Petrology and Geochemistry of the Early Mesozoic Pyroxene Andesites in the Maixiu Area, West Qinling, China: Products of Subduction Or Syn-Collision?

Lithos 172–173 (2013) 158–174

Contents lists available at SciVerse ScienceDirect

Lithos

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

Petrology and geochemistry of the early Mesozoic pyroxene andesites in the Maixiu Area, West Qinling, China: Products of subduction or syn-collision?

Xiao-Wei Li a,b,⁎, Xuan-Xue Mo a,⁎⁎, Xue-Hui Yu a, Yi Ding a, Xiong-Fei Huang a, Ping Wei a, Wen-Yan He a a State Key Laboratory of Geological Processes and Mineral Resources, School of Earth Sciences and Resources, China University of Geosciences, Beijing 100083, PR China b School of Earth and Space Sciences, Peking University, Beijing 100871, PR China article info abstract

Article history: The Qinling–Dabie–Sulu Orogen is dotted with Mesozoic igneous rocks and its exact tectonic origin is still Received 13 September 2012 controversial, especially the precise timing of initial collision between the North China Block (NCB) and the Accepted 10 April 2013 Yangtze Block (YB) and the subsequent closure of the Paleo-Tethys Ocean in Qinling, China. This paper Available online 20 April 2013 presents geochronological and geochemical data for pyroxene andesites in Maixiu area, West Qinling. Laser fusion 40Ar/39Ar dating for matrix glass yields an isochron age of 234 ± 3 Ma. The Maixiu pyroxene andesites Keywords: (MPAs) display a hyalopilitic texture, and the predominant phenocryst phases are plagioclase, orthopyroxene High-Mg andesite Laser fusion 40Ar/39Ar dating and clinopyroxene. Orthopyroxene generally displays delicately normal zoning, whereas some clinopyroxene In-situ trace elements grains exhibit reverse zonings. Textural relations indicate that magma mixing plays a key role for the genesis Triassic of the MPAs. The MPAs, with 53.75–57.29 wt.% SiO2, 0.6–0.82 wt.% TiO2 and 48–72 Mg#, are characterized by Tectonic implication high magnesium contents in some samples. The MPAs display enriched light rare earth elements (LREEs) and West Qinling, China relatively high (La/Yb)N ratios (5–9). Clinopyroxene phenocrysts are depleted in some HFSE (e.g., Nb, Zr, Hf, and Ti) and some LILE (i.e., Ba, K and Sr), and are enriched in some other HFSE (e.g., Th and U), REE (e.g., Nd

and Sm) and some other LILE (e.g., Rb and Pb). The MPAs have uniformly low εNd(t) values (−7.74 to −9.27) 87 86 and high ( Sr/ Sr)t ratios (0.70788 to 0.71225), implying a continental rather than oceanic type magma source. Based on data for clinopyroxene phenocrysts, we estimate a temperature range of 956 to 1087 °C with the mean value of 1032 ± 39 °C (1σ), and a pressure range from 5.9 to 13.6 kbar with an average of 9.8 ± 1.9 kbar (1σ). We conclude that the petrogenesis of the MPAs in West Qinling Orogen may have involved magma mixing between melts derived from the sedimentary cover of the northward-subducting A'nyemaqen–Mianlue oceanic slab and peridotite-derived basaltic melts from the overriding mantle wedge during the initial collision stage between the NCB and the YB. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Graham and Cole, 1991; Tiepolo et al., 2011); (3) partial melting of the hydrated mantle wedge peridotite (Kelemen, 1995). In particular, Andesite is the second most common volcanic rock type on Earth the fluid dehydration from water-bearing minerals (such as carpholites, and provides abundant information about the interaction between amphiboles, and chlorites), plays a key role in the genesis of arc the mantle and crust in the subduction zones (Grove and Kinzler, andesite, which are enriched in large ion lithophile elements (LILE) 1986). However, the petrogenesis of subduction-related andesite pet- and light rare earth elements (LREE), and depletion of high field rogenesis is being debated (Boettcher, 1973; Grove and Kinzler, 1986; strength elements (HFSE) (Chiaradia et al., 2011). Moreover, andesites Reubi and Blundy, 2009; Tatsumi and Eggins, 1995; Thorpe, 1982), and coeval basalt–dacite–rhyolite volcanic suites can serve as excellent since andesite can form via different processes, such as (1) magma proxies for the paleotectonic and paleogeographical reconstruction, mixing between felsic and mafic/ultramafic melt (Reubi and Blundy, especially in ancient orogens that underwent multiple deformation 2009; Streck et al., 2007); (2) fractional melting or assimilation stages (Bailey, 1981). fractional crystallization (AFC) from basaltic composition (Gill, 1981; Stretching across the central part of China, the Qinling–Dabie–Sulu Orogen (Central Orogen) marks the final amalgamation of the North China Block (NCB) and Yangtze Block (YB) (Fig. 1A and B). It is generally ⁎ Correspondence to: X.-W. Li, State Key Laboratory of Geological Processes and Mineral accepted that the Qinling Orogenic belt can be divided into two parts Resources, School of Earth Sciences and Resources, China University of Geosciences, along the Baoji–Chengdu railway (Meng and Zhang, 2000; Zhang et Beijing 100083, PR China. Tel.: +86 10 6274 2179. al., 2001, 2007; Fig. 1C), namely the East Qinling Orogenic Belt and the ⁎⁎ Corresponding author. Tel.: +86 10 6230 6299. E-mail addresses: [email protected] (X.-W. Li), [email protected] West Qinling Orogenic Belt. During the past four decades, much re- (X.-X. Mo). search was done in the eastern part of this orogen (Dong et al., 2011;

Fig. 1. A and B. Subdivision of the major tectonic units in China showing the location of the Qinling Orogenic Belt. C. Simpliﬁed regional geological map of the West Qinling, showing distribution of the Mesozoic granitoids in the West Qinling Orogen (modiﬁed from Feng et al., 2002; Zhang et al., 2006, 2007). Abbreviations in Fig. 1A: CB: Cathaysia Block; KL: Kunlun; LB: Lhasa Block; NCB: North China Block; QB: Qiangtang Block; QD: Qaidam; QL: Qilian; QOB: Qinling Orogenic Belt; YB: Yangtze Block. Pluton names and age data sources in Fig. 1C: Wenquan (218 Ma) and Heimahe (235 Ma) are from Zhang et al. (2006); Zhongchuan (232–236 Ma) is from Zhu et al. (in press); Daerzang (238 Ma) and Meiwu (245 Ma) are from Jin et al. (2005); Tongren (240 Ma) are from Feng et al. (2002); Guangtoushan (216 Ma) is from Sun et al. (2000); Miba (220 Ma) is from Sun et al. (2002); Mishuling (213 Ma) is from Qin et al. (2009); Wenquan (216.2 Ma) is from Zhu et al. (2011); and Gangcha (234–243 Ma) is from Guo et al. (2011, 2012) and Luo et al. (2012). D. Geological map of the Maixiu area, West Qinling.

Li et al., 1978; Mattauer et al., 1985). However, knowledge on the 2. Geological background western counterpart is still cryptic till now (Zheng et al., 2010), and this largely restricts our understanding of the general geological evolu- West Qinling is the westward extension of East Qinling and is tion of the entire orogen. Speciﬁcally, the precise time of tectonic tran- bounded by Qilian terrane to the north, Qaidam terrane to the west, sition from the Mianlue Paleo-ocean's northward subduction to the and the Songpan–Garzê terrane to the south (Ratschbacher et al., Yangtze Block–North China Block collision is still debated due to a lack 2003; Zhang et al., 2001; Zheng et al., 2010)(Fig. 1 A and B), Moreover, of geochronological data. For example, several authors considered the the A'nimaque–Mianlue suture zone along the southern margin of West early Indosinian granitoids in the Central Orogen as products of a Qinling is interpreted to represent a fossil ocean, belonging to one post-collisional regime (Luo et al., 2012; Zhang et al., 2006). However, branch of the Paleo-Tethys, and extends to the Buqingshan–A'nimaque Jin et al. (2005) and Guo et al. (2011, 2012) contended that the coeval paleo-ocean in the Eastern Kunlun Orogen. adakite-like granitoids occurring in the same tectonic unit (Fig. 1)are West Qinling is interpreted to be a microcontinental block that consistent with an active destructive margin regime during the early originally split from the NCB, migrated towards the northern YB Indosinian in West Qinling. during the Meso-Neoproterozoic, and then collided with the NCB Here, we report, the geochronology, mineral chemistry, whole- during the Triassic (Dong et al., 2011; Meng et al., 2005; Zheng et rock major-trace element chemistry, and Sr–Nd isotopic data to con- al., 2010). As the main component of the Foping metamorphic terrane, strain the timing of volcanism and petrogenesis of the Mesozoic vol- the oldest exposed crystalline basement in West Qinling is the Qinling canic rocks from the Maixiu area, West Qinling. Our main objective Group, which predominantly consists of gneisses, amphibolites and is to develop a paleotectonic model that is consistent with the coeval marbles, with U–Pb ages ranging from 2172 to 2267 Ma (Dong et al., intrusive rocks in the same regions (e.g. Gangcha complex and Daerzang 2011; Meng and Zhang, 2000). However, the well exposed stratigraphic pluton; Fig. 1). sequence in our study area is the Upper Permian Maomaolong 160 X.-W. Li et al. / Lithos 172–173 (2013) 158–174

Table 1 The Ar/Ar dating for matrix glass separated from Maixiu pyroxene andesite sample ZK10-1.

Run ID 40Ar* (%) Age (Ma) ±1σ 39Ar (Moles) 40Ar ±1σ 39Ar ±1σ 38Ar ±1σ 37Ar ±1σ 36Ar ±1σ

J = 0.00495626 ± 0.0000263 2062-01 28.40 219.26 2.48 0.27 324.43 0.10 3.54 0.00 0.23 0.00 4.89 0.07 0.79 0.00 2062-02 36.14 227.52 1.87 0.36 351.39 0.10 4.69 0.00 0.22 0.00 6.58 0.07 0.76 0.00 2062-03 21.23 218.49 3.45 0.23 374.18 0.14 3.06 0.00 0.24 0.00 3.16 0.06 1.00 0.00 2062-04 27.58 219.84 2.58 0.30 367.70 0.22 3.88 0.01 0.25 0.00 8.10 0.07 0.90 0.00 2062-05 46.55 229.00 1.37 0.38 286.71 0.09 4.89 0.01 0.18 0.00 8.44 0.06 0.52 0.00 2062-06 39.25 227.66 1.73 0.32 286.32 0.10 4.15 0.00 0.19 0.00 8.06 0.06 0.59 0.00 2062-07 25.69 215.33 2.64 0.39 506.47 0.12 5.09 0.00 0.33 0.00 5.15 0.06 1.28 0.00 2062-08 30.99 223.68 2.26 0.29 321.89 0.10 3.75 0.00 0.20 0.00 5.48 0.06 0.75 0.00 2062-09 20.39 212.01 3.42 0.34 538.96 0.12 4.37 0.01 0.35 0.00 5.31 0.07 1.45 0.00 2062-10 34.91 221.75 1.86 0.41 406.53 0.08 5.38 0.00 0.26 0.00 5.97 0.07 0.90 0.00 2062-11 24.96 220.91 2.77 0.45 614.03 0.15 5.84 0.01 0.41 0.00 8.50 0.07 1.56 0.00 2062-12 27.44 225.58 2.61 0.28 351.74 0.10 3.60 0.00 0.24 0.00 5.23 0.06 0.87 0.00 2062-13 22.32 216.99 3.21 0.29 438.38 0.16 3.80 0.00 0.29 0.00 4.76 0.06 1.15 0.00 2062-14 45.47 229.51 1.37 0.44 341.27 0.10 5.67 0.01 0.23 0.00 8.20 0.07 0.63 0.00 2062-15 39.29 221.33 1.60 0.54 468.28 0.24 6.99 0.01 0.30 0.00 8.78 0.07 0.96 0.00 2062-16 32.70 220.19 2.02 0.40 419.12 0.12 5.24 0.01 0.28 0.00 6.24 0.06 0.96 0.00 2062-17 28.53 217.70 2.50 0.46 544.89 0.62 6.01 0.01 0.35 0.00 6.56 0.06 1.32 0.00

Formation (Fig. 1C), which mainly consists of sandstone, siltstone, argil- 560 laceous slate, shale, limestone and brecciated limestone (Li, 2011; Zhao A 6 and Yang, 1992). The Triassic stratigraphy has been investigated in 540 16 West Qinling (Meng et al., 2007). The Lower Triassic Longwuhe Group is featured by siliciclastic rocks (Meng et al., 2007), and the overlying 520 conformable stratum is Gulangdi Formation (middle Triassic), which is mainly composed of sandstone interlayered with limestone and 500 siltstone. The pyroxene andesites analyzed in this study are exposed 7 17 480

within the Maixiu Volcanic Basin near Duofuntun Town (Fig. 1D), Ar where they have a total thickness of ca. 2200–3400 m. The MPAs belong 36 460 2 11 to Maixiu Group, which occurs unconformably over the Gulangdi Ar/

40 Formation, and consists of fluvial-lacustrine sandstone, carbonaceous 440 20 slate, and intermediate-acidic volcanic rocks. This study focused on a 9 420 suite of pyroxene andesite (i.e. MPAs) in the lower part of the Maixiu 1 21 13 Group. The upper most part of the Maixiu Group, containing late- 5 400 8 Triassic flora fossils (QBGMR, 1991), consists of amphibole andesites 12 Age=234 ± 3 Ma intercalated with coal seams. 380 14 4 MSWD=1.2(n=17) 10 360 3. Analytical methods 2.5 3.5 4.55.5 6.5 7.5 8.5 9.5 39Ar/36Ar Sample ZK10-1 was analyzed by Automatic 40Ar/39Ar Laser-Probe dating method for matrix glass at Peking University. The sample was fi fi 0.0027 10 rst crushed, then the phenocrysts and amygdule llings were picked 4 B out under a binocular microscope, and sieve fraction (60–80 μm) was 14 0.0026 separated from the sample. The sample was wrapped in Al foil tubes, 12 8 and was then sealed into the evacuated quartz bottle. Then the sam- 0.0025 5 ple was irradiated in the 49-2 Nuclear Reactor at the Institute of 13 1 21 Chinese Atomic Energy. Neutron flux variation (J) was measured 0.0024 using the standard mineral Zhoukoudian biotite (ZBH-25, Age: 9 132.7 Ma). The detail procedures of the Ar isotope analyses are simi- Ar 0.0023 40 20 lar to those reported by Hall and Farrell (1995) and Zhu et al. (2007). Ar/ 0.0022 11 All analyses were corrected for fusion-system blank levels at the five Ar 36 2 mass positions; blanks were run after each three analyses. Blank levels −16 40 −17 0.0021 were approximately 3.9 × 10 mol for Ar and 1.8 × 10 mol 17 36 7 for Ar. The data processing is similar to that described by Nomade et 0.0020 al. (2005). Microprobe analyses were carried out on polished thin section 0.0019 Age=234 ± 5 Ma MSWD=2.1(n=17) using JXA-8100 electron microprobe at Peking University in a wave 16 length-dispersive mode with 15 kV acceleration potential and a 0.0018 6 0.008 0.010 0.012 0.014 0.016 10 nA beam current. Backscattered electron images and a subset of 39 40 the clinopyroxene and orthopyroxene data were obtained on a Ar/ Ar JXA-8230 electron microprobe at the Chinese Academy of Geological Fig. 2. (A) Ar–Ar Isochron for glass matrix separated from Maixiu pyroxene andesite Sciences in a wave length-dispersive mode with 15 kV acceleration sample ZK10-1. (B) Inverse isochron plot for glass matrix separated from Maixiu potential and a 20 nA beam current. pyroxene andesite sample ZK10-1. X.-W. Li et al. / Lithos 172–173 (2013) 158–174 161

In-situ trace element analyses were performed by LA–ICP–MS at mesh for major, trace elements and Sr–Nd isotopic analyses. Then Loss the Key Laboratory of Orogenic Belts and Crustal Evolution at Peking on ignition was determined by heating ca. 0.5 g of rock powder in plat- University. The laser ablation ICP–MS system consists of a 193 nm inum crucibles in a furnace for 1 h at 1000 °C. Major element analyses ArF-excimer laser coupled to an Agilent 7500ce ICP–MS. Helium were determined by inductive coupled plasma–atomic emission spec- was served as the carrier gas to maintain stable and optimum excita- troscopy (ICP–AES) at China University of Geosciences, Beijing. The tion conditions. A spot size of 60 μm, with a repetition rate of 5 Hz operating procedures were described by Song et al. (2010). and a laser energy of 100 mJ, was used in this study. The background Whole rock trace element analyses were measured by inductively acquisition (gas blank) time and the data acquisition interval were coupled plasma–mass spectrometry (ICP–MS) at the State Key 20 s (including 5 s pre-ablation) and 60 s, respectively. The glass Laboratory of Geological Processes and Mineral Resources, China standards, NIST 610, were employed as an external calibration stan- University of Geosciences, Wuhan. The operating procedures were dard, with NIST 612 and 614 serving as monitoring standards at the described at length by Liu et al. (2008). same time. Silicon (29Si) was selected as the internal standard. Sr–Nd isotopic analyses were carried out on a Thermo-Finnigan Reference values of NIST 610, 612 and 614 were taken from GeoREM TRITON® at Tianjin Institute of Geology and Mineral Resources. The (http://georem.mpch-mainz.gwdg.de). The data were reduced using Sr and Nd isotopic ratios were normalized to 86Sr/88Sr = 0.1194 the GLITTER 4.4.1 developed by GEMOC, Macquarie University (van and 146Nd/144Nd = 0.7219, respectively. The detailed analytical Achterbergh et al., 2001). procedures were given in Niu et al. (2012). A subset of our data was The least-altered rock samples were ﬁrstly crushed into coarse, determined using a Finnigan MAT-261 mass spectrometer at China freshly broken chips 0.5–1 cm in diameter, which were rinsed in dis- University of Geosciences (Wuhan) and the detailed analytical proce- tilled water and dried, and then grounded in agate mills to about 200 dures were given in Liu et al. (2004) and Zhao et al. (2009a).

A D

Cpx Pl

B E

Cpx

C F

Opx Pl

Fig. 3. Photomicrographs of characteristic petrographic features of the MPAs. (A) Sieve texture of plagioclase in the MPAs; (B) crossed twinning of plagioclase in the MPAs; (C) oscillatory zoning of plagioclase in the MPAs; (D) plagioclase inclusions are surrounded by clinopyroxene; (E) clinopyroxene with an irregular corrosion rim; (F) orthopyroxene phenocryst. Abbreviations: Pl, plagioclase; Cpx, clinopyroxene; Opx, orthopyroxene. 162 X.-W. Li et al. / Lithos 172–173 (2013) 158–174

4. Results with An68–83 (Fig. 4A, Appendix 1). Most phenocrysts display crossed twinning (Fig. 3B) but some possess delicate oscillatory zoning 4.1. 40Ar–39Ar isotope dating (Fig. 3C) and simple twinning. In contrast, plagioclase in the matrix

shows relatively low CaO with An60–81 (Fig. 4,Appendix1). The 40Ar–39Ar isotopic data for matrix glass separated from sample Although less abundant than plagioclase phenocrysts, clinopyroxene ZK10-1 are listed in Table 1. These isochron age data yield a weighed is another common phenocryst phase in the basaltic andesites. Some mean age of 234 ± 3 Ma (Fig. 2A;MSWD=1.2,n=17)withan clinopyroxene grains have been replaced by chlorite, and often have initial ratio of 287.9 ± 1.5, which is slightly below the Nier value (Nier, plagioclase inclusions (Fig. 3D). Clinopyroxene phenocrysts are usually 1950; Steiger and Jäger, 1977). This age indicates that the Maixiu pyrox- surrounded by an irregular corrosion rim (Fig. 3E), which may be a ene andesites erupted during the early stage of the late Triassic, coincid- result of rapid decompression (remelting) in the volcanic conduits. ing with the stratigraphic succession in the field. In the inverse isochron, Clinopyroxene phenocrysts from basaltic andesite shows variable SiO2 36 40 39 40 Ar/ Ar is plotted against Ar/ Ar for each analysis (Fig. 2B), giving an (50.59–53.64 wt.%), Al2O3 (1.47–4.27 wt.%), MgO (14.47–18.07 wt.%) age of 234 ± 5 Ma, which is consistent with the isochron age above. contents and Mg# (73–91) values, but the CaO (18.85–23.12 wt.%) contents seem to be relatively constant (Appendix 2), whereas the corrosion 4.2. Petrography and mineral chemistry rims have relatively low contents of CaO (3.92–4.43 wt.%), and high contents of FeO (16.84–21.68 wt.%) and MgO (19.81–22.97 wt.%) The volcanic rocks in the Maixiu area are classified as pyroxene (Appendix 2). Most clinopyroxene grains do not exhibit compositional andesites and display a hyalopilitic texture. Plagioclase (30–40 vol.%; zoning. However, some reversely zoned clinopyroxene crystals are also Fig. 3A–C) and clinopyroxene (10–15 vol.%; Fig. 3D–E) are the observed (Fig. 5A). Except for two clinopyroxene crystals belonging dominant phenocryst phases, with minor orthopyroxene (0–5vol.%; to diopside, all the other clinopyroxene crystals fall in the augite field Fig. 3F), biotite (0–3 vol.%), quartz (2–5 vol.%), and Fe–Ti oxides (1– near the endiopside/augite boundary, with a compositional range of

5 vol.%). Plagioclase in the basaltic andesite is present both as pheno- Wo38–47 En42–50 Fs6–17 (Fig. 5C, Appendix 2). Orthopyroxene phenocrysts and within the groundmass, with phenocrysts having cores crysts are mainly present in andesite samples with high magnesium with a pronounced sieve texture (Fig. 3A) overgrown by clean rims, contents (Fig. 3F). Orthopyroxene phenocrysts generally exhibit nor- which is generally interpreted as a result of rapid decompression mal compositional zonings (Fig. 5B; Appendix 3). They are character- (Nelson and Montana, 1992). The ubiquitous appearance of plagioclase ized by a uniform magnesian interior (Mg# = 86–92) surrounded by phenocrysts enclosed by clinopyroxenes indicates that plagioclase a very thin and iron-rich rim (Mg# = 83–84; Fig. 5B; Appendix 3). appears to have been a near-liquidus phase, which in turn implies Nearly all of the orthopyroxene crystals fall in the bronzite ﬁeld. less than 2 to 5 wt.% water content in the magma prior to eruption Meanwhile, some orthopyroxene crystals can be classiﬁed as enstatite (Gill, 1981). Plagioclase phenocrysts occur as euhedral to subhedral (Fig. 5C, Appendix 3). Fe–Ti oxides are disseminated in the ground- crystals, and they are mainly composed of labradorite and bytownite mass, and scarcely occur as microphenocrysts.

4.3. Whole-rock major and trace elements A Or Whole-rock major element compositions of representative volcanic rocks are listed in Table 2. All the samples in this study are mod- erately altered, as suggested by high and variable loss-on-ignition (LOI) values of 3.92–6.63 wt.%, probably reflecting the chloritization of clinopyroxene and the sericitization of plagioclase, which are consistent with field and petrographic observations. However, no sys- tematic correlations exist between LOI and fluid mobile elements, such as K, Na, Rb, Sr, and Ba. Further, a good linear regression is Sanidine observed between Th and U, implying the limited influence on changing the valence of uranium during the post-magmatic process. Consequently, the magmatic process is still the first-order control on the compositional variations for the MPAs. The MPAs are dominated

by basaltic andesite and andesite in the Zr/TiO2–Nb/Y diagram (Fig. 6). The pyroxene andesites show a wide range of major element contents

(Fig. 7)(SiO2 =53.75–57.29 wt.%; TiO2 =0.62–0.8 wt.%; Al2O3 = Matrix Phenocryst Anorthoclase 14.96–18.46 wt.%; Fe2O3T = 5.90–8.95 wt.%; MgO = 2.78–8.05 wt.%) Albite Oligoclase Andesine Labradorite Bytownite and Mg# = 48–72. Notably, considerably positive correlations are Anorthite observed between Al2O3 plus TiO2 and MgO (Fig. 7). Ab An The high Mg# values for some samples are higher than those of the B experimental melts (Rapp and Watson, 1995), suggesting the contribu- 6 tion from mantle-derived components in their source region. All the ﬁ matrix samples could be classi ed into two groups according to the SiO2 versus MgO diagram of McCarron and Smellie (1998) (Fig. 8A). One group 4 phenocrysts belongs to high magnesian andesites (HMA) with MgO > 6 wt.% (Kelemen et al., 2003). This group resembles HMAs from elsewhere ﬁ counts around the globe (Fig. 8). The other group can be classi ed as low-Mg 2 andesite (LMA; MgO b 6 wt.%), similar to normal arc andesites (Fig. 8A and B). The partial melting from the hydrous oceanic slab alone can- 0 not account for the origin of the MPAs, as is illustrated in Figs. 8AandB. 60 62 64 66 68 70 72 74 76 78 80 82 84 Chondrite-normalized (REE) patterns of MPAs are relatively smooth Fig. 4. (A) Or–Ab–An ternary diagram showing plagioclase compositions of the MPAs with shallow negative slopes (Fig. 9A) and mean (La/Yb)N ratios of in West Qinling. (B) Histogram of plagioclase anorthite contents for the MPAs. 6.94 ± 1.02 (1σ) (where N means chondrite-normalized). The heavy X.-W. Li et al. / Lithos 172–173 (2013) 158–174 163

A 100 µm B 200 µm

Cpx

Opx 90

77 84 87 84

C Diopside Hedenbergite

Salite Ferrosalite

Augite Ferroaugite

Endiopside

SubcalcicAugite Subcalcic Ferroaugite

Pigeonite

Hypersthene

Enstatite Bronzite

clinopyroxene orthopyroxene

Fig. 5. A and B Backscattered electron images of clinopyroxene (reverse zoning) and orthopyroxene (normal zoning); (C) CaSiO3–MgSiO3–FeSiO3 plot showing the compositions of pyroxene (Morimoto et al., 1988) for the MPAs in West Qinling. earth rare elements (HREE) patterns are similar to those of the lower implying considerable contribution from the mantle endmember. In crust, whereas the light rare earth elements (LREE) concentrations are contrast, the high Th/La ratios from the MPAs with the mean value of higher than those of the lower crust. The pyroxene andesites generally 0.33 ± 0.06 (2σ) are almost identical with the average upper conti- reveal weakly negative to positive Eu anomalies (Eu/Eu* = 0.72– nental crust value of 0.33 ± 0.05 (2σ)(Plank, 2005), suggesting the 1.04). More speciﬁcally, the LMA exhibits more pronounced REE enrich- substantial involvement of the upper continental crust and/or its ment than the HMA, and the LMA have relatively lower contents of equivalents (sediments) into magma generation. compatible elements than those of the HMA (e.g. Ni). Primitive mantle normalized trace element patterns are shown in 4.4. In-situ trace element contents of clinopyroxene phenocrysts Fig. 9B. All the samples are collectively depleted in Nb, Ta and Ti and enriched in Th, U and Pb. Some HFSE concentrations, such as Th and In-situ trace element compositions for clinopyroxene phenocrysts U, are systematically lower than those of the lower crust. The MPAs are listed in Appendix 4. REE and spider diagrams for pyroxene andes- have relatively high Cr (133–845 ppm) and Ni (59–136 ppm) contents, ites are shown in Figs. 10A and B. Clinopyroxene phenocrysts display 164 X.-W. Li et al. / Lithos 172–173 (2013) 158–174

Table 2 Major (wt.%), trace element (ppm), and Sr–Nd isotopic compositions of selected samples from the MPAs.

Sample name ZK10-1 ZK10-2 ZK10-3 ZK10-36 ZK11-03 ZK11-04 ZK11-06 ZK11-09 ZK11-12 ZK11-15 ZK11-18 ZK11-20

Rock type Pyroxene Pyroxene Pyroxene Pyroxene Pyroxene Pyroxene Pyroxene Pyroxene Pyroxene Pyroxene Pyroxene Pyroxene andesite andesite andesite andesite andesite andesite andesite andesite andesite andesite andesite andesite

SiO2 53.75 56.96 56.22 57.29 55.78 56.31 55.08 55.64 n.d. 55.16 n.d. 55.58

TiO2 0.62 0.82 0.71 0.82 0.69 0.64 0.73 0.74 n.d. 0.60 n.d. 0.72

Al2O3 15.34 18.46 17.40 17.94 15.99 15.90 17.53 17.88 n.d. 14.96 n.d. 18.01 T Fe2O3 7.56 6.05 5.90 5.96 8.95 7.06 6.82 6.58 n.d. 7.45 n.d. 6.54 MnO 0.09 0.06 0.10 0.04 0.05 0.09 0.05 0.06 n.d. 0.09 n.d. 0.05 MgO 6.39 2.89 4.64 3.01 6.14 8.05 4.53 3.59 n.d. 6.88 n.d. 2.78 CaO 7.06 6.60 5.89 6.06 2.10 3.12 6.31 8.09 n.d. 7.94 n.d. 8.62

Na2O 1.81 2.60 2.53 2.44 1.90 2.76 2.29 2.34 n.d. 1.67 n.d. 2.45

K2O 0.71 1.90 1.72 1.19 2.05 0.77 1.83 1.69 n.d. 0.28 n.d. 1.24

P2O5 0.17 0.02 0.07 0.05 0.06 0.10 0.09 0.10 n.d. 0.09 n.d. 0.08 LOI 6.63 3.92 4.01 4.40 6.37 5.89 4.88 4.11 n.d. 5.10 n.d. 4.80 Total 100.13 100.29 99.20 99.20 100.06 100.69 100.15 100.82 n.d. 100.21 n.d. 100.87 Mg# 65 51 63 53 60 72 59 55 n.d. 67 n.d. 48 Li 46.4 67.1 25.4 34.3 40.2 50.2 41.6 20.7 40.6 17.2 107 25.2 Be 1.28 2.01 1.65 1.04 1.24 1.28 1.45 1.59 1.24 1.01 1.36 1.49 Sc 23.1 15.9 16.0 29.3 23.0 21.2 21.9 19.9 19.6 24.2 18.7 22.5 V 141 108 113 176 126 112 121 126 131 137 81.5 133 Cr 288 133 258 491 283 253 288 291 845 797 236 410 Co 32.7 26.0 22.1 36.5 30.9 26.6 28.8 28.5 43.5 35.1 25.5 34.4 Ni 106 59.0 78.5 72.0 79.8 65.2 103 106 136 106 70.1 72.8 Cu 95.6 50.7 56.0 32.7 40.2 25.2 40.9 46.1 32.6 33.5 22.6 25.7 Zn 68.5 68.7 44.9 71.0 70.2 64.7 67.3 72.6 86.3 69.7 55.1 78.5 Ga 17.0 21.0 17.2 16.4 17.3 16.8 18.7 19.4 17.8 16.3 15.3 18.8 Rb 20.7 32.6 124 25.4 91.9 32.4 68.4 63.4 21.4 7.79 62.6 45.2 Sr 185 252 144 240 60.7 170 201 251 254 260 61.0 341 Y 17.2 22.7 16.7 19.0 13.0 15.2 15.0 16.2 13.9 15.6 16.3 14.7 Zr 99.5 142 113 110 104 96.6 109 108 99.6 93.9 99.6 106 Nb 4.36 7.15 5.45 5.26 5.43 4.94 5.17 5.34 5.26 4.94 5.68 5.50 Cs 12.4 25.1 6.13 7.24 10.8 18.4 17.8 7.56 21.7 16.1 18.5 13.2 Ba 186 166 113 251 76.6 190 305 323 170 142 106 273 La 14.4 23.9 21.7 15.0 15.0 13.9 15.5 15.5 12.4 13.4 15.5 15.8 Ce 32.1 48.2 42.1 33.6 29.2 29.3 31.4 31.3 24.0 27.2 30.8 31.6 Pr 3.76 5.43 4.74 3.95 3.18 3.47 3.70 3.71 2.78 3.17 3.49 3.67 Nd 14.8 21.1 17.5 15.8 11.9 13.5 14.6 14.1 10.3 12.6 13.4 14.0 Sm 3.26 4.41 3.38 3.39 2.11 2.58 2.99 2.99 2.07 2.76 2.61 2.72 Eu 0.80 1.02 0.93 0.91 0.50 0.72 0.83 0.92 0.74 0.73 0.65 0.84 Gd 3.20 4.12 3.30 3.31 2.00 2.58 2.95 2.99 2.23 2.55 2.69 2.85 Tb 0.50 0.65 0.54 0.60 0.32 0.41 0.44 0.47 0.36 0.42 0.43 0.43 Dy 3.15 3.90 3.21 3.56 2.19 2.58 2.75 2.84 2.35 2.57 2.80 2.75 Ho 0.62 0.76 0.65 0.73 0.48 0.53 0.52 0.56 0.48 0.52 0.56 0.53 Er 1.82 2.20 1.76 2.00 1.40 1.43 1.48 1.56 1.41 1.56 1.66 1.56 Tm 0.28 0.32 0.27 0.32 0.23 0.24 0.22 0.24 0.20 0.24 0.25 0.24 Yb 1.81 2.09 1.66 1.97 1.53 1.42 1.53 1.61 1.49 1.50 1.77 1.50 Lu 0.25 0.30 0.25 0.30 0.24 0.24 0.23 0.22 0.22 0.24 0.23 0.23 Hf 2.63 3.80 3.09 2.84 2.93 2.74 2.98 2.98 2.77 2.61 2.73 2.88 Ta 0.32 0.54 0.42 0.37 0.39 0.34 0.36 0.36 0.37 0.36 0.40 0.40 Pb 12.1 19.9 12.5 9.14 9.54 10.4 17.2 18.6 13.8 10.6 11.2 19.5 Th 4.09 8.04 6.46 4.52 4.76 4.89 4.72 4.64 4.60 4.70 5.49 5.39 U 1.06 2.03 1.57 0.85 1.18 1.19 1.18 1.15 1.05 1.03 1.71 1.20 ΣREE 81 118 102 85 70 73 79 79 61 69 77 79

(La/Yb)N 689577776668 Eu/Eu* 0.75 0.72 0.84 0.82 0.73 0.84 0.84 0.93 1.04 0.83 0.74 0.91 87Rb/86Sr 0.33370 0.38562 0.25628 0.31448 4.51131 0.56851 1.01260 0.75218 0.25098 0.08921 3.05968 0.39430 87Sr/86Sr 0.71037 0.71329 0.71311 0.71065 0.72289 0.71106 0.71271 0.71185 0.71052 0.71018 0.71870 0.71128 ±2σ 45654484194712 87 86 ( Sr/ Sr)t 0.70926 0.71200 0.71225 0.70961 0.70788 0.70916 0.70933 0.70934 0.70968 0.70988 0.70852 0.70997 147Sm/144Nd 0.13305 0.12636 0.11677 0.12994 0.10700 0.11603 0.12398 0.12833 0.12168 0.13227 0.11794 0.11745 143Nd/144Nd 0.51208 0.51207 0.51207 0.51206 0.51208 0.51211 0.51211 0.51212 0.51213 0.51210 0.51206 0.51209 ±2σ 75673185310322 143 144 ( Nd/ Nd)t 0.51187 0.51187 0.51190 0.51186 0.51191 0.51193 0.51192 0.51192 0.51194 0.51189 0.51188 0.51191 εNd(t) −9.0 −9.0 −8.6 −9.3 −8.3 −8.0 −8.1 −8.1 −7.7 −8.7 −8.9 −8.3

TDM (Ga) 2.02 1.88 1.69 1.98 1.53 1.63 1.76 1.84 1.69 1.97 1.73 1.67

T 2+ 3+ 87 86 87 86 Fe2O3, total iron as Fe2O3. Mg# = Mg/(Mg + Fe ), assuming Fe /Fe = 0.1, molar ratio, Eu/Eu* = EuN ∗ 2 / (SmN + GdN). n.d. not determined ( Sr/ Sr)t =( Sr/ Sr)sample − 87 86 λt 143 144 143 144 147 144 λt 143 144 147 144 λt ( Rb/ Sr)sample ×(e − 1), ( Nd/ Nd)t =( Nd/ Nd)sample − ( Sm/ Nd)sample ×(e − 1), εNd(t) = {[( Nd/ Nd)sample − ( Sm/ Nd)sample ×(e − 1)] / 143 144 147 144 λt 143 144 147 144 [( Nd/ Nd)CHUR(0) − ( Sm/ Nd)CHUR(0) ×(e − 1)] − 1} × 10,000, TDM =1/λ ×ln{[1+[( Nd/ Nd)sample − 0.51315] / [( Sm/ Nd)sample − 0.2137)]}, where εNd(t) 143 144 147 144 147 144 values were calculated using Nd/ Nd)CHUR(0) = 0.512638 and Sm/ Nd)CHUR(0) = 0.1967, and TDM values were calculated using present-day ( Sm/ Nd)DM =0.2137and 143 144 −1 −1 ( Nd/ Nd)DM = 0.51315, λSm =6.54×10–12 year (Lugmair and Marti, 1978), λRb =1.42×10–11 year (Steiger and Jäger, 1977), t = 234 Ma. roughly ﬂat REE patterns with negative Eu anomalies (Eu/Eu* = 0.55– Moreover, clinopyroxene phenocrysts are depleted in some HFSE (for 0.73). The LREE concentrations are slightly lower than those of example Nb, Zr, Hf, and Ti) and some LILE, i.e., Ba, K and Sr, and are both middle REE and HREE, as shown by (La/Yb)N ratios of 0.47–0.64, enriched in some HFSE (e.g., Th and U) and REE (e.g., Nd and Sm) and (La/Sm)N ratios of 0.32–0.40, and (Sm/Yb)N ratios of 1.29–1.66. some other LILE (e.g., Rb and Pb). X.-W. Li et al. / Lithos 172–173 (2013) 158–174 165

10 of both terrigenous sediments and hydrous oceanic crust (Tatsumi and Hanyu, 2003).

Phonolite 5.3. The nature of the parent magma 1 Com/Pant

The major and trace elements of the primitive parent magma of Rhyolite the MPAs can be estimated using the composition of clinopyroxene. 0.1 Trachyte Partitioning of Fe and Mg between clinopyroxene and melt can be * 0. 0001 Rhyodacite Dacite 2 / Trachy And used to evaluate the MgO/FeO molar ratios of the melt in equilibrium with the clinopyroxene. Putirka (1999) and Putirka et al. (2003)

Zr/TiO Andesite reported the partition coefficient between clinopyroxene and melt 0.01 Bsn/Nph Andesite/Basalt (in molar) as follows: ¼ = ¼ : : : SubAlkaline Basalt Alk-Bas Kd cpx MgliqFecpx MgCpxFeliq 0 275 0 067 0.001 0.01 0.1 1 10 Similarly, Bédard (2007) summarized the exchange coefficient be- Nb/Y tween orthopyroxene and melt for Fe and Mg as follows: Fig. 6. Classification of the MPAs using the Nb/Y–Zr/TiO2 diagram (Winchester and ¼ : – : : Floyd, 1977). ln KdOpx 0 0308 MgO 1 53

Using the most Mg-rich clinopyroxene composition from the 4.5. Sr–Nd isotopes of whole rocks MPAs (Mg# = 91), we computed that the melt in equilibrium with it has a Mg-value of 74, implying that the parent magma of the Whole rock Sr–Nd isotopic compositions for representative sam- MPAs is basaltic. Meanwhile, the melt in equilibrium with the core ples of the basaltic andesites are listed in Table 2. Initial isotopic of orthopyroxene (Mg# = 92) has a Mg-value of 75, whereas the values were calculated at t = 234 Ma on the basis of laser fusion melt in equilibrium with the rim of orthopyroxene (Mg# = 84) has 40Ar/39Ar dating result. The selected samples all display high initial a Mg-value of 58. 87 86 Sr/ Sr isotopic ratios (0.70788–0.71225) and low εNd(t) values Meanwhile, the Na2O and TiO2 contents of parental magmas can (−7.74 to −9.27), similar to some Triassic felsic intrusions in West also be estimated based on the clinopyroxene–liquid partition coefﬁ- Qinling, as reported by Zhang et al. (2006) and Cretaceous Bamco cients as follows (Putirka, 1999): andesites in the northern Gangdese (Chen et al., 2010). The samples yield TDM ages ranging from 1.53 Ga to 2.02 Ga, with an average age ln Na =Na ¼ −2:48 þ 0:11ðÞ PðÞ bars =TKðÞ−5 cpx liq of 1.78 Ga. − 10 7 PðÞ bars 2=TKðÞ 5. Discussions = ¼ : : : Ticpx Tiliq 0 47 0 28 5.1. Timing of the volcanism Assuming that clinopyroxene phenocrysts are in equilibrium with

Our results indicate that the MPAs started forming during the early the parental magma, their Na2O and TiO2 contents can be used to stage of the late Triassic (ca. 234 ± 3 Ma) according to the latest geo- calculate those of the parental magma. The calculated mean values logical timescale (Walker et al., 2013). Combined with the evidence of Na2O and TiO2 for the parental magma are 0.87 ± 0.12 (1σ) wt.% from the local fossil record (QBGMR, 1991), the volcanism terminated and 1.28 ± 0.34 (1σ) wt.% TiO2. no earlier than late-Triassic. The contemporaneous intrusive counter- Trace element abundances of the parental magma deduced from parts (Fig. 1C) include the Xiahe Pluton (238 ± 4 Ma) (Jin et al., the early clinopyroxene phenocrysts can be computed based on the 2005), the Heimahe Pluton (235 ± 2 Ma; Zhang et al., 2006)and composition of individual clinopyroxene and clinopyroxene–liquid altered potassic-diorite from the Gangcha Complex (234 ± 0.6 Ma; partition coefﬁcients (Bédard, 2001; Dunn, 1987; Hart and Dunn, Guo et al., 2012). The slightly older so-called “ophiolite” with a matured 1993; Ross and Elthon, 1993). As shown in Fig. 10A, the calculated arc-signature was identiﬁed recently and is dated at 250.1 ± 2.2 Ma REE patterns of the parental magmas are characterized by LREE

(Wang et al., 2009, 2010). enrichment and ﬂat HREE distributions with [Gd/Yb]N =1.24±0.07 (1σ) and Eu/Eu* = 0.65 ± 0.04 (1σ), similar to that of whole rock 5.2. Pressure and temperature conditions deduced from clinopyroxene REE patterns (except for La). Moreover, the calculated parental melts major elements display trace element patterns typical for arc lavas (Fig. 10B). They are characterized by an enrichment in LILE (Rb, Pb and Ba) and HFSE The physical conditions of parent magmas can be approximately (e.g., Th and U) and REE (e.g., Nd), and a depletion in other HFSE (Nb, deduced from their early crystallizing phases. For example, average Zr, Hf and Ti) and LILE (Sr and K). crystallization temperatures and pressures can be calculated from experimentally established clinopyroxene and clinopyroxene–liquid 5.4. Origin of the MPAs thermobarometers (Appendix 2; Putirka et al., 2003; Putirka, 2008). Based on microprobe analyses of clinopyroxene, we estimate a The MPAs are featured by low Yb/Lu (6.78 ± 0.44 (1σ)), Dy/Yb temperature range of 956 to 1087 °C with the mean value of 1032 ± (1.74 ± 0.14 (1σ)) and (Ho/Yb)N (1.04 ± 0.07 (1σ)), suggesting 39 °C (1σ), and a pressure range from 5.9 to 13.6 kbar with an average that amphibole was the prevailing buffering phase in their source of 9.8 ± 1.9 kbar (1σ) (Appendix 2). These values are close to the region (Moyen, 2009; Qin et al., 2010b). This deduction is further experimentally predicted solidus temperatures of hydrated amphibo- conﬁrmed in the (La/Yb)N vs YbN diagram (Fig. 11), which is also consis- lite (Wyllie and Wolf, 1993) and hydrous peridotite (Hirose, 1997). tent with our temperature estimates above. Moreover, the linear corre- Furthermore, this temperature is well above the solidus temperature lation between Sc versus V indicates clinopyroxene and/or amphibole 166 X.-W. Li et al. / Lithos 172–173 (2013) 158–174

60 0.9

58 0.8

56 0.7

54 0.6

52 0.5

50 0.4

16 4

14 2

12 0

8 10

6 8

4 6

2 4

0 2 246810246810

Fig. 7. Plots of major elements versus MgO for the MPAs in West Qinling.

were involved as the main fractionation phase(s) for the magma gene- As for the HMA, they cannot be generated by partial melting of sis (Fig. 12A), as is suggested by high Amph/melt Kd and Cpx/melt Kd a basaltic slab alone, because the experiments on natural hydrous values for both of these elements. Similarly, the high ilmenite/melt Kd basalts at 8–32 kbar and 10–40% melting only produce tonalitic to values for both Ta and Hf suggest the ilmenite was also a fractionation trondhjemitic melts (Kelemen, 1995; Rapp et al., 1991, 1999; Zhao phase (Fig. 12B). Petrographic evidence of late plagioclase crystalliza- et al., 2009b). Combined with high Ni and Cr contents, the high tion, deduced from the fact that the great majority of the plagioclase Mg# andesite (HMA) can be produced by (1) partial melting of occurs as microcrystals in the groundmass, is also indicated in Fig. 12C hydrated peridotite (Kelemen, 1995; Kushiro, 1974; Mysen and and D. Boettcher, 1975; Straub et al., 2011; Tatsumi, 1981, 1982; Wood and Phenocryst minerals exhibit strong textural and compositional Turner, 2009); (2) interaction between lower crust and astheno- disequilibrium. Some plagioclase crystals show sieve-like appearance sphere via delamination or break-off (Gao et al., 2004); (3) crustal- (Fig. 3A), and bimodal distribution of An numbers between pheno- level magma mixing and potential ascension of mafic/ultramafic crysts and groundmass. The observed normal and reverse zonings in cumulates (Shellnutt and Zellmer, 2010; Streck et al., 2007; Tiepolo pyroxene phenocrysts can be explained by magma mixing. The uni- et al., 2011); (4) interaction between melt (slab and/or sediment) form magnesian interior of orthopyroxene and the magnesium-rich and overlying mantle wedge (Ayabe et al., 2012; Kay, 1978; Kelemen, rim of clinopyroxene represent the products of the mafic-end mem- 1995; Shimoda et al., 1998; Tatsumi, 2001; Tsuchiya et al., 2005; ber, whereas the more iron-rich rim of orthopyroxene and the core Wang et al., 2011; Yogodzinski et al., 1995). of clinopyroxene with reverse zoning could stand for the felsic-end We argue that the first two models are not applicable to the MPAs member. according to the evidence below. First, the MPAs require parental X.-W. Li et al. / Lithos 172–173 (2013) 158–174 167

A Setouchi HMA Troodos HMA California & Bonin HMA

Antarctic Peninsula HMA

Hybridized slab melts(ca.4Gpa)

Alexander Island HMA

normal arcs

Mg#

Slab melts(1-4Gpa)

SiO2

12 B This study Aonaegawa HMA

10 Setouchi HMA Kadavu Island HMA

Partial melts of metabasalts and HMA amphibolites at 1.5 - 3.2 Gpa 8 Average of partial melts of metabasalt at 1100 °C @ 3.8GPa and basalt at 1100 °C @ 3.2GPa Melts of basalt hybridized with peridotite at 3.8GPa 6

MgO(wt. %) normal arcs 4

0 50 55 60 65 70 75 80

SiO2(wt. %)

Fig. 8. (A) Plot of SiO2 versus MgO for the MPAs in West Qinling: HMAs and normal arc andesites from other areas are from McCarron and Smellie (1998) and references therein; the

ﬁelds for the slab melts at 1–4 GPa and hydradized slab melt at ca. 4 GPa are from Rapp et al. (1999) (B) SiO2 vs. MgO (wt.%) for the MPAs and HMAs from other areas and normal arc andesites compared with experimentally produced partial melt data. The brown and gray areas are from McCarron and Smellie (1998). HMA in Aonaegawa, SW and NE Japan arcs, Sato et al. (2013); HMA in the Setouchi volcanic zone, SW Japan arc, Shimoda et al. (1998); HMA from Kadavu Island, Danyushevsky et al. (2008); partial melts of metabasalt and amphibolite at 1.5–3.2 GPa, Rapp et al. (1991), Winther and Newton (1991), Sen and Dunn (1994) and Rapp and Watson (1995); average of partial melts of metabasalt at 1100 °C and 3.2 GPa and basalt at 1100 °C and 3.8 GPa, Rapp and Watson (1995) and Rapp et al. (1999); melts of basalt hybridized with peridotite at 3.8 GPa, Rapp et al. (1999).

magma with high initial 87Sr/86Sr ratios, readily ruling out the ﬁrst 2010a). Third, the low ratios of Sr/Y and Ba/Th, and high ratios of Th/La model. Second, the second model requires the lower crust to descend and (La/Sm)N reveal major contribution of sediments (Fig. 13A and B), into the asthenosphere followed by partial melting immediately and rather than contribution from an altered oceanic crust (Tatsumi, interacting with the depleted mantle to obtain high Mg#, Ni and Cr 2006). Moreover, the crustal contamination in the magma chamber contents. However, this model must produce the depleted signatures, would predict high Mg in the core and low Mg in the rim of the which is inconsistent with our observation (Gao et al., 2004; Qin et al., clinopyroxene phenocrysts, which is not observed. Instead, reverse 168 X.-W. Li et al. / Lithos 172–173 (2013) 158–174

– A incorporation of 10 20% sediment melt to the mantle wedge-derived 100 magma can successfully create the trace element patterns of the MPAs. Both the LMA and HMA have indistinguishable isotope data and very similar REE and trace element distribution patterns. The Mg # Lower crust values of the LMA is just a little higher than those of the rocks derived from slab melt (Fig. 8), and it implies that the LMA magmas were slightly hybridized by mantle peridotite-derived melts via magma 10 mixing. However, the HMA group of the MPAs have pronounced higher Mg# and compatible element concentrations, indicating that

samples/chondrite their parental magmas were intensively hybridized (e.g. Wang et al., 2011).

5.5. Tectonic implication and tentative model 1 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu The tectonic background in the Qinling Orogen during the middle– late Triassic is hotly debated at present. Jiang et al. (2010, 2011) pos- tulated that oceanic subduction did not cease until the Norian (228– B 209 Ma; Walker et al., 2013). Whereas several authors argue that 100 the syn-collision stage started at ca. 237 Ma and ceased at ca. 215 Ma lower crust (Dong et al., 2011; Qin et al., 2010a,b; Sun et al., 2002; Zhang et al., 2008). This was followed by a post-collision stage from ca. 215 Ma to 10 ca. 200 Ma. However, Zhang et al. (2006) and Luo et al. (2012) believe that the middle-Triassic Heimahe pluton formed during the post- collision regime. The conclusions, nevertheless, solely deduced from

1 samples/primitive mantle A

0.1 calculated parental melt for ZK10-36-17 (PM17) Rb Ba Th U Nb Ta La Ce Pb Pr Sr Nd Zr Hf Sm Eu Ti Gd Tb Dy Y Ho Er Tm Yb Lu

Fig. 9. (A) Chondrite-normalized REE patterns and (B) primitive mantle-normalized spidergrams for the MPAs in West Qinling. The normalized values are from Sun and calculated parental melt for ZK10-36-30 (PM30) McDonough (1989); lower crust values are from Rudnick and Gao (2003).

compositional zoning is observed in clinopyroxene phenocrysts. Hence, the third model is also not perfectly applicable to explain the genesis of the MPAs. However, we recognized that magma mixing has played a key role for the genesis of the MPAs (Streck et al., 2007; Zhang and Shao, 2008). Hence, we tentatively propose that the MPAs were probably produced by the interaction between the mantle peridotite-derived basaltic melts and sediment-derived melts. Firstly, the high Mg-value of the calculated parental melt for ZK10-36-30 (PM30) B melts in equilibrium with both the core of orthopyroxene and the rim of clinopyroxene collectively require that these kinds of melts must calculated parental melt for ZK10-36-17 (PM17) have been derived from partial melting of peridotites. Secondly, the compositions of sediment-derived melts can form different hybridized magmas varying from rhyolitic to andesitic as they dissolve olivine and pyroxene in peridotites (Tatsumi, 2001; Wang et al., 2011). Thirdly, they exhibit relatively high initial 87Sr/86Sr isotopic ratios and low

εNd(t) values (Fig. 14A), and low Nb/Ta ratios, similar to those of terrigenous sediments. Fourthly, pioneering research has shown that subduction-related magmas with a robust ﬁngerprint of subducted sediments in their source area will produce a clearly positive correlation between Th/Ce versus Th/Sm ratios (Gao et al., 2008), as illustrated in Fig. 13C. Moreover, to quantify the process that is potentially responsible for generating the MPAs, we employ several different models using 87 86 Fig. 10. (A) Rare earth and (B) other trace element compositions of the parental isotopic and trace elements. In the diagram of εNd(t)versus( Sr/ Sr)t magmas (PM 17 and PM 30) calculated from the trace element abundances in (Fig. 14A), one end member is the original mantle wedge (MW), the clinopyroxenes; also shown are the actual REE and other trace element compositions other one represents the sediments (SED). The mixing calculation of clinopyroxenes. Clinopyroxene/liquid partition coefﬁcients used in the calculation shows that about 10–20% involvement of sediment melts can generate are La 0.0536, Ce 0.0858, Pr 0.124, Nd 0.1873, Sm 0.291, Eu 0.3288, Gd 0.367, Tb 0.404, Dy 0.38, Ho 0.4145, Er 0.387, Tm 0.4085, Yb 0.43, Lu 0.433, Rb 0.006, Ba the hybridized melts with varying (Sr/Nd)SED ratios of 2–10. In the 0.00068, Th 0.0021, Nb 0.008, Sr 0.1283, P 0.13, Ti 0.34, Y 0.412, Rb 0.006, Ba N-MORB normalized diagram, the high magnesian basalt (e.g. basalt 0.00068, Th 0.0021, Nb 0.008, Sr 0.1283, P 0.13, Ti 0.34, Y 0.412 (Bédard, 2001, and ref- SD438: Tatsumi and Hanyu, 2003) and subducting sediment values are erences therein). The chondrite values and primitive mantle values are from Sun and assumed to follow Tatsumi and Hanyu (2003). As shown in Fig. 14B, McDonough (1989). X.-W. Li et al. / Lithos 172–173 (2013) 158–174 169

Adakite

10% garnet amphibolite

amphibolite 25 “Normal”arc andesite

50 10 50 MORB

Fig. 11. (La/Yb)N versus YbN for the MPAs in West Qinling (modiﬁed after Defant and Drummond, 1990) Adakite ﬁelds from Martin (1999). Red dash lines with arrows indicate partial melting of an amphibolite and a garnet-bearing (10% garnet) amphibolite.

0.55 A 180 B

0.50 160

140 0.45

Ta 120 0.40

100 0.35 ilm cpx or amph 80 0.30 2.6 2.8 3.0 3.2 3.4 3.6 3.8 14 16 18 20 22 24 26 28 30 Hf Sc

8 1.0

7 C D 0.8 6

5 opx 0.6 opx 4

Lu 0.4 3 pl 2 0.2 1 pl cpx cpx 0 0.0 1000 2000 3000 4000 5000 6000 0 5 10 15 20 25 30 35 40 Ti Y

Fig. 12. (A) V versus Sc; (B) Ta versus Hf; (C) Tu versus Ti; (D) Lu versus Y. The line with arrow indicates calculated fractional crystallization path for mineral, using partition coefﬁcients from Ersoy and Helvacı (2010), and references therein. Mineral abbreviations: cpx: clinopyroxene, opx: orthopyroxene, amph: amphibole, ilm: ilmenite, pl: plagioclase. 170 X.-W. Li et al. / Lithos 172–173 (2013) 158–174

2000; Li et al., 2007; Liu et al., 2005; Meng et al., 2005). The initial colli- A sion in East Qinling began during the early Triassic (Li et al., 1993; Yin and Nie, 1993), whereas the western counterpart might only have

upper crust and marine sediments started during the middle Triassic (Meng et al., 2005). In a broader con- text, the well deﬁned tectono-magmatic evolution in the westward extending A'nimaque-Mianlue paleo-ocean, East Kunlun paleo-ocean (Buqingshan–A'nimaque Ocean), shows the initial collision commenced during the middle Triassic whereas the post-collision stage was during the late-Triassic (Guo et al., 1998; Mo et al., 2007; Yang et al., 2009; Zhang et al., 2012). MORB The meta-volcanic rocks in the Mianlue suture zone with the metamorphic ages of 226.9 ± 0.9 Ma and 219.4 ± 1.4 Ma readily indicate the closure of Mianlue Ocean (Lai and Qin, 2010),indicating that the hypothesis as proposed by Jiang et al. (2010, 2011) that oceanic subduction lasted till 211 Ma is unlikely. Besides, the remnant A'nimaque–Mianlue oceanic basin dramatically shrinked during the Ladinian (241–237 Ma; Walker et al., 2013), and was exclusively limited to the Diebu–Songpan area during the Carnian (237–228 Ma; B Walker et al., 2013)–Early Norian (Liu et al., 2005). Moreover, the inherited arc-signature of some igneous rocks in the Qinling Orogeny Altered oceanic crust during syn-collision and post-collision should not be neglected. Besides, the relatively continuous and steady deposition of the Gulangdi Formation ruled out the possibility of a post-collision regime in West

10.0 Original MantleWedge A Arc 5.0 this study Wang Qiang et al,2011

Sediment Aleutian HMA 0.0

N-MORB Bonin HMA (t) (Sr/Nd)s=10

Nd -5.0 ε (Sr/Nd)s=5.7

-10.0 (Sr/Nd)s=2

-15.0 C Sediments -20.0 0.702 0.704 0.706 0.708 0.710 0.712 0.714 0.716 0.718 (87Sr/86Sr)t

R=0.85 B 1000 MPAs

100

N-MORB Normallized mantle wedge 1 sediment melt 10% sediments Fig. 13. (A) Th/La versus Th (after Plank, 2005); (B) Ba/Th versus (La/Sm)N (after Tatsumi, ﬁ 20% sediments 2006); (C) Plot of Th/Ce versus Th/Sm showing a signi cant linear relationship (after Gao 30% sediments et al., 2008). 0.1 Rb Ba Th Nb K Pb Sr Nd

87 86 petrogeochemistry sometimes are ambiguous, because the arc-signature Fig. 14. (A) Plot of εNd(t) versus ( Sr/ Sr)t showing two-component mixing for the in terms of trace elements (the depletion of HFSE such as Nb, Ta, Ti and MPAs in West Qinling. HMA data from Wang et al. (2011) are also shown. (B) Plot of Hf) can be preserved in the continuous process from oceanic subduction incompatible element plot normalized to normal mid-ocean ridge basalt (N-MORB; “ ” Sun and McDonough, 1989) for the MPAs in West Qinling. Data indicate that the to incipient collision ( soft touch style), then to post-collision (e.g. Mo et MPAs can be generated by mixing of melts from sediments and mantle wedge al., 2008). When the paleogeographic evolution is taken into account, the peridotites. Inferred original mantle wedge values (trace elements and isotopes) and dramatic facies change from marine strata to non-marine strata (ubiqui- trace element values of sediment-derived melts are taken from Tatsumi and Hanyu tous angular unconformity) during the latest middle-Triassic should not (2003). We assume that the isotope ratios of sediment-derived melt with different Sr/Nd ratios (2, 5.7, and 10) are: (87Sr/86Sr) = 0.71703 and εNd(t) = −20.2. Marks be neglected (Li, 2011; Meng et al., 2005, 2007; Pan et al., 1997; Yan et al., t along the mixing lines indicate every 10% increments of sediment melt; 10% to 20% 2012; Zhou and Graham., 1996). The east-to-west diachronous collision sediment melt mixed with melt from the mantle wedge, represented by Basalt between YB and NCB has been proposed by some authors (Hacker et al., DS438 (Tatsumi and Hanyu, 2003), span the range of MPA compositions. X.-W. Li et al. / Lithos 172–173 (2013) 158–174 171

Qinling during the middle Triassic as proposed by Zhang et al. (2006) subduction, and essential volcano-tectonic events within the subduction and Luo et al. (2012). More specifically, the model proposed by Zhang zone (Zhang et al, 2012, and references therein). Therefore, we argue et al. (2006) and Luo et al. (2012) cannot be reconciled with the follow- that northward subduction of the A'nimaque–Mianlue paleo-ocean ing lines of evidence: (1) The presence of a subduction-related andesitic could last at least till the middle Triassic, and the MPAs are probably the ignimbrite from Yazhagou with an age of 246.2 ± 2.8 Ma (Dong et al., products of initial collision when the oceanic slab was still subducting. 2011; Qin et al., 2008). (2) Oceanic subduction induced maficdyke Combined with a late-Triassic slab break-off model proposed by swarms with the age of 251 ± 2 Ma in East Kunlun (Xiong et al., Qin et al. (2010b) and numerical modeling results (Baumann et al., 2011). (3) The subduction of the Buqingshan–A'nimaque beneath East 2010; Duretz et al., 2011), we suggest that the regional unconformity Kunlun lasted at least until Early Triassic as proposed by Wang et al. between the Gulangdi Fm. and the Maixiu Group is a consequence of (1997). (4) The paleomagnetic age data confirms that the NCB and YB the initial collision of the YB with the NCB while the subduction was moved further apart during the Middle Permian to the Middle–Late still active (Fig. 15). We suggest the MPAs were generated by slab Triassic (Enkin et al., 1992; Hacker et al., 2004; Lin and Fuller, 1990; roll-back during the initial collision process, similar to the scenario Zhao and Coe, 1987). (5) The sedimentary facies change from turbiditic during 65–45 Ma in Tibet proposed by Ding et al. (2003) and Mo et deposites during the Early–Middle Triassic to shallow marine-terrestrial al. (2008). In the arc-wedge corner, the asthenospheric upwelling deposits during the Middle–Late Triassic in West Qinling (Yan et al., resulted in decompression (Baumann et al., 2010; Kincaid and 2012). Although tectonic uplift existed during the Permian, immediate Griffiths, 2003). Roll-back of the subducting slab could cause a loss subsequent subsidence was also observed during the Ladinian (Yang et of lithosphere which should have been replaced by asthenosphere al., 1994) in the southern Qinling and Songpan Terrane, as evidence (Gueguen et al., 1997). This resulted in the rapid uplift in the north from flysch deposition in large turbidite basins indicates. (6) Apart part of West Qinling, and the back-arc extensional basin formed with- from collision, the late Permian regional unconformity between the in the northern margin of West Qinling during the middle-Triassic (Li, Gequ Formation and underlying A'nimaque ophiolitic mélange in the 2011). EasternKunlunOrogencouldalsobetheresultofsomesubduction- This interpretation is geochemically and geologically reasonable related processes, such as changing subduction angle/speed, ridge because (1) hybrid melts from the contemporary pluton reveal a

A'nimaque-Mianlue Paleo-ocean

Qinling Continental Crust

subducting sediments

magma source

amphibole stable sub-continental lithosphere mantle

oceanic lithosphere

oceanic crust asthenosphere

asthenosphere

slab roll-back

Fig. 15. Proposed tectonic model for the generation of the MPAs in West Qinling during the late Triassic. See text for explanation. 172 X.-W. Li et al. / Lithos 172–173 (2013) 158–174 considerable depleted mantle source manifested in zircon Hf isotopes Boettcher, A.L., 1973. Volcanism and orogenic belts — the origin of andesites. Tectonophysics – fi 17, 223 240. (Guo et al., 2012); (2) mantle contribution was also exempli ed by Chen, Y., Zhu, D., Zhao, Z., Zhang, L., Liu, M., Yu, F., Guan, Q., Mo, X., 2010. Geochronology, the high Cr and Ni contents in the MPAs; (3) south-eastward roll- geochemistry and petrogenesis of the Bamco andesites from the northern Gangdes, back, as evidenced by the progressively younger magmatism from Tibet. Acta Petrologica Sinica 26, 2193–2206 (in Chinese with English Abstract). Chiaradia, M., Müntener, O., Beate, B., 2011. Enriched basaltic andesites from mid- the northern to the southern margin (Fig. 1C). This commenced as crustal fractional crystallization, recharge, and assimilation (Pilavo Volcano, Western early as ca. 245 Ma according to available data up to now. Cordillera of Ecuador). Journal of Petrology 52, 1107–1141. The linear distribution of the Early Indosinian intrusive rocks in Danyushevsky, L.V., Falloon, T.J., Crawford, A.J., Tetroeva, S.A., Leslie, R.L., Verbeeten, A., fi West Qinling suggests that slab roll-back played a key role in the 2008. High-Mg adakites from Kadavu Island Group, Fiji, southwest Paci c: evidence for the mantle origin of adakite parental melts. Geology 36, 499–502. formation of the MPAs and it could also be responsible for rapid uplift Defant, M.J., Drummond, M.S., 1990. Derivation of some modern arc magmas by melting of in northern Qinling as suggested by fission track analysis (Zheng et young subducted lithosphere. Nature 347, 662–665. al., 2004). This is a response to influx of mantle materials during Ding, L., Kapp, P., Zhong, D., Deng, W., 2003. Cenozoic volcanism in Tibet: evidence for a transition from oceanic to continental subduction. Journal of Petrology 44, 1833–1865. slab roll-back. Additionally, the rapid cooling history around 220 Ma Dong, Y., Zhang, G., Neubauer, F., Liu, X., Genser, J., Hauzenberger, C., 2011. Tectonic (6–15 °C/Ma) (Zheng et al., 2004) in the north margin of Qinling evolution of the Qinling Orogen, China: review and synthesis. Journal of Asian probably matches the slab break-off model as proposed by Qin et al. Earth Sciences 41, 213–237. Dunn, T., 1987. Partitioning of Hf, Lu, Ti, and Mn between olivine, clinopyroxene and (2010b). basaltic liquid. Contributions to Mineralogy and Petrology 96, 476–484. Duretz, T., Gerya, T.V., May, D.A., 2011. Numerical modelling of spontaneous slab 6. Conclusions breakoff and subsequent topographic response. Tectonophysics 502, 244–256. Enkin, R.J., Yang, Z., Chen, Y., Courtillot, V., 1992. Paleomagnetic constraints on the geodynamic history of the major blocks of China from the Permian to the Present. (1) Laser fusion 40Ar/39Ar dating for matrix glass indicates the Journal of Geophysical Research 97, 13953–13989. Ersoy, Y., Helvacı, C., 2010. FC–AFC–FCA and mixing modeler: a Microsoft® Excel© MPAs, mainly consisting of pyroxene andesite with high magne- spreadsheet program for modeling geochemical differentiation of magma by crys- sium contents in some samples, erupted during the late Triassic tal fractionation, crustal assimilation and mixing. Computers and Geosciences 36, (234 ± 3 Ma). 383–390. fi Feng, Y., Cao, X., Zhang, E., Hu, Y., Pan, X., Yang, J., Jia, Q., Li, W., 2002. Structure, Orogenic (2) The MPAs have typical arc-af nity geochemical features, with Processes and Geodynamic of the Western Qinling Orogen. Xi'an Map Press, Xi'an, 87 86 uniformly low εNd(t) values and high ( Sr/ Sr)t ratios. p. 263 (in Chinese). (3) The MPAs were probably produced by mixing of melts from sed- Gao, S., Rudnick, R.L., Yuan, H.-L., Liu, X.-M., Liu, Y.-S., Xu, W.-L., Ling, W.-L., Ayers, J., iments and mantle wedge peridotites. Wang, X.-C., Wang, Q.-H., 2004. Recycling lower continental crust in the North China Craton. Nature 432, 892–897. (4) The early Mesozoic arc-related volcanic events in West Qinling Gao, Y., Wei, R., Hou, Z., Tian, S., Zhao, R., 2008. Eocene high-MgO volcanism in southern show that the continental–continental collision between the Tibet: new constraints for mantle source characteristics and deep processes. Lithos – NCB and the YB likely occurred during the middle-Triassic. 105, 63 72. Gill, J.B., 1981. Orogenic Andesites and Plate Tectonics. Springer-Verlag Berlin, New This interpretation is consistent with the evidence from paleo- York, p. 390. geography and the tectono-sedimentary history, numerical Graham, I.J., Cole, J.W., 1991. Petrogenesis of andesites and dacites of White Island modeling and paleomagnetism. volcano, Bay of Plenty, New Zealand, in the light of new geochemical and isotopic data. New Zealand Journal of Geology and Geophysics 34, 303–315. Grove, T.L., Kinzler, R.J., 1986. Petrogenesis of andesites. Annual Review of Earth and Supplementary data to this article can be found online at http:// Planetary Sciences 14, 417–454. dx.doi.org/10.1016/j.lithos.2013.04.010. Gueguen, E., Doglioni, C., Fernandez, M., 1997. Lithospheric boudinage in the Western Mediterranean back-arc basin. Terra Nova 9, 184–187. Guo, Z., Deng, J., Xu, Z., Mo, X., Luo, Z., 1998. Late Palaeozoic–Mesozoic intracontinental Acknowledgments orogenic process and intermediate-acidic igneous rocks from the Eastern Kunlun Mountains in north Tibet, northwestern China. Geoscience 12, 344–352 (in Chinese We are particularly grateful to Professor Nelson Eby and two with English Abstract). Guo, X., Yan, Z., Wang, Z., Wang, T., Wu, F., Fu, C., Hou, K., Li, J., 2011. Geological char- anonymous reviewers for their kind and critically constructive com- acteristics and associated magmatic ages of the Xiekeng skarn-type Cu–Au deposit ments and suggestions, which greatly improved the quality of our in the West Qinling terrane. Acta Petrologica Sinica 27, 3811–3822 (in Chinese manuscript. We profoundly appreciate Yong-Sheng Liu, Hai-Hong with English Abstract). Chen and Shuo Liu for the assistance of trace element analyses, and Guo, X., Yan, Z., Wang, Z., Wang, T., Hou, K., Fu, C., Li, J., 2012. Middle Triassic arc magmatism along the northeastern margin of the Tibet: U–Pb and Lu–Hf zircon Hong Qin and Shao-Long Mo for major element analyses, Gui-Ming characterization of the Gangcha complex in the West Qinling terrane, central Shu, Zhen-Yu Chen and Xiao-Dan Chen for the assistance with the China. Journal of the Geological Society 169, 327–336. EPMA analyses, and Jian-Qing Ji and Jing Zhou for the 40Ar/39Ar Hacker, B.R., Ratschbacher, L., Webb, L., McWilliams, M.O., Ireland, T., Calvert, A., Dong, S., Wenk, H.-R., Chateigner, D., 2000. Exhumation of ultrahigh-pressure continental dating. Mark Scheltens, Thomas Bader, Jin-Yu Zhang and Yi Cao are crust in east central China: Late Triassic–Early Jurassic tectonic unroofing. Journal also profoundly thanked for providing helpful comments and word of Geophysical Research 105, 13339–13364. polishing. This work is financially supported by China Geological Survey Hacker, B.R., Ratschbacher, L., Liou, J.G., 2004. Subduction, collision and exhumation in the ultrahigh-pressure Qinling–Dabie Orogen. Geological Society, London, Special (1212011220928 and 1212011121260), National Key Project for Basic Publications 226, 157–175. Research (2009CB421002 and 2011CB403102) and NSF of China Hall, C.M., Farrell, J.W., 1995. Laser 40Ar39Ar ages of tephra from Indian Ocean deep-sea (41121062, 40972049, and 40830317). sediments: tie points for the astronomical and geomagnetic polarity time scales. Earth and Planetary Science Letters 133, 327–338. Hart, S.R., Dunn, T., 1993. Experimental cpx/melt partitioning of 24 trace elements. References Contributions to Mineralogy and Petrology 113, 1–8. Hirose, K., 1997. Melting experiments on lherzolite KLB-1 under hydrous conditions Ayabe, M., Takanashi, K., Shuto, K., Ishimoto, H., Kawabata, H., 2012. Petrology and and generation of high-magnesian andesitic melts. Geology 25, 42–44. geochemistry of adakitic dacites and high-MgO andesites, and related calc-alkaline Jiang, Y.-H., Jin, G.-D., Liao, S.-Y., Zhou, Q., Zhao, P., 2010. Geochemical and Sr–Nd–Hf dacites from the Miocene Okoppe Volcanic Field, N Hokkaido, Japan. Journal of isotopic constraints on the origin of Late Triassic granitoids from the Qinling Petrology 53, 547–588. Orogen, central China: implications for a continental arc to continent–continent Bailey, J.C., 1981. Geochemical criteria for a refined tectonic discrimination of orogenic collision. Lithos 117, 183–197. andesites. Chemical Geology 32, 139–154. Jiang, Y.-H., Jin, G.-D., Liao, S.-Y., Zhou, Q., Zhao, P., 2011. Petrogenesis and tectonic Baumann, C., Gerya, T.V., Connolly, J.A.D., 2010. Numerical modelling of spontaneous implications of ultrapotassic microgranitoid enclaves in Late Triassic arc granitoids, slab breakoff dynamics during continental collision. Geological Society, London, Qinling Orogen, central China. International Geology Review 54, 208–226. Special Publications 332, 99–114. Jin, W., Zhang, Q., He, D., Jia, X., 2005. SHRIMP dating of adakites in western Qinling and their Bédard, J., 2001. Parental magmas of the Nain Plutonic Suite anorthosites and mafic implications. Acta Petrologica Sinica 21, 959–966 (in Chinese with English Abstract). cumulates: a trace element modelling approach. Contributions to Mineralogy and Kay, R.W., 1978. Aleutian magnesian andesites: melts from subducted Pacific Ocean Petrology 141, 747–771. crust. Journal of Volcanology and Geothermal Research 4, 117–132. Bédard, J.H., 2007. Trace element partitioning coefficients between silicate melts and Kelemen, P.B., 1995. Genesis of high Mg# andesites and the continental crust. Contributions orthopyroxene: parameterizations of D variations. Chemical Geology 244, 263–303. to Mineralogy and Petrology 120, 1–19. X.-W. Li et al. / Lithos 172–173 (2013) 158–174 173

Kelemen, P.B., Hanghøj, K., Greene, A.R., 2003. One view of the geochemistry of Plank, T., 2005. Constraints from thorium/lanthanum on sediment recycling at subduction subduction-related magmatic arcs, with an emphasis on primitive andesite and zones and the evolution of the continents. Journal of Petrology 46, 921–944. lower crust. In: Heinrich, D.H., Karl, K.T. (Eds.), Treatise on Geochemistry. Pergamon, Putirka, K.D., 1999. Clinopyroxene + liquid equilibria to 100 kbar and 2450 K. Contributions Oxford, pp. 1–70. to Mineralogy and Petrology 135, 151–163. Kincaid, C., Griffiths, R.W., 2003. Laboratory models of the thermal evolution of the Putirka, K.D., 2008. Thermometers and barometers for volcanic systems. Reviews in mantle during rollback subduction. Nature 425, 58–62. Mineralogy and Geochemistry 69, 61–120. Kushiro, I., 1974. Melting of hydrous upper mantle and possible generation of andesitic Putirka, K.D., Mikaelian, H., Ryerson, F., Shaw, H., 2003. New clinopyroxene–liquid magma: an approach from synthetic systems. Earth and Planetary Science Letters thermobarometers for mafic, evolved, and volatile-bearing lava compositions, 22, 294–299. with applications to lavas from Tibet and the Snake River Plain, Idaho. American Lai, S., Qin, J., 2010. Ophiolites and Associated Volcanics in the Mianlue Tectonic Zone of Mineralogist 88, 1542–1554. the South Qinling Belt. Science Press, Beijing, p. 257 (in Chinese). QBGMR (Qinghai Bureau of Geology and Mineral Resources), 1991. Regional Geology of Li, L., 2011. Late Permian–Middle Triassic deep-marine sedimentation of the West Qinghai Province. Geology Publishing House, Beijing 662 (in Chinese). Qinling, central China. (Master Thesis) Chinese Academy of Sciences, p. 86 (in Qin, J., Lai, S., Zhang, G., Diwu, C., Li, Y., 2008. Zircon LA–ICP MS U–Pb dating of the Chinese). Longkang andesitic ignimbrites from Jiuzhaigou: evidence of the Mianlue Suture Li, C., Liu, Y., Zhu, B., Feng, Y., Wu, H., 1978. Tectonic evolution of Qinling and Qilian westward extension. Journal of China University of Geosciences 19, 47–53. Orogen. Northwestern Geology, 1–12 (in Chinese). Qin, J., Lai, S., Grapes, R., Diwu, C., Ju, Y., Li, Y., 2009. Geochemical evidence for origin of Li, S., Xiao, Y., Liou, D., Chen, Y., Ge, N., Zhang, Z., Sun, S.-S., Cong, B., Zhang, R., Hart, S.R., magma mixing for the Triassic monzonitic granite and its enclaves at Mishuling in Wang, S., 1993. Collision of the North China and Yangtse Blocks and formation of the Qinling Orogen (central China). Lithos 112, 259–276. coesite-bearing eclogites: timing and processes. Chemical Geology 109, 89–111. Qin, J.-F., Lai, S.-C., Diwu, C.-R., Ju, Y.-J., Li, Y.-F., 2010a. Magma mixing origin for the Li, S., Kusky, T.M., Wang, L., Zhang, G., Lai, S., Liu, X., Dong, S., Zhao, G., 2007. Collision post-collisional adakitic monzogranite of the Triassic Yangba pluton, Northwestern leading to multiple-stage large-scale extrusion in the Qinling Orogen: insights margin of the South China block: geochemistry, Sr–Nd isotopic, zircon U–Pb dating from the Mianlue suture. Gondwana Research 12, 121–143. and Hf isotopic evidences. Contributions to Mineralogy and Petrology 159, 389–409. Lin, J.-L., Fuller, M., 1990. Palaeomagnetism, North China and South China collision, and Qin, J.-F., Lai, S.-C., Grapes, R., Diwu, C., Ju, Y.-J., Li, Y.-F., 2010b. Origin of Late Triassic the Tan–Lu Fault. Philosophical Transactions of the Royal Society of London. Series high-Mg adakitic granitoid rocks from the Dongjiangkou area, Qinling Orogen, central A Mathematical and Physical Sciences 331, 589–598. China: implications for subduction of continental crust. Lithos 120, 347–367. Liu, Y., Gao, S., Yuan, H., Zhou, L., Liu, X., Wang, X., Hu, Z., Wang, L., 2004. U–Pb zircon ages Rapp, R.P., Watson, E.B., 1995. Dehydration melting of metabasalt at 8–32 kbar: implications and Nd, Sr, and Pb isotopes of lower crustal xenoliths from North China Craton: for continental growth and crust–mantle recycling. Journal of Petrology 36, 891–931. insights on evolution of lower continental crust. Chemical Geology 211, 87–109. Rapp, R.P., Watson, E.B., Miller, C.F., 1991. Partial melting of amphibolite/eclogite and Liu, S., Steel, R., Zhang, G., 2005. Mesozoic sedimentary basin development and tectonic the origin of Archean trondhjemites and tonalites. Precambrian Research 51, implication, northern Yangtze Block, eastern China: record of continent–continent 1–25. collision. Journal of Asian Earth Sciences 25, 9–27. Rapp, R.P., Shimizu, N., Norman, M.D., Applegate, G.S., 1999. Reaction between slab-derived Liu, Y., Zong, K., Kelemen, P.B., Gao, S., 2008. Geochemistry and magmatic history of melts and peridotite in the mantle wedge: experimental constraints at 3.8 GPa. eclogites and ultramafic rocks from the Chinese continental scientific drill hole: Chemical Geology 160, 335–356. subduction and ultrahigh-pressure metamorphism of lower crustal cumulates. Ratschbacher,L.,Hacker,B.R.,Calvert,A.,Webb,L.E.,Grimmer,J.C.,McWilliams,M.O.,Ireland, Chemical Geology 247, 133–153. T., Dong, S., Hu, J., 2003. Tectonics of the Qinling (Central China): tectonostratigraphy, Lugmair, G.W., Marti, K., 1978. Lunar initial 143Nd/144Nd: differential evolution of the geochronology, and deformation history. Tectonophysics 366, 1–53. lunar crust and mantle. Earth and Planetary Science Letters 39, 349–357. Reubi, O., Blundy, J., 2009. A dearth of intermediate melts at subduction zone volcanoes Luo, B., Zhang, H., Lü, X., 2012. U–Pb zircon dating, geochemical and Sr–Nd–Hf isotopic and the petrogenesis of arc andesites. Nature 461, 1269–1273. compositions of Early Indosinian intrusive rocks in West Qinling, central China: Ross, K., Elthon, D., 1993. Cumulates from strongly depleted mid-ocean-ridge basalt. petrogenesis and tectonic implications. Contributions to Mineralogy and Petrology Nature 365, 826–829. 164, 551–569. Rudnick, R.L., Gao, S., 2003. Composition of the continental crust. In: Heinrich, D.H., Martin, H., 1999. The adakitic magma: modern analogue of Archean granitoids. Lithos Karl, K.T. (Eds.), Treatise on Geochemistry. Pergamon, Oxford, pp. 1–64. 46, 411–429. Sato, M., Shuto, K., Uematsu, M., Takahashi, T., Ayabe, M., Takanashi, K., Ishimoto, H., Mattauer, M., Matte, P., Malavieille, J., Tapponnier, P., Maluski, H., Qin, X.Z., Lun, L.Y., Kawabata, H., 2013. Origin of Late Oligocene to Middle Miocene adakitic andesites, Qin, T.Y., 1985. Tectonics of the Qinling Belt: build-up and evolution of eastern high magnesian andesites and basalts from the back-arc margin of the SW and NE Asia. Nature 317, 496–500. Japan arcs. Journal of Petrology 54, 481–524. McCarron, J.J., Smellie, J.L., 1998. Tectonic implications of fore-arc magmatism and generation Sen, C., Dunn, T., 1994. Dehydration melting of a basaltic composition amphibolite at of high-magnesian andesites: Alexander Island, Antarctica. Journal of the Geological 1.5 and 2.0 GPa: implications for the origin of adakites. Contributions to Mineralogy Society 155, 269–280. and Petrology 117, 394–409. Meng, Q.-R., Zhang, G.-W., 2000. Geologic framework and tectonic evolution of the Shellnutt, J.G., Zellmer, G.F., 2010. High-Mg andesite genesis by upper crustal differen- Qinling Orogen, central China. Tectonophysics 323, 183–196. tiation. Journal of the Geological Society 167, 1081–1088. Meng, Q.-R., Wang, E., Hu, J.-M., 2005. Mesozoic sedimentary evolution of the northwest Shimoda, G., Tatsumi, Y., Nohda, S., Ishizaka, K., Jahn, B.M., 1998. Setouchi high-Mg Sichuan basin: implication for continued clockwise rotation of the South China andesites revisited: geochemical evidence for melting of subducting sediments. Earth block. Geological Society of America Bulletin 117, 396–410. and Planetary Science Letters 160, 479–492. Meng, Q., Qu, H., Hu, J., 2007. Triassic deep-marine sedimentation in the western Song, S., Su, L., Li, X.-h., Zhang, G., Niu, Y., Zhang, L., 2010. Tracing the 850-Ma continental Qinling and Songpan terrane. Science in China Series D: Earth Sciences 50 (Suppl. flood basalts from a piece of subducted continental crust in the North Qaidam UHPM II), 246–263. belt, NW China. Precambrian Research 183, 805–816. Mo, X., Luo, Z., Deng, J., Yu, X., Liu, C., Chen, H., Yuan, W., Liu, Y., 2007. Granitoids and Steiger, R.H., Jäger, E., 1977. Subcommission on geochronology: convention on the use crustal growth in the East-Kunlun Orogenic Belt. Geological Journal of China of decay constants in geo- and cosmochronology. Earth and Planetary Science Universities 13, 403–414 (in Chinese with English Abstract). Letters 36, 359–362. Mo, X., Niu, Y., Dong, G., Zhao, Z., Hou, Z., Zhou, S., Ke, S., 2008. Contribution of Straub, S.M., Gomez-Tuena, A., Stuart, F.M., Zellmer, G.F., Espinasa-Perena, R., Cai, Y., syncollisional felsic magmatism to continental crust growth: a case study of the Iizuka, Y., 2011. Formation of hybrid arc andesites beneath thick continental Paleogene Linzizong volcanic succession in southern Tibet. Chemical Geology 250, crust. Earth and Planetary Science Letters 303, 337–347. 49–67. Streck, M.J., Leeman, W.P., Chesley, J., 2007. High-magnesian andesite from Mount Morimoto, N., Fabries, J., Ferguson, A.K., Ginzburg, I.V., Ross, M., Seifert, F.A., Zussman, J., Shasta: a product of magma mixing and contamination, not a primitive mantle Aoki, K., Gottardi, G., 1988. Nomenclature of pyroxenes. American Mineralogist 73, melt. Geology 35, 351–354. 1123–1133. Sun, S.-S., McDonough, W.F., 1989. Chemical and isotopic systematics of oceanic basalts: Moyen, J.-F., 2009. High Sr/Y and La/Yb ratios: the meaning of the “adakitic signature”. implications for mantle composition and processes. Geological Society, London, Lithos 112, 556–574. Special Publications 42, 313–345. Mysen, B.O., Boettcher, A.L., 1975. Melting of a hydrous mantle: II. Geochemistry of Sun, W., Li, S., Chen, Y., Li, Y., 2000. Zircon U–Pb dating of granitoids from South Qinling, crystals and liquids formed by anatexis of mantle peridotite at high pressures Central China and their geological significance. Geochimica 29, 209–216 (in Chinese and high temperatures as a function of controlled activities of water, hydrogen, with English Abstract). and carbon dioxide. Journal of Petrology 16, 549–593. Sun, W., Li, S., Chen, Y., Li, Y., 2002. Timing of synorogenic granitoids in the South Nelson, S.T., Montana, A., 1992. Sieve-textured plagioclase in volcanic rocks produced Qinling, Central China: constraints on the evolution of the Qinling–Dabie Orogenic by rapid decompression. American Mineralogist 77, 1242–1249. Belt. Journal of Geology 110, 457–468. Nier, A.O., 1950. A redetermination of the relative abundances of the isotopes of Tatsumi, Y., 1981. Melting experiments on a high-magnesian andesite. Earth and Planetary carbon, nitrogen, oxygen, argon, and potassium. Physical Review 77, 789–793. Science Letters 54, 357–365. Niu, X., Chen, B., Liu, A., Suzuki, K., Ma, X., 2012. Petrological and Sr–Nd–Os isotopic Tatsumi, Y., 1982. Origin of high-magnesian andesites in the Setouchi volcanic belt, constraints on the origin of the Fanshan ultrapotassic complex from the North southwest Japan, II. Melting phase relations at high pressures. Earth and Planetary China Craton. Lithos 149, 146–158. Science Letters 60, 305–317. Nomade, S., Renne, P.R., Vogel, N., Deino, A.L., Sharp, W.D., Becker, T.A., Jaouni, A.R., Mundil, Tatsumi, Y., 2001. Geochemical modeling of partial melting of subducting sediments R., 2005. Alder Creek sanidine (ACs-2): a Quaternary 40Ar/39Ar dating standard tied to and subsequent melt–mantle interaction: generation of high-Mg andesites in the the Cobb Mountain geomagnetic event. Chemical Geology 218, 315–338. Setouchi volcanic belt, southwest Japan. Geology 29, 323–326. Pan, G., Chen, Z., Li, X., Yan, Y., Xu, X., Xu, Q., Jiang, X., Wu, Y., Luo, J., Zhu, T., Peng, Y., Tatsumi, Y., 2006. High-Mg andesites in the Setouchi volcanic belt, southwestern 1997. Geological-Tectonic Evolution in the Eastern Tethys. Geological Press, Beijing, Japan: analogy to archean magmatism and continental crust formation? Annual p. 218 (in Chinese). Review of Earth and Planetary Sciences 34, 467–499. 174 X.-W. Li et al. / Lithos 172–173 (2013) 158–174

Tatsumi, Y., Eggins, S., 1995. Subduction Zone Magmatism. Blackwell Science, Boston Yogodzinski, G.M., Kay, R.W., Volynets, O.N., Koloskov, A.V., Kay, S.M., 1995. Magnesian 211. andesite in the western Aleutian Komandorsky region: implications for slab melting Tatsumi, Y., Hanyu, T., 2003. Geochemical modeling of dehydration and partial melting and processes in the mantle wedge. Geological Society of America Bulletin 107, of subducting lithosphere: toward a comprehensive understanding of high-Mg 505–519. andesite formation in the Setouchi volcanic belt, SW Japan. Geochemistry, Geophysics, Zhang, H.F., Shao, J.A., 2008. Volcanic lavas of the Yixian Formation in western Liaoning Geosystems 4, 1081. province, China: products of lower crust delamination or magma mixing? Acta Thorpe, R.S., 1982. Andesites: Orogenic Andesites and Related Rocks. Wiley, Chichester, Petrologica Sinica 24, 37–48 (in Chinese with English Abstract). p. 697. Zhang, G., Zhang, B., Yuan, X., et al., 2001. Qinling Orogenic Belt and Continental Dynamics Tiepolo, M., Tribuzio, R., Langone, A., 2011. High-Mg andesite petrogenesis by amphibole Beijing. Science Press, Beijing 729 (in Chinese). crystallization and ultramafic crust assimilation: evidence from Adamello Hornblendites Zhang, H.F., Chen, Y.L., Xu, W.C., Liu, R., Yuan, H.L., Liu, X.M., 2006. Granitoids around (Central Alps, Italy). Journal of Petrology 52, 1011–1045. Gonghe basin in Qinghai province: petrogenesis and tectonic implications. Acta Tsuchiya, N., Suzuki, S., Kimura, J.-I., Kagami, H., 2005. Evidence for slab melt/mantle Petrologica Sinica 22, 2910–2922 (in Chinese with English Abstract). reaction: petrogenesis of Early Cretaceous and Eocene high-Mg andesites from Zhang, H., Jin, L., Zhang, L., Nigel, H., Zhou, L., Hu, S., Zhang, B., 2007. Geochemical and the Kitakami Mountains, Japan. Lithos 79, 179–206. Pb–Sr–Nd isotopic compositions of granitoids from western Qinling belt: constraints van Achterbergh, E., Ryan, C.G., Jackson, S.E., Griffin, W.L., 2001. Data reduction software on basement nature and tectonic affinity. Science in China Series D: Earth Sciences 50, for LA–ICPMS. In: Sylvester, P. (Ed.), Laser Ablation ICPMS in the Earth Sciences; 184–196. Principles Applications: Laser Ablation ICPMS in the Earth Sciences, Principles Zhang, H.-F., Parrish, R., Zhang, L., Xu, W.-C., Yuan, H.-L., Gao, S., Crowley, Q.G., 2008. Applications, 29, pp. 239–243. Reply to the comment by Zhang et al. on: “First finding of A-type and adakitic Walker, J.D., Geissman, J.W., Bowring, S.A., Babcock, L.E., 2013. The Geological Society of magmatism association in Songpan–Garze fold belt, eastern Tibetan Plateau: implication America Geologic Time Scale. Geological Society of America Bulletin 125, 259–272. for lithospheric delamination”.Lithos103,565–568. Wang, Y., Huang, J., Luo, M., Tian, J., Bai, Y., 1997. Paleo-Ocean evolution of the southern Zhang, J.-Y., Ma, C.-Q., Xiong, F.-H., Liu, B., 2012. Petrogenesis and tectonic significance eastern Kunlun Orogenic belt during Hercy-Early Indosinian. Journal of China Uni- of the Late Permian–Middle Triassic calc-alkaline granites in the Balong region, versity of Geosciences 22, 369–372 (in Chinese with English Abstract). eastern Kunlun Orogen, China. Geological Magazine 149, 892–908. Wang, H., Zhu, Y., Lin, Q., Li, Y., 2009. Mineral characteristics and tectonic environment Zhao, X., Coe, R.S., 1987. Palaeomagnetic constraints on the collision and rotation of of Longwuxia Gorge ophiolite in Tongren, West Qinling area. Acta Petrologica et North and South China. Nature 327, 141–144. Mineralogica 28, 316–328 (in Chinese with English Abstract). Zhao, J.T., Yang, F.Q., 1992. Permian sedimentary features and environmental analysis Wang, H., Zhu, Y., Lin, Q., Li, Y., Wang, K., 2010. LA–ICP–MS zircon U–Pb dating of the of Maomaolong Formation in Malonggou, Xiahe, Gansu. Acta Sedimentologica gabbro from Longwu Gorge ophiolite, Jianzha–Tongren area, Qinghai, China and Sinica 10, 47–56 (in Chinese with English Abstract). its geological significance. Geological Bulletin of China 29, 86–92. Zhao, Z., Mo, X., Dilek, Y., Niu, Y., DePaolo, D.J., Robinson, P., Zhu, D., Sun, C., Dong, G., Wang, Q., Li, Z.-X., Chung, S.-L., Wyman, D.A., Sun, Y.-L., Zhao, Z.-H., Zhu, Y.-T., Qiu, H.-N., Zhou, S., Luo, Z., Hou, Z., 2009a. Geochemical and Sr–Nd–Pb–O isotopic compositions 2011. Late Triassic high-Mg andesite/dacite suites from northern Hohxil, North of the post-collisional ultrapotassic magmatism in SW Tibet: petrogenesis and impli- Tibet: geochronology, geochemical characteristics, petrogenetic processes and tec- cations for India intra-continental subduction beneath southern Tibet. Lithos 113, tonic implications. Lithos 126, 54–67. 190–212. Winchester, J.A., Floyd, P.A., 1977. Geochemical discrimination of different magma Zhao, Z., Wang, Q., Xiong, X., Niu, H., Zhang, H., Qiao, Y., 2009b. Magnesian andesites in series and their differentiation products using immobile elements. Chemical Geology north Xinjiang, China. International Journal of Earth Sciences 98, 1325–1340. 20, 325–343. Zheng, D.W., Zhang, P.Z., Wan, J.L., Li, D.M., Wang, F., Yuan, D.Y., Zhang, G.L., 2004. The Winther, K.T., Newton, R.C., 1991. Experimental melting of hydrous low-K tholeiite: Ar-40/Ar-39, fission track evidence of Mesozoic tectonic in northern margin of evidence on the origin of Archaean cratons. Bulletin of the Geological Society of West Qinling Mountain. Acta Petrologica Sinica 20, 697–706 (in Chinese with Denmark 39, 213–228. English Abstract). Wood, B.J., Turner, S.P., 2009. Origin of primitive high-Mg andesite: constraints from Zheng, J.P., Griffin, W.L., Sun, M., O'Reilly, S.Y., Zhang, H.F., Zhou, H.W., Xiao, L., Tang, natural examples and experiments. Earth and Planetary Science Letters 283, 59–66. H.Y., Zhang, Z.H., 2010. Tectonic affinity of the West Qinling terrane (central Wyllie, P.J., Wolf, M.B., 1993. Amphibolite dehydration-melting: sorting out the solidus. China): North China or Yangtze? Tectonics 29 TC2009. Geological Society, London, Special Publications 76, 405–416. Zhou, D., Graham, S.A., 1996. The Songpan–Ganzi complex of the West Qinling Shan as Xiong, F., Ma, C., Zhang, J., Liu, B., 2011. LA–ICP–MS zircon U–Pb dating, elements and a Triassic remnant ocean basin. In: Yin, A., Harrison, T.M. (Eds.), The Tectonic Evolution Sr–Nd–Hf isotope geochemistry of the Early Mesozoic mafic dyke swarms in East of Asia. Cambridge University Press, New York, pp. 281–299. Kunlun Orogenic Belt. Acta Petrologica Sinica 27, 3350–3364 (in Chinese with Zhu, Y., Zhou, J., Zeng, Y., 2007. The Tianger (Bingdaban) shear zone hosted gold deposit, English Abstract). west Tianshan, NW China: petrographic and geochemical characteristics. Ore Geology Yan, Z., Wang, Z., Li, J., Xu, Z., Deng, J., 2012. Tectonic settings and accretionary orogenesis Reviews 32, 337–365. of the West Qinling Terrane, northeastern margin of the Tibet Plateau. Acta Zhu, L.-M., Zhang, G.-W., Chen, Y.-J., Ding, Z.-J., Guo, B., Wang, F., Lee, B., 2011. Zircon U–Pb Petrologica Sinica 28, 1808–1828 (in Chinese with English Abstract). ages and geochemistry of the Wenquan Mo-bearing granitioids in West Qinling, Yang, F., Yin, H., Yang, H., Lai, X., 1994. The Songpan–Garzê massif: its relationship with China: constraints on the geodynamic setting for the newly discovered Wenquan the Qinling fold belt and Yangtze platform and development. Acta Geologica Sinica Mo deposit. Ore Geology Reviews 39, 46–62. 68, 208–218 (in Chinese with English Abstract). Zhu, L., Zhang, G., Yang, T., Wang, F., Gong, H., 2013. Geochronology, petrogenesis and Yang, J., Shi, R., Wu, C., Wang, X., Robinson, P., 2009. Dur'ngoi ophiolite in East Kunlun, tectonic implications of the Zhongchuan granitic pluton in the Western Qinling Northeast Tibetan plateau: evidence for paleo-Tethyan suture in Northwest China. metallogenic belt, China. Geological Journal (in press). Journal of Earth Science 20, 303–331. Yin, A., Nie, S., 1993. An indentation model for the North and South China collision and the development of the Tan–Lu and Honam Fault Systems, eastern Asia. Tectonics 12, 801–813.