Janney, P.E., Solidum, R.U., 1999. Petrology and geochemistry of Camiguin obtained by ICP-MS. Island, southern Philippines: insights to the source of adakites and other lavas in a complex arc setting. Contributions to Mineralogy and Petrology 134, 33–51. Whole-rock oxygen isotope analyses were performed on 51 samples Chappell, B.W., 1996. Magma mixing and the production of compositional variation with- at the Isotope Geology Lab, Chinese Academy of Geological Science in granite suites: evidence from the granites of Southeastern Australia. Journal of Pe- – (Beijing). Oxygen from silicate was liberated by reaction with BrF5 in Ni trology 37, 449 470. bombs. Reactions were conducted at 550 °C for more than 15 h. The oxy- Chappell, B.W., White, A.J.R., 1992. I- and S-type granites in the Lachlan Fold Belt. Trans- actions of the Royal Society of Edinburgh: Earth Sciences 83, 1–26. gen was converted to CO2 by reaction with a hot carbon rod, and then iso- Chappell, B.W., White, A.J.R., Wyborn, D., 1987. The importance of residual source – tope ratios were analyzed as CO2 on a Finnigan MAT 252 mass material (restite) in granite petrogenesis. Journal of Petrology 28, 1111 1138. – – – spectrometer, and the results are expressed relative to V-PDB (C) and Chen, L., 2011. The metallogenesis of Oligocene Nuri skarn porphyry Cu W Mo deposit in south Gangdese. Ph.D. Thesis 1–228. V-SMOW (O). Internal and international standards run at the same time Chu, M.F., Chung, S.L., Song, B., Liu, D.Y., O'Reilly, S.Y., Pearson, N.J., Ji, J.Q., Wen, D.J., 2006. with precision of δ18Oandδ13C for the samples being better than 0.2‰. Zircon U–Pb and Hf isotope constraints on the Mesozoic tectonics and crustal evolu- tion of southern Tibet. Geology 34, 745–748. Chu, M.F., Chung, S.L., O'Rilly, S.Y., Pearson, N.J., Wu, F.Y., Li, X.H., Liu, D.Y., Ji, J.Q., Chu, C.H., References for Appendix I Lee, H.Y., 2011. India's hidden inputs to Tibetan orogeny revealed by Hf isotopes of Transhimalayan zircons and host rocks. Earth and Planetary Science Letters 307, 204 479–486. Andersen, T., 2002. Correction of common lead in U–Pb analyses that do not report Pb. Chung, S.L., Liu, D.Y., Ji, J.Q., Chu, M.F., Lee, Y.Y., Wen, D.J., Lo, C.H., Lee, T.Y., Qian, Q., Zhang, Chemical Geology 192, 59–79. Q., 2003. Adakites from continental collision zones, melting of thickened lower crust Davis, D.W., Amelin, Y., Nowell, G.M., Parrish, R.R., 2005. Hf isotopes in zircon from the beneath southern Tibet. Geology 31, 1021–1024. western Superior province, Canada: implications for Archean crustal development Chung, S.L., Chu, M.F., Zhang, Y.Q., Xie, Y.W., Lo, C.H., Lee, T.Y., Lan, C.Y., Li, X.H., Zhang, Q., and evolution of the depleted mantle reservoir. Precambrian Research 140, 132–156. Wang, Y.Z., 2005. Tibetan tectonic evolution inferred from spatial and temporal var- Jackson, S.E., Pearson, N.J., Griffin, W.L., Belousova, E.A., 2004. The application of laser iations in post-collisional magmatism. Earth Science Reviews 68, 173–196. ablation–inductively coupled plasma-mass spectrometry to in situ U–Pb zircon 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. The geochronology. Chemical Geology 211, 47–69. nature and timing of crustal thickening in Southern Tibet. Tectonophysics 477, 36–48. Kemp, A.I.S., Foster, G.L., Scherstén, A., Whitehouse, M.J., Darling, J., Storey, C., 2009. Con- Cloos, M., 1993. Lithospheric buoyancy and collisional orogenesis: subduction of oceanic current Pb–Hf isotope analysis of zircon by laser ablation multi-collector ICPMS, with plateaus, continental margins, island arcs, spreading ridges and seamounts. Geologi- implications for the crustal evolution of Greenland and Himalayas. Chemical Geology cal Society of America Bulletin 105, 715–737. 261, 244–260. Condie, K.C., 2005. TTGs and adakites: are they both slab melts? Lithos 80, 33–44. Stacey, J.S., Kramers, J.D., 1975. Approximation of terrestrial lead isotope evolution by a Coulon, C., Maluski, H., Bollinger, C., Wang, S., 1986. Mesozoic and cenozoic volcanic rocks two-stage model. Earth and Planetary Science Letters 26, 207–221. from central and southern Tibet: 39Ar/40Ar dating, petrological characteristics and Wiedenbeck, M., Allé, P., Corfu, F., Griffin, W.L., Meier, M., Oberli, F., Vonquadt, A., Roddick, geodynamical significance. Earth and Planetary Science Letters 79, 281–302. J.C., Spiegel, W., 1995. Three natural zircon standards for U–Th–Pb, Lu–Hf, trace Defant, M.J., Drummond, M.S., 1990. Derivation of some modern arc magmas by melting element and REE analyses. Geostandards Newsletter 19, 1–23. of young subducted lithosphere. Nature 347, 662–665. Wiedenbeck, M., Hanchar, J.M., Peck, W.H., Sylvester, P., Valley, J., Whitehouse, M., Kronz, Delong, S.E., Schwarz, W.M., Anderson, R.N., 1979. Thermal effects of ridge subduction. A., Morishita, Y., Nasdala, L., Fiebig, J., Franchi, I., Girard, J.P., Greenwood, R.C., Hinton, Earth and Planetary Science Letters 44, 239–246. R., Kita, N., Mason, P.R.D., Norman, M., Ogasawara, M., Piccoli, P.M., Rhede, D., Satoh, DePaolo, D.J., 1981. Neodymium isotopes in the Colorado Front Range and crust–mantle H., Schulz-Dobrick, B., Skår, O., Spicuzza, M.J., Terada, K., Tindle, A., Togashi, S., evolution in the Proterozoic. Nature 291, 193–196. Vennemann, T., Xie, Q., Zheng, Y.F., 2004. Further characterisation of the 91500 zircon Didier, J., 1987. Contribution of enclave studies to the understanding of origin and crystal. Geostandards and Geoanalytical Research 28, 9–39. evolution of granitic magmas. Geologische Rundschau 76, 41–50. Woodhead, J., Hergt, J., Shelley, M., Eggins, S., Kemp, R., 2004. Zircon Hf-isotope analysis Ding, L., Lai, Q.Z., 2003. Geological evidence for thickening and uplift of the Gangdese crust with an excimer laser, depth profiling, ablation of complex geometries, and concom- before collision—constraints of arc convergence on the uplift and extension of the itant age estimation. Chemical Geology 209, 121–135. Qinghai–Tibetan plateau. Chinese Science Bulletin 48, 836–842 (in Chinese with Wu, F.Y., Yang, Y.H., Xie, L.W., Yang, J.H., Xu, P., 2006. Hf isotopic compositions of the English abstract). standard zircons and baddeleyites used in U–Pb geochronology. Chemical Geology Ding, L., Kapp, P., Zhong, D.L., Deng, W.M., 2003. Cenozoic volcanism in Tibet: evidence for a 234, 105–126. transition from oceanic to continental subduction. Journal of Petrology 44, 1833–1865. Xie, L.W., Zhang, Y.B., Zhang, Y.H., Sun, J.F., Wu, F.Y., 2008. In-situ simultaneous determi- Eichelberger, J.C., 1980. Vesiculation of mafic magma during replenishment of silicic nation of trace elements, U–Pb and Lu–Hf isotopes in zircon and baddeleyite. Chinese magma reservoirs. Nature 288, 446–450. Science Bulletin 53, 1565–1573. Elburg, M.A., Van Bergen, M., Hoogewerff, J., Foden, J., Vroon, P., Zulkarnain, I., Nasution, A., 2002. Geochemical trends across an arc–continent collision zone: magma sources References and slab-wedge transfer processes below the Pantar Strait volcanoes, Indonesia. Aggarwal, P.K., Longstaffe, F.J., 1987. Oxygen-isotope geochemistry of metamorphosed, Geochimica et Cosmochimica Acta 66, 2771–2789. massive sulfide deposits of the Flon-Snow Lake Belt, Manitoba. Contributions to Min- Gao, S., Rudnick, R.L., Yuan, H.L., Liu, X.M., Liu, Y.S., Xu, W.L., Liang, W.L., Ayers, J., Wang, eralogy and Petrology 96, 314–325. X.C., Wang, Q.H., 2004. Recycling lower continental crust in the North China craton. Allègre, C.J., Rousseau, D., 1984. The growth of the continent through geological time Nature 432, 892–897. studied by Nd isotope analysis of shales. Earth and Planetary Science Letters 67, Gao, Y.F., Hou, Z.Q., Kamber, B.S., Wei, R.H., Meng, X.J., Zhao, R.S., 2007. Adakite-like 19–34. porphyries from the southern Tibetan continental collision zones: evidence for slab Appleby, S., Graham, C., Gillespie, M., Hinton, R., Oliver, G., 2008. A cryptic record of melt metasomatism. Contributions to Mineralogy and Petrology 153, 105–120. magma mixing in diorites revealed by high-precision SIMS oxygen isotope analysis Gao, Y.F., Wei, R.H., Hou, Z.Q., Tian, S.H., Zhao, R.S., 2008. Eocene high-MgO volcanism in of zircons. Earth and Planetary Science Letters 269, 105–117. southern Tibet: new constraints for mantle source characteristics and deep processes. Atherton, M.P., Petford, N., 1993. Generation of sodium-rich magmas from newly Lithos 105, 63–72. underplated basaltic crust. Nature 362, 144–146. Gao, Y.F., Yang, Z.S., Santosh, M., Hou, Z.Q., Wei, R.H., Tian, S.H., 2010. Adakitic rocks from Bachman, O., Dungan, M.A., Bussy, F., 2005. Insights into shallow magmatic processes in slab melt-modified mantle sources in the continental collision zone of southern Tibet. large silicic magma bodies: the trace element record in the Fish Canyon magma Lithos 119, 651–663. body, Colorado. Contributions to Mineralogy and Petrology 149, 338–349. Griffin, W., O'Reilly, S.Y., 1987. Is the continental Moho the crust–mantle boundary? Baker, M.B., Grove, T.L., Price, R., 1994. Primitive basalts and from the Mt. Shasta Geology 15, 241–244. region, N. California: products of varying melt fraction and water content. Contribu- Griffin,W.L.,Pearson,N.J.,Belousova,E.,Jackson,S.E., van Achterbergh, E., O'Reilly, S.Y., Shee, tions to Mineralogy and Petrology 118, 111–129. S.R., 2000. The Hf isotope composition of cratonic mantle: LA-MC-ICPMS analysis of zir- Baker, M., Hirschmann, M., Ghiorso, M., Stolper, E., 1995. Compositions of near-solidus pe- con megacrysts in kimberlites. Geochimica et Cosmochimica Acta 64, 133–147. ridotite melts from experiments and thermodynamic calculation. Nature 375, 308–311. Griffin, W.L., Pearson, N.J., Belousova, E.A., Saeed, A., 2006. Comment: Hf-isotope hetero- Barth, M.G., McDonough, W.F., Rudnick, R.L., 2000. Tracking the budget of Nb and Ta in geneity in standard zircon 91500. Chemical Geology 233, 358–363. the continental crust. Chemical Geology 165, 197–213. Guan, Q., Zhu, D.C., Zhao, Z.D., Zhang, L.L., Liu, M., Li, X.W., Yu, F., Mo, X.X., 2010. Late Cre- Ben Othman, D., White, W.M., Patchett, J., 1989. The geochemistry of marine sediments: taceous adakites in the eastern segment of the Gangdese Belt, southern Tibet: prod- island arc magma genesis and crust–mantle recycling. Earth and Planetary Science ucts of Neo-Tethyan ridge subduction? Acta Petrologica Sinica 26, 2165–2179 (in Letters 94, 1–21. Chinese with English abstract). Besse, J., Courtillot, V., Pozzi, J.P., Westphal, M., Zhou, Y.X., 1984. Palaeomagnetic estimates Guan, Q., Zhu, D.C., Zhao, Z.D., Dong, G.C., Mo, X.X., Liu, Y.S., Hu, Z.C., Yuan, H.L., 2011. Zir- of crustal shortening in the Himalayan thrusts and Zangbo suture. Nature 311, con U–Pb chronology, geochemistry of the Late Cretaceous mafic magmatism in the 621–625. southern Lhasa Terrane and its implications. Acta Petrologica Sinica 27, 2083–2094 Bourdon, E., Eissen, J.P., 2003. Magmatic response to early aseismic ridge subduction: the (in Chinese with English abstract). Ecuadorian margin case (South America). Earth and Planetary Science Letters 205 Guan, Q., Zhu, D.C., Zhao, Z.D., Dong, G.C., Zhang, L.L., Li, X.W., Liu, M., Liu, M.H., Mo, X.X., (3–4), 123–138. Liu, Y.S., Yuan, H.L., 2012. Crustal thickening prior to 38 Ma in southern Tibet: Y. Zheng et al. / Lithos 190–191 (2014) 240–263 261

evidence from lower crust-derived adakitic magmatism in the Gangdese Batholith. Kay, S.M., Ramos, V.A., Marquez, M., 1993. Evidence in Cerro Pampa volcanic rocks for Gondwana Research 21, 88–99. slab-melting prior to ridge–trench collision in southern South America. Journal of Guivel, C., Lagabrielle, Y., Bourgois, J., Maury, R.C., Fourcade, S., Martin, H., Arnaud, N., Geology 101, 703–714. 1999. New geochemical constraints for the origin of ridge subduction-related pluton- Keleman, P.B., 1990. Reaction between ultramafic rock and fractionated basaltic magma. I. ic and volcanic suites from the Chile Triple Junction (Taitao Peninsula and Site 862, Phase relations, the origin of calc-alkaline magma series, and the formation of discor- LEG ODP141 on the Taitao Ridge). Tectonophysics 311, 83–111. dant dunite. Journal of Petrology 31, 51–98. Guivel, C., Morata, D., Pelleter, E., Espinoza, F., Maury, R.C., Lagabrielle, Y., Polvé, M., Bellon, Keleman, P.B., 1995. Genesis of high Mg# andesites and the continental crust. Contribu- H.,Cotten,J.,Benoit,M.,Suárez,M.,delaCruz,R.,2006.Miocene to Late Quaternary tions to Mineralogy and Petrology 120, 1–19. Patagonian basalts (46–47°S): geochronometric and geochemical evidence for slab Keleman, P.B., Shimizu, N., Dunn, T., 1993. Relative depletion of niobium in some arc tearing due to active spreading ridge subduction. Journal of Volcanology and magmas and the continental crust, partitioning of K, Nb, La and Ce during meltyrock Geothermal Research 149, 346–370. reaction in the upper mantle. Earth and Planetary Science Letters 120, 111–134. Guo, Z., Hertogen, J., Liu, J., Pasteels, P., Boven, A., Punzalan, L., He, H., Luo, X., Zhang, W., Kemp, A.I.S., Hawkesworth, C.J., Foster, G.L., Paterson, B.A., Woodhead, J.D., Hergt, J.M., 2005. Potassic magmatism in western Sichuan and Yunnan provinces, SE Tibet, Gray, C.M., Whitehouse, M.J., 2007. Magmatic and crustal differentiation history of China: petrological and geochemical constraints on petrogenesis. Journal of Petrology granitic rocks from Hf–O isotopes in zircon. Science 315, 980–983. 46, 33–78. Kind, R., Ni, J., Zhao, W., Wu, J., Yuan, X., Zhao, L., Sandvol, E., Reese, C., Nabelek, J., Hearn, Guo, Z.F., Wilson, M., Liu, J.Q., 2007. Post-collisional adakites in south Tibet: products of T., 1996. Evidence from earthquake data for partially molten crustal layer in Southern partial melting of subduction-modified lower crust. Lithos 96, 205–224. Tibet. Science 274, 1692–1694. Harrison, T.M., Yin, A., Grove, M., Lovera, O.M., Ryerson, F., Zhou, X., 2000. The Zedong Kinoshita, O., 1995. Migration of igneous activities related to ridge subduction in South- window: a record of superposed Tertiary convergence in southeastern Tibet. Journal west Japan and the East Asian continental margin from the Mesozoic to the Paleo- of Geophysical Research 105, 9211–9230. gene. Tectonophysics 245, 25–35. Hawkesworth, C., Turner, S., Peate, D., McDermott, F., Van Calsteren, P., 1997. Elemental U Klootwijk, C.T., Gee, J.S., Peirce, J.W., Smith, G.M., McFadden, P.L., 1992. An early India–Asia and Th variations in island arc rocks: implications for U-series isotopes. Chemical contact: Paleomagnetic constraints from Ninetyeast Ridge, ODP leg 121. Geology 20, Geology 139, 207–221. 395–398. Hildreth, W., Moorbath, S., 1988. Crustal contributions to arc magmatism in the Andes of Kushiro, I., 1990. Partial melting of mantle wedge and evolution of island arc crust. Journal central Chile. Contributions to Mineralogy and Petrology 98, 455–489. of Geophysical Research 95, 15929–15939. Hofmann, A., Jochum, K., Seufert, M., White, W., 1986. Nb and Pb in oceanic basalts: new Lackey, J.S., Valley, J.W., Saleeby, J.B., 2005. Supracrustal input to magmas in the deep crust constraints on mantle evolution. Earth and Planetary Science Letters 79, 33–45. of Sierra Nevada batholith: evidence from high δ18O zircon. Earth and Planetary Sci- Holden, P., Halliday, A.N., Stephens, W.E., 1987. Neodymium and strontium isotope con- ence Letters 235, 315–330. tent of microdiorite enclaves points to mantle input to granitoid production. Nature Lackey, J., Valley, J., Hinke, H., 2006. Deciphering the source and contamination history of 330, 53–56. peraluminous magmas using δ18O of accessory minerals: examples from garnet- Honegger, K., Dietrich, V., Frank, W., Gansser, A., Thöni, M., Trommsdorff, V., 1982. bearing plutons of the Sierra Nevada batholith. Contributions to Mineralogy and Magmatism and metamorphism in the Ladakh Himalayas (the Indus–Tsangpo suture Petrology 151, 20–44. zone). Earth and Planetary Science Letters 60, 253–292. Lackey, J.S., Valley, J.W., Chen, J.H., Stockli, D.F., 2008. Dynamic magma systems, crustal Hou, Z.Q., Gao, Y.F., Qu, X.M., Rui, Z.Y., Mo, X.X., 2004. Origin of adakitic intrusives gener- recycling, and alteration in the central Sierra Nevada batholith: the oxygen isotope ated during mid-Miocene east–west extension in southern Tibet. Earth and Planetary record. Journal of Petrology 49, 1397–1426. Science Letters 220, 139–155. Leech, M.L., Singh, S., Jain, A.K., Klemperer, S.L., Manickavasagam, R.M., 2005. The onset of Hou, Z.Q., Zhao, Z.D., Gao, Y.F., Yang, Z.M., Jiang, W., 2006. Tearing and subduction of the India–Asia continental collision: early, steep subduction required by the timing of UHP Indian continental slab: evidence from Cenozoic Gangdese igneous rocks in Tibet. metamorphism in the western Himalaya. Earth and Planetary Science Letters 234, Acta Petrologica Sinica 22, 761–774 (in Chinese with English abstract). 83–97. Hou, Z.Q., Yang, Z.M., Qu, X.M., Meng, X.J., Li, Z.Q., Beaudoin, G., Rui, Z.Y., Gao, Y.F., Khin, Z., Leier, A.L., He, S., Kapp, P., Ding, L., 2002. Jurassic–Cretaceous deposits of the central Lhasa 2009. The Miocene Gangdese Porphyry Copper Belt generated during post-collisional terrane: implications for the tectonic evolution of southern Asia prior to the Indo- extension in the Tibetan Orogen. Ore Geology Reviews 36, 25–51. Asian collision. Geological Society of America Abstracts with Programs 34 (6), 412. Hou, Z.Q., Zheng, Y.C., Zeng, L.S., Gao, L.E., Huang, K.X., Li, W., Li, Q.Y., Fu, Q., Liang, W., Sun, Leier, A.L., DeCelles, P.G., Kapp, P., 2007. The Takena Formation of the Lhasa terrane, Q.Z., 2012. Eocene–Oligocene granitoids in southern Tibet: constraints on crustal southern Tibet: the record of a Late Cretaceous retroarc foreland basin. Geological anatexis and tectonic evolution of the Himalayan orogen. Earth and Planetary Science Society of America Bulletin 119, 31–48. http://dx.doi.org/10.1130/B25974.1. Letters 349, 38–52. Li, X.H., Liu, Y., Li, Q.L., Guo, C.H., Chamberlain, K.R., 2009. Precise determination of Phan- Hou, Z.Q., Zheng, Y.C., Yang, Z.M., Rui, Z.Y., Zhao, Z.D., Jiang, S.H., Qu, X.M., Sun, Q.Z., 2013. erozoic zircon Pb/Pb age by multicollector SIMS without external standardization. Contribution of mantle components within juvenile lower-crust to collisional zone Geochemistry, Geophysics, Geosystems 10, Q04010. porphyry Cu systems in Tibet. Mineralium Deposita 48, 173–192. Li, X.H., Li, W.X., Li, Q.L., Wang, X.C., Liu, Y., Yang, Y.H., 2010. Petrogenesis and tectonic signif- Iwamori, H., 2000. Thermal effects of ridge subduction and its implications for the origin icance of the 850 Ma Gangbian alkaline complex in South China: evidence from in situ of granitic batholith and paired metamorphic belts. Earth and Planetary Science zircon U–Pb dating, Hf–O isotopes and whole-rock geochemistry. Lithos 114, 1–15. Letters 181, 131–144. Ma, L., Wang, Q., Wyman, D.A., Li, Z.X., Jiang, Z.Q., Yang, J.H., Gou, G.N., Guo, H.F., 2013. Late Jagoutz, O., 2010. Construction of the granitoid crust of an island arc. Part II: a quantitative Cretaceous (100–89 Ma) magnesian charnockites with adakitic affinities in the Milin petrogenetic model. Contributions to Mineralogy and Petrology 160, 359–381. area, eastern Gangdese: partial melting of subducted oceanic crust and implications Jagoutz, O., Müntener, O., Schmidt, M.W., Burg, J.P., 2011. The roles of flux- and de- for crustal growth in southern Tibet. Lithos 175–176, 315–332. compression melting and their respective fractionation lines for continental Macpherson, C.G., Dreher, S.T., Thirlwall, M.F., 2006. Adakites without slab melting: high crust formation: evidence from the Kohistan arc. Earth and Planetary Science Let- pressure differentiation of island arc magma, Mindanao, the Philippines. Earth and ters 303, 25–36. Planetary Science Letters 243, 581–593. Ji, W.Q., Wu, F.Y., Chung, S.L., Li, J.X., Liu, C.Z., 2009. Zricon U–Pb geochronogy and Hf iso- Mahoney, J.J., Frei, R., Tejada, M.L.G., Mo, X.X., Leat, P.T., Nägler, T.F., 1998. Tracing the topic constraints on petrogensis of the Gangdese batholith, southern Tibet. Chemical Indian Ocean mantle domain through time isotopic results from old West Indian, Geology 262, 229–245. East Tethyan and South Pacificseafloor. Journal of Petrology 39, 1285–1306. Ji, W.Q., Wu, F.Y., Chung, S.L., Liu, C.Z., 2012. Identification of early Carboniferous granit- Martin, H., 1987. Petrogenesis of Archaean trondhjemites, tonalites and granodiorites from oids from southern Tibet and implications for terrane assembly related to the eastern Finland: major and trace element geochemistry. Journal of Petrology 28, Paleo-Tethyan evolution. Journal of Geology 120, 531–541. 921–953. Jiang, Z.Q., Wang, Q., Li, Z.X., Wyman, D.A., Tang, G.J., Jia, X.H., Yang, Y.H., 2012. Late Martin, H., 1999. Adakitic magmas, modern analogues of Archaean granitoids. Lithos 46, Cretaceous (ca. 90 Ma) adakitic intrusive rocks in the Kelu area, Gangdese Belt 411–429. (southern Tibet): slab melting and implications for Cu–Au mineralization. Journal Martin, H., Smithies, R., Rapp, R., Moyen, J.F., Champion, D., 2005. An overview of adakite, of Asian Earth Sciences 53, 67–81. tonalite–trondhjemite–granodiorite (TTG) and sanukitoid: relationships and some Kapp, P., Yin, A., Harrison, T.M., Ding, L., 2005. Cretaceous-Tertiary shortening, basin implications for crustal evolution. Lithos 79, 1–24. development, and volcanism in central Tibet. Geological Society of America Bulletin McNamara, D.E., Owens, T.J., Walter, W.R., 1995. Observations of regional phase prop- 117, 865–878. agation across the Tibetan plateau. Journal of Geophysical Research 100, Kapp, P., DeCelles, P.G., Gehrels, G.E., Heizler, M., Ding, L., 2007. Geological records of the 22215–22229. Lhasa–Qiangtang and Indo-Asian collisions in the Nima area of central Tibet. Miller, G., Schuster, R., Klötzli, U., Frank, W., Purtscheller, F., 1999. Post-collisional potassic Geological Society of America Bulletin 119, 917–933. and ultrapotassic magmatism in SW Tibet: geochemical and Sr–Nd–Pb–O isotopic Karsten, J.L., Klein, E.M., Sherman, S.B., 1996. Subduction zone geochemical characteristics constraints for mantle source characteristics and petrogenesis. Journal of Petrology in ocean ridge basalts from the southern Chile Ridge: implications of modern ridge 40, 1399–1424. subduction systems for the Archean. Lithos 37, 143–161. Mo, X.X., Dong, G.C., Zhao, Z.D., Guo, T.Y., Wang, L.L., Chen, T., 2005. Timing of magma Kay, S.M., Kay, R.W., 1985. Role of crystal cumulates and the oceanic crust in the forma- mixing in the Gangdisê magmatic belt during the India–Asia collision: zircon tion of the lower crust of the Aleutian arc. Geology 13, 461–464. SHRIMP U–Pb dating. Acta Geologica Sinica 79, 66–76. Kay, R.W., Kay, S.M., 2002. Andean adakites: three ways to make them. Acta Petrologica Mo, X.X., Hou, Z.Q., Niu, Y.L., Dong, G.C., Qu, X.M., Zhao, Z.D., Yang, Z.M., 2007. Mantle con- Sinica 18, 303–311. tributions to crustal thickening during continental collision: evidence from Cenozoic Kay, S.M., Mpodozis, C., 2001. Central Andean ore deposits linked to evolving shallow igneous rocks in southern Tibet. Lithos 96, 225–242. subduction systems and thickening crust. GSA Today 11, 4–9. Molnar, P., Houseman, G., Clinton, C., 1998. Rayleigh–Taylor instability and convective Kay, S.M., Maksaev, V., Moscoso, R., Mpodozis, C., Nasi, C., 1987. Probing the evolving Andean thinning of mechanically thickened lithosphere: effects of non-linear viscosity de- lithosphere: mid-late Tertiary magmatism in Chile (29°-30°30′S) over the modern zone creasing exponentially with depth and of horizontal shortening of the layer. Geo- of subhorizontal subduction. Journal of Geophysical Research 92, 6173–6189. physical Journal International 133, 568–584. 262 Y. Zheng et al. / Lithos 190–191 (2014) 240–263

Murphy, M.A., Yin, A., Harrison, T.M., Dürr, S.B., Chen, Z., Ryerson, F.J., Kidd, W.S.F., Wang, Sekine, T., Wyllie, P.J., 1983. Experimentasl similation of mantle hybridization in subduc- X., Zhou, X., 1997. Did the Indo-Asian collision alone create the Tibetan plateau? Ge- tion zones. Journal of Geology 91, 511–528. ology 25, 719–722. Sisson, V.B., Pavlis, T.L., Roeske, S.M., Thorkelson, D.J., 2003. Introduction: an overview of Nábělek, J., Hetényi, G., Vergne, J., Sapkota, S., Kafle, B., Jiang, M., Su, H., Chen, J., Huang, ridge–trench interactions in modern and ancient settings. Geological Society of B.S., 2009. Underplating in the Himalaya–Tibet collision zone revealed by the Hi- America, Special Paper 371, 1–18. CLIMB experiment. Science 325, 1371–1374. Soderlund, U., Patchett, P.J., Vervoort, J.D., Isachsen, C.E., 2004. The 176Lu decay constant Nebel, O., Münker, C., Nebel-Jacobsen, Y.J., Kleine, T., Mezger, K., Mortimer, N., 2007. Hf–Nd– determined by Lu–Hf and U–Pb isotope systematics of Precambrian mafic intrusions. Pb isotope evidence from Permian arc rocks for the long-term presence of the Indian–Pa- Earth and Planetary Science Letters 219, 311–324. cific mantle boundary in the SW Pacific. Earth and Planetary Science Letters 254, Stern, R.A., Hanson, G.N., 1991. Archean high-Mg granodiorite: a derivative of light rare 377–392. earth element-enriched monzodiorite of mantle origin. Journal of Petrology 32, Nelson, K.D., Zhao, W.J., Brown, L.D., Kuo, J., Che, J.K., Liu, X.W., Klemperer, S.L., Makovsky, 201–238. Y., Meissner, R., Mechie, J., Kind, R., Wenzel, F., Ni, J., Nabelek, J., Chen, L.S., Tan, H.D., Stern, C.R., Kilian, R., 1996. Role of the subducted slab, mantle wedge and continental Wei, W.B., Jones, A.G., Booker, J., 1996. Partially molten middle crust beneath south- crust in the generation of adakites from the Andean Austral volcanic zone. Contribu- ern Tibet: synthesis of Project INDEPTH Results. Science 274, 1684–1688. tions to Mineralogy and Petrology 123, 263–281. Nelson, W.R., Dorais, M.J., Christiansen, E.H., Hart, G.L., 2013. Petrogenesis of Sierra Nevada Sun, S.S., McDonough, W.F., 1989. Chemical and isotopic systematics of oceanic basalts: plutons inferred from the Sr, Nd, and O isotopic signatures of maficigneouscomplexes implications for mantle composition and processes. Geological Society, London, Spe- in Yosemite Valley, California. Contributions to Mineralogy and Petrology 165, 397–417. cial Publications 42, 313–345. Nimaciren, X., Xie, Y.W., Sha, Z.L., Xiluolangjie, Qiangbazaxi, Peng D.P., Gesangsuolang, Sun, X., Zheng, Y.Y., Wu, S., You, Z.M., Wu, X., Li, M., Zhou, T.C., Dong, J., 2013. Mineraliza- Luosongzhandui, 2005. 1: 250, 000 geological report of Nagqu County with geological tion age and petrogenesis of associated intrusions in the Mingze–Chengba porphyry– map. Xizang Institute of Geological Survey, Lhasa, pp 103–107 (unpublished, in Chinese). skarn Mo–Cu deposit, Gangdese. Acta Petrologica Sinica 29, 1392–1406 (in Chinese Owens, T.J., Zandt, G., 1997. Implications of crustal property variations for models of with English abstract). Tibetan plateau evolution. Nature 387, 37–43. Tang, G.J., Wang, Q., Wyman, D.A., Li, Z.X., Zhao, Z.H., Yang, Y.H., 2012a. Late Carbon- Pan, G.T., Ding, J., Yao, D.S., Wang, L.Q., 2004. Guidebook of 1: 1,500,000 geologic map of iferous high εNd(t)–εHf(t) granitoids, enclaves and dikes in western Junggar, NW the Qinghai–Xizang (Tibet) plateau and adjacent areas. Cartographic Publishing China: ridge-subduction-related magmatism and crustal growth. Lithos 140–141, House, Chengdu, China, pp 1–148 (in Chinese). 86–102. Pan, G.T., Mo, X.X., Hou, Z.Q., Zhu, D.C., Wang, L.Q., Li, G.M., Zhao, Z.D., Geng, Q.R., Liao, Z.L., Tang, G.J., Wyman, D.A., Wang, Q., Li, J., Li, Z.X., Zhao, Z.H., Sun, W.D., 2012b. Astheno- 2006. Spatial–temporal framework of the Gangdese Orogenic Belt and its evolution. sphere–lithosphere interaction triggered by a slab window during ridge subduction: Acta Petrologica Sinica 22, 521–533 (in Chinese with English abstract). trace element and Sr–Nd–Hf–Os isotopic evidence from Late Carboniferous tholeiites Pan, G.T., Wang, L.Q., Li, R.S., Yuan, S.H., Ji, W.H., Yin, F.G., Zhang, W.P., Wang, B.D., 2012. in the western Junggar area (NW China). Earth and Planetary Science Letters Tectonic evolution of the Qinghai–Tibet Plateau. Journal of Asian Earth Sciences 53, 329–330, 84–96. 3–14. Tatsumi, Y., 1982. Origin of high-magnesian andesites in the Setouchi volcanic belt, south- Patriat, P., Achache, J., 1984. India–Eurasia collision chronology has implications for crust- west Japan, II. Melting phase relations at high pressures. Earth and Planetary Science al shortening and driving mechanism of plates. Nature 311, 615–621. Letters 60, 305–317. Pavlis, T.L., Sisson, V.B., 1995. Structural history of the Chugach metamorphic complex in Tatsumi, Y., 2006. High-Mg andesites in the Setouchi volcanic belt, Southwestern Japan: the Tana River region, eastern Alaska: a record of Eocene ridge subduction. Geological analogy to Archean magmatism and continental crust formation? Annual Review of Society of America Bulletin 107, 1333–1355. Earth and Planetary Sciences 34, 467–499. Peacock, S.M., 1996. Thermal and petrologic structure of subduction zones. Geophysical Tatsumi, Y., Eggins, S., 1995. Subduction zone magmatism. Blackwell Science, Oxford 211. Monograph Series 96, 119–133. Taylor, S.R., 1967. The origin and growth of continents. Tectonophysics 4, 17–34. Peacock, S.M., Rushmer, T., Thompson, A.B., 1994. Partial melting of subducting oceanic Taylor, S.R., 1977. Island arc models and the composition of the continental crust. Maurice crust. Earth and Planetary Science Letters 121, 227–244. Ewing Series 1, 325–335. Peccerillo, A., Taylor, S.R., 1976. Geochemistry of Eocene calc-alkaline volcanic rocks from Taylor, S.R., McLennan, S.M., 1995. The geochemical evolution of the continental crust. the Kastamonu area, Northern Turkey. Contributions to Mineralogy and Petrology 58, Reviews of Geophysics 33, 241–265. 63–81. Turner, S., Arnaud, N., Liu, J., Rogers, N., Hawkesworth, C., Harris, N., Kelley, S., Calsteren, Peccerillo, A., Barberio, M.R., Yirgu, G., Ayalew, D., Barbieri, M., Wu, T.W., 2003. Relation- P.V., Deng, W., 1996. Post–collision, shoshonitic volcanism on the Tibetan plateau: ship between mafic and peralkaline felsic magmatism in continental rift settings: a implications for convective thinning of the lithosphere and the source of ocean island petrological, geochemical and isotopic study of the Gedemsa Volcano, Central basalts. Journal of Petrology 37, 45–71. Ethiopian Rift. Journal of Petrology 44, 2003–2032. Valley, J.W., 2003. Oxygen isotopes in zircon. Reviews in Mineralogy and Geochemistry Pennington, W.D., 1984. The effect of oceanic crustal structure on phase changes and 53, 343–385. subduction. Tectonophysics 102, 377–398. Valley, J.W., Chiarenzelli, J.R., McLelland, J.M., 1994. Oxygen isotope geochemistry of zir- Plank, T., Langmuir, C.H., 1998. The chemical composition of subducting sediment and its con. Earth and Planetary Science Letters 126, 187–206. consequences for the crust and mantle. Chemical Geology 145, 325–394. Vernon, R.H., 1984. Microgranitoid enclaves in granites—globules of hybrid magma Priestley, K., Debayle, E., McKenzie, D., Pilidou, S., 2006. Upper mantle structure of eastern quenched in a plutonic environment. Nature 304, 438–439. Asia from multimode surface waveform tomography. Journal of Geophysical Re- Vervoort, J.D., Patchett, P.J., Blichert-Toft, J., Albarède, F., 1999. Relationships between Lu– search 111, B10304. http://dx.doi.org/10.1029/2005JB004082. Hf and Sm–Nd isotopic systems in the global sedimentary system. Earth and Plane- Prowatke, S., Klemme, S., 2006. Rare earth element partitioning between titanite and tary Science Letters 168, 79–99. silicate melts: Henry's law revisited. Geochimica et Cosmochimica Acta 70, Wang, Q., Wyman, D.A., Xu, J.F., Jian, P., Zhao, Z.H., Li, C.F., Xu, W., Ma, J.L., He, B., 2007. 4997–5012. Early cretaceous adakitic granites in the Northern Dabie complex, central China: im- Qu, X.M., Hou, Z.Q., Li, Y.G., 2004. Melt components derived from a subducted slab in Late plications for partial melting and delamination of thickened lower crust. Geochimica Orogenic ore-bearing porphyries in the Gangdese copper belt, southern Tibetan et Cosmochimica Acta 71, 2609–2636. plateau. Lithos 74, 131–148. Wang,Q.,Wyman,D.A.,Xu,J.F.,Dong,Y.H.,Vasconcelos,P.M.,Pearson,N.,Wan, Rapp, R.P., Watson, E.B., Miller, C.F., 1991. Partial melting of amphibolite/eclogite and the Y.S.,Dong,H.,Li,C.F.,Yu,Y.S.,Zhu,T.X.,Feng,X.T.,Zhang,Q.Y.,Zi,F.,Chu, origin of Archean trondhjemites and tonalities. Precambrian Research 51, 1–25. Z.Y., 2008. Eocene melting of subducting continental crust and early uplifting Rapp, R.P., Shimizu, N., Norman, M.D., Applegate, G.S., 1999. Reaction between slab- of central Tibet: evidence from central–western Qiangtang high–Kcalc– derived melts and peridotite in the mantle wedge: experimental constraints at alkaline andesites, dacites and rhyolites. Earth and Planetary Science Letters 3.8 GPa. Chemical Geology 145, 117–152. 272, 158–171. Ravikant, V., Wu, F.Y., Ji, W.Q., 2009. Zircon U–Pb and Hf isotopic constraints on petrogen- Wang, J.L., Zhang, Z.M., Dong, X., Liu, F., Yu, F., Wang, W., Xu, F.J., Shen, K., 2009. Discovery esis of the Cretaceous–Tertiary granites in eastern Karakoram and Ladakh, India. Lith- of Late Cretaceous garnet two-pyroxene granulite in the southern Lhasa terrane, os 110, 153–166. Tibet and its tectonic significances. Acta Petrologica Sinica 25, 1678–1694 (in Chinese Richards, J.P., Kerrich, R., 2007. Adakite-like rocks: their diverse origins and questionable with English abstract). role in metallogenesis. Economic Geology 102, 537–576. Wen, D.R., Chung, S.L., Song, B., Iizuka, Y., Yang, H.J., Ji, J.Q., Liu, D.Y., Gallet, S., 2008. Late Ringwood, A.E., 1977. Petrogenesis in island arc systems. Maurice Ewing Series 1, Cretaceous Gangdese intrusions of adakitic geochemical characteristics, SE Tibet: pet- 311–324. rogenesis and tectonic implications. Lithos 105, 1–11. Rudnick, R.L., 1995. Making continental crust. Nature 378, 571–578. Williams, H.M., Turner, S.P., Pearce, J.A., Kelley, S.P., Harris, N.B.W., 2004. Nature of the Rudnick, R.L., Fountain, D.M., 1995. Nature and composition of the continental crust: a source regions for post–collisional, potassic magmatism in Southern and Northern lower crustal perspective. Reviews of Geophysics 33, 267–309. Tibet from geochemical variations and inverse trace element modeling. Journal of Pe- Rudnick, R.L., Gao, S., 2003. Composition of the continental crust. Treatise on Geochemis- trology 45, 555–607. try 3, 1–64. Wilson, M., 1991. Igneous Petrogenesis: a global tectonic approach. Harper Collins Aca- Sajona, F.G., Maury, R.C., Bellon, H., Cotton, J., Defant, M.J., Pubellier, M., 1993. Initiation of demic 466. subduction and the generation of slab melts in western and eastern Mindanao, Woodhead, J., Hergt, J., Davidson, J., Eggins, S., 2001. Hafnium isotope evidence for ‘con- Philippines. Geology 21, 1007–1010. servative’ element mobility during subduction zone processes. Earth and Planetary Schulte-Pelkum, V., Monsalve, G., Sheehan, A., Pandey, M., Sapkota, S., Bilham, R., Wu, F., Science Letters 192, 331–346. 2005. Imaging the Indian subcontinent beneath the Himalaya. Nature 435, 1222–1225. Wu, F.Y., Ji, W.Q., Liu, C.Z., Chung, S.L., 2010. Detrital zircon U–Pb and Hf isotopic data from Searle, M.P., Noble, S.R., Cottle, J.M., Waters, D.J., Mitchell, A.H.G., Hlaing, T., Horstwood, the Xigaze fore-arc basin: constraints on Transhimalayan magmatic evolution in M.S.A., 2007. Tectonic evolution of the Mogok metamorphic belt, Burma southern Tibet. Chemical Geology 271, 13–25. (Myanmar) constrained by U–Th–Pb dating of metamorphic and magmatic rocks. Wyllie, P.J., 1978. Water and magma generation at subduction zones. Bulletin Tectonics 26, TC3014. Volcanologique 41, 360–377. Y. Zheng et al. / Lithos 190–191 (2014) 240–263 263

Xu, J.F., Shinjo, R., Defant, M.J., Wang, Q., Rapp, R.P., 2002. Origin of Mesozoic adakitic in- Zhang, Z.M., Shen, K., Santosh, M., Dong, X., 2011. High density carbonic fluids in a slab trusive rocks in the Ningzhen area of east China: partial melting of delaminated lower window: evidence from the Gangdese charnockite, Lhasa terrane, southern Tibet. continental crust? Geology 30, 1111–1114. Journal of Asian Earth Sciences 42, 515–524. Xu, W.C., Zhang, H.F., Guo, L., Yuan, H.L., 2010. Miocene high Sr/Y magmatism, south Zhang, Z.M., Dong, X., Santosh, M., Zhao, G.C., 2014. Metamorphism and tectonic evolu- Tibet: product of partial melting of subducted Indian continental crust and its tecton- tion of the Lhasa terrane, Central Tibet. Gondwana Research 25, 170–189. ic implication. Lithos 114, 293–306. Zhao, Z.D., Mo, X.X., Dilek, Y., Niu, Y.L., DePaolo, D.J., Robinson, P., Zhu, D.C., Sun, C.G., Yang, J.H., Wu, F.Y., Chung, S.L., Wilde, S.A., Chu, M.F., 2004. Multiple sources for the origin Dong, G.C., Zhou, S., Luo, Z.H., Hou, Z.Q., 2009. Geochemical and Sr–Nd–Pb–O isotopic of granites: geochemical and Nd/Sr isotopic evidence from the Gudaoling granite and compositions of the post-collisional ultrapotassic magmatism in SW Tibet: petrogen- its mafic enclaves, northeast China. Geochimica et Cosmochimica Acta 68, 4469–4483. esis and implications for India intra-continental subduction beneath southern Tibet. Yang, T.N., Zhang, H.R., Liu, Y.X., Wang, Z.L., Song, Y.C., Yang, Z.S., Tian, S.H., Xie, H.Q., Hou, Lithos 113, 190–212. K.J., 2011. Permo-Triassic arc magmatism in central Tibet: evidence from zircon U–Pb Zheng, Y.C., Gu, L.X., Tang, X.Q., Wu, C.Z., Li, C.H., Liu, S.H., 2011a. Geology and geochem- geochronology, Hf isotopes, rare earth elements, and bulk geochemistry. Chemical istry of highly metamorphosed footwall alteration zones in the Hongtoushan Geology 284, 270–282. Vocanogenic Massive Sulfide deposit, Liaoning Province, China. Resource Geology Yin, A., Harrison, T.M., 2000. Geologic evolution of the Himalayan–Tibetan orogen. Annual 61, 113–139. Review of Earth and Planetary Sciences 28, 211–280. Zheng, Y.C., Gu, L.X., Tang, X.Q., Liu, S.H., Wu, C.Z., Li, C.H., 2011b. Oxygen isotope Yogodzinski, G.M., Kay, R.W., Volynets, O.N., Koloskov, A.V., Kay, S.M., 1995. Magnesian an- characterististics of the footwall alteration zones in the Hongtoushan volanogenic desite in the western Aleutian Komandorsky region: implications for slab melting and massive sulfide deposit, Liaoning Province, Chian and restoration of their formation processes in the mantle wedge. Geological Society of America Bulletin 107, 505–519. temperatures. Acta Geologica Sinica-English Edition 85, 683–693. Yogodzinski, G.M., Lees, J.M., Churikova, T.G., Dorendorf, F., Wöerner, G., Volynets, O.N., Zheng, Y.C., Hou, Z.Q., Li, Q.Y., Sun, Q.Z., Liang, W., Fu, Q., Li, W., Huang, K.X., 2012a. Origin 2001. Geochemical evidence for the melting of subducting oceanic lithosphere at of Late Oligocene adakitic intrusives in the southeastern Lhasa terrane, evidence from plate edges. Nature 409, 500–504. in situ zircon U–Pb dating: Hf–O isotopes, and whole-rock geochemistry. Lithos 148, Zeng, L.S., Gao, L.E., Xie, K.J., Zeng, J.L., 2011. Mid–Eocene high Sr/Y granites in the North- 296–311. ern Himalayan gneiss domes: melting thickened lower continental crust. Earth and Zheng, Y.C., Hou, Z.Q., Li, W., Liang, W., Huang, K.X., Li, Q.Y., Sun, Q.Z., Fu, Q., Zhang, S., Planetary Science Letters 303, 251–266. 2012b. Petrogenesis and geological implications of the Oligocene Chongmuda– Zhang, S.Q., Mahoney, J.J., Mo, X.X., Ghazi, A.M., Milani, L., Crawford, A.J., Guo, T.Y., Zhao, Mingze adakite-like intrusions and their mafic enclaves, southern Tibet. Journal of Z.D., 2005. Evidence for a widespread Tethyan upper mantle with Indian-Ocean- Geology 120, 647–669. type isotopic characteristics. Journal of Petrology 46, 829–858. Zhu,D.C.,Zhao,Z.D.,Pan,G.T.,Lee,H.Y.,Kang,Z.Q.,Liao,Z.L.,Wang,L.Q.,Li,G.M., Zhang, H.F., Harris, N., Guo, L., Xu, W.C., 2010a. The significance of Cenozoic magmatism Dong, G.C., Liu, B., 2009a. Early cretaceous subduction-related adakite-like from the western margin of the eastern syntaxis, southeast Tibet. Contributions to rocks of the Gangdese Belt, southern Tibet: products of slab melting and subse- Mineralogy and Petrology 160, 83–98. quent melt–peridotite interaction? Journal of Asian Earth Sciences 34, Zhang, Z.M., Zhao, G.C., Santosh, M., Wang, J.L., Dong, X., Shen, K., 2010b. Late Cretaceous 298–309. charnockite with adakitic affinities from the Gangdese batholith, southeastern Tibet: ev- Zhu, D.C., Mo, X.X., Niu, Y.L., Zhao, Z.D., Wang, L.Q., Liu, Y.S., Wu, F.Y., 2009b. Geochemical idence for Neo-Tethyan mid-ocean ridge subduction? Gondwana Research 17, 615–631. investigation of early Cretaceous igneous rocks along an east–west traverse through- Zhang, Z.M., Zhao, G.C., Santosh, M., Wang, J.L., Dong, X., Liou, J.G., 2010c. Two stages of out the central Lhasa Terrane, Tibet. Chemical Geology 268, 298–312. granulite facies metamorphism in the eastern Himalayan syntaxis, south Tibet: pe- Zhu, D.C., Zhao, Z.D., Niu, Y.L., Mo, X.X., Chung, S.L., Hou, Z.Q., Wang, L.Q., Wu, F.Y., 2011. trology, zircon geochronology and implications for the subduction of Neo-Tethys The Lhasa Terrane: record of a microcontinent and its histories of drift and growth. and the Indian continent beneath Asia. Journal of Metamorphic Geology 28, 719–733. Earth and Planetary Science Letters 301, 241–255.