The Magmatic Plumbing System for Mesozoic High-Mg Andesites, Garnet-bearing Dacites and Porphyries, Rhyolites and Leucogranites from West Qinling, Central China

The Magmatic Plumbing System for Mesozoic High-Mg Andesites, Garnet-bearing Dacites and... ABSTRACT An integrated study of the petrography, mineral composition, zircon geochronology, whole-rock geochemistry and Sr–Nd–Hf isotopes was carried out for an unusual suite of igneous rocks, including high-Mg andesites, garnet-bearing dacites and porphyries, rhyolites and leucogranites, from West Qinling, central China. These data, particularly observations from garnets, are used to demonstrate the petrogenetic links among the associated magmatic components which eventually formed the observed lithologies, evaluate the influence of recycling crystal populations and reconstruct the whole magmatic plumbing system. The crystallization ages of these igneous rocks are ∼239–244 Ma. The high-Mg andesites are phenocryst-rich and characterized by high Mg# (>40; Mg#=100*mol. MgO/(MgO + FeO)), Cr and Ni abundances, low Sr/Y and (La/Yb)N ratios, and relatively high ISr and negative εNd(t) and εHf(t) values. The petrography, mineral chemistry and geochemical data indicate that the high-Mg andesites were generated by mixing between mantle-derived magmas and crustal melts, with subsequent entrainment of xenocrysts (e.g. high Mg# pyroxenes, high An plagioclase and some glomerocrysts) from various sources within the crust. The chemical compositions of the garnet-bearing dacites and porphyries and crystal-poor rhyolites define a common differentiation trend. They become more strongly peraluminous and have more evolved Sr–Nd–Hf isotopic compositions with increasing SiO2 content. Petrological and geochemical data indicate that these peraluminous magmas were likely produced by fractional crystallization of andesitic magma, accompanied by assimilation of crustal materials and, or, entrainment of various phenocryst/xenocryst assemblages. Four types of garnets have been identified, including antecrysts, orthocrysts, peritectic phases and xenocrysts, and the variations in mineral composition and inclusion assemblage indicate a complicated history of magma mixing and mineral-melt interaction/re-equilibrium. The leucogranites are strongly depleted in HREE ((La/Yb)N > 300) and show remarkable negative Eu anomalies (Eu/Eu* = 0·42–0·52). These geochemical features are indicative of the presence of both residual garnet and plagioclase in the magma source resulting from muscovite dehydration melting of metapelitic rocks. Together, all these observations consistently reflect magma evolution in several dispersed but interconnected magma reservoirs which formed a complicated trans-crustal magmatic plumbing system. Local magma compositions have been influenced by multiple processes, including crystallization and accumulation, recharging, anatexis, magma mixing and mingling, assimilation, remobilization of crystal mushes and random entrainment of various phenocryst assemblages and crustal xenoliths. Therefore, detailed petrographic information and mineral composition data are needed for interpreting the whole-rock geochemistry properly. The rapid ascent and eruption of crystal-rich and garnet-bearing magmas have been closely associated with an extensional regime in a post-collisional tectonic setting and facilitated by active fault systems. INTRODUCTION The origin of intermediate to silicic magmatic rocks is fundamental for understanding the evolution and differentiation of the continental crust (Annen et al., 2006; Kemp et al., 2007). Geophysical, geochemical, petrological, and volcanological studies have demonstrated that magmatic systems are trans-crustal, dispersed throughout the entire crust, and dominated by crystal mushes (Dahren et al., 2012; Annen et al., 2015; Marsh, 2015; Cashman et al., 2017). Rapidly ascending magmas can contain abundant mush fragments, which can be present as single crystals, crystal clusters (glomerocrysts), cumulate nodules, or restite (Cashman et al., 2017). These crystal populations may have contrasting origins and can be classified as orthocrysts, antecrysts, xenocrysts or peritectic phases (Davidson et al., 2007; Stevens et al., 2007; Jerram & Martin, 2008; Lackey et al., 2012). Orthocrysts are directly crystallized from their host magmas, whereas antecrysts are not in equilibrium with the host magmas but nevertheless related to the magmatic system (Miller et al., 2007; Jerram & Martin, 2008; Bach et al., 2012). Xenocrysts are derived from assimilated xenolithic rocks and peritectic phases form by dehydration melting of metamorphic rocks or xenoliths (Harangi et al., 2001; Stevens et al., 2007; Erdmann et al., 2009). These crystal populations reflect the diverse magma compositions with which they were in equilibrium (Davidson et al., 2007; Jerram & Martin, 2008), and they may provide important insights into the complex magmatic processes and pressure and temperature conditions in which the host magmas have evolved (Jerram & Davidson, 2007; Putirka, 2008a, b). Moreover, the incorporation of complex crystal populations influences final bulk-rock compositions (e.g. Stevens et al., 2007; Smith et al., 2010; Price et al., 2012; Larrea et al., 2013). Thus, evaluating the origin of complex crystal populations in igneous rocks can offer more robust information than can be obtained from whole-rock compositions, revealing the petrogenetic processes and prevailing physicochemical conditions by which the host magmas were generated and evolved, and facilitating the reconstruction of the whole magmatic plumbing system. In this contribution we focus on an unusual suite of garnet-bearing igneous rocks from the West Qinling Orogen, central China (Fig. 1a and b). Garnet is a common mineral in peraluminous dacitic to rhyolitic rocks (e.g. Clemens & Wall, 1984; Gilbert & Rogers, 1989; Zeck, 1992; Cesare, 2000; Acosta-Vigil et al., 2010), but rare in metaluminous andesites worldwide (e.g. Fitton, 1977; Day et al., 1992; Harangi et al., 2001; Kawabata & Takafuji, 2005; Bach et al., 2012). Previous studies have shown that garnets in igneous rocks can occur as orthocrysts, antecrysts, xenocrysts and, or, peritectic phases and usually crystallize at ∼5–12 kbar. Compositional data for garnets may record petrogenetic processes from source to emplacement or eruption (Harangi et al., 2001; Stevens et al., 2007; Acosta-Vigil et al., 2010; Bach et al., 2012; Lackey et al., 2012). The garnet-bearing andesitic rocks in the Hezuo and Maixiu areas from the West Qinling Orogen were first described by Wang & Ding (1990) (Fig. 1c). Recently, Li et al. (2013) suggested that some garnet-free andesites from the Maixiu area are high-Mg andesites. Our observations suggest that garnet-bearing dacites, granodioritic porphyries and pyroclastic rocks are spatially associated with high Mg-andesites, rhyolites and leucogranites (Fig. 1c). We present a comprehensive study of these igneous rocks based on combined petrography and mineral compositions, laser ablation inductively coupled mass spectrometric (LA-ICP-MS) U–Pb zircon dating, whole-rock geochemical and Sr–Nd–Hf isotopic compositions. These rock assemblages provide a rare opportunity to examine the petrogenetic links among these associated igneous rocks, and they provide an opportunity to examine the effects of recycled crystal populations on whole-rock compositions and to reconstruct the magmatic plumbing system in which the various magmas have evolved. Fig. 1 View largeDownload slide (a) Simplified geological map, showing the major tectonic units of China. (b) The distribution of Early Mesozoic magmatic rocks in the Qinling orogen (Feng et al., 2002). (c) Simplified geological map of the studied area. WQ, West Qinling; EQ, East Qinling; DB, Dabie belt; SL, Sulu belt; QD, Qaidam; QL, Qilian belt; KL (EKL), Kunlun belt; NQL, North Qinling; SQL, South Qinling. Pluton names: MB, Miba; MSL; Mishuling; WQ, Wenquan; XH, Xiahe; XK, Xiekeng; SPX, Shuanpengxi; TR, Tongren; JZ, Jianzha; HMH, Heimahe; WQ, Wenquan. Zircon U–Pb ages are indicated in Fig. 1 b (Luo et al., 2015 and references therein). Ages of the Maixiu andesite and Duowa granodiorite in Fig. 1c are from Li et al. (2013, 2015). Fig. 1 View largeDownload slide (a) Simplified geological map, showing the major tectonic units of China. (b) The distribution of Early Mesozoic magmatic rocks in the Qinling orogen (Feng et al., 2002). (c) Simplified geological map of the studied area. WQ, West Qinling; EQ, East Qinling; DB, Dabie belt; SL, Sulu belt; QD, Qaidam; QL, Qilian belt; KL (EKL), Kunlun belt; NQL, North Qinling; SQL, South Qinling. Pluton names: MB, Miba; MSL; Mishuling; WQ, Wenquan; XH, Xiahe; XK, Xiekeng; SPX, Shuanpengxi; TR, Tongren; JZ, Jianzha; HMH, Heimahe; WQ, Wenquan. Zircon U–Pb ages are indicated in Fig. 1 b (Luo et al., 2015 and references therein). Ages of the Maixiu andesite and Duowa granodiorite in Fig. 1c are from Li et al. (2013, 2015). GEOLOGICAL BACKGROUND The West Qinling Orogen is separated from the East Kunlun and Qaidam terranes by the Wenquan-Wahongshang fault to the west, bounded by the Qilian Orogen along the Qinghai Lake-Baoji fault to the north, and separated from the Songpan-Ganze block to the south by the A’nimaque-Mianlue suture zone (Fig. 1). The suture zone contains abundant ophiolite fragments, considered to represent fragments of a Late Palaeozoic Palaeo-Tethys oceanic subducted slab (Xu et al., 2002; Guo et al., 2007). The West Qinling Orogen is primarily covered by Devonian to Cretaceous sedimentary rocks and the Precambrian basement is rarely exposed (Feng et al., 2002; Liu et al., 2008a). Early Mesozoic igneous rocks are widespread in the Qinling Orogen (Fig. 1b). Abundant geochronological and geochemical studies have been performed on the granitoids in East Qinling and its neighbouring areas (Fig. 1b) (Luo et al., 2015, and references therein). Recent studies have revealed that the early Mesozoic igneous rocks in the West Qinling Orogen can be divided into two stages: (1) Middle Triassic (∼246–234 Ma), mainly distributed in the central and west parts of the West Qinling Orogen (Jin et al., 2005; Zhang et al., 2006; Guo et al., 2012; Luo et al., 2012, 2015; Li et al., 2013, 2014); (2) Late Triassic (∼228–205 Ma), widespread throughout the West Qinling Orogen (Fig. 1b) (Zhang et al., 2006, 2007; Qin et al., 2009; Cao et al., 2011). There are two hypotheses for the generation of the Middle Triassic magmatism in West Qinling: (a) an active continental margin setting related to subduction of the A’nimaque oceanic slab (Jin et al., 2005; Guo et al., 2012; Li et al., 2013, 2014); (b) an early-stage post-collisional setting induced by delamination of thickened lithosphere (Zhang et al., 2008) or break-off of the subducted A’nimaque oceanic slab (Zhang et al., 2006; Luo et al., 2012, 2015) after the collision between the West Qinling Orogen and the Songpan-Ganze block. FIELD GEOLOGY AND PETROGRAPHY The Mesozoic intermediate to felsic volcanic rocks are discontinuously distributed in a NW–SE direction over an area 140 km long and 10–16 km wide in the central part of the West Qinling Qrogen, extending from Dewulu, through Saierqingou to Maixiu (Fig. 1c). The garnet-bearing andesitic rocks from Maixiu and Dewulu (Fig. 1c) have been described by Wang & Ding (1990). In this study, the Saierqingou garnet-bearing volcanic rocks, the Fandelongwa garnet-bearing granodioritic porphyries and the Sangke leucogranites are newly recognized rock suites in the Xiahe area (Fig. 1c). Volcanic rocks The Maixiu volcanic rocks The Maixiu volcanic rocks belong to the Maixiu group, which unconformably overlies Lower to Middle Triassic strata and has a total thickness of 2·2–3·2 km and an exposed area of ∼272 km2 (Xu, 1994) (Fig. 1c), The Maixiu group can be divided into three sub-cycles (Xu, 1994). The lower unit (∼2·58 m to 1·12 km thick), located within the southern Maixiu Basin, consists of andesites, tuffaceous and pyroclastic rocks, intercalated with sandstone, slate and/or coal seams at the base overlain by andesites and dacites intercalated with sandstone, slate and pyroclastic rocks (Xu, 1994). The middle unit (> 0·78 km thick), mainly located in the centre of the Maixiu Basin, comprises dacites, dacitic lava breccias and brecciated lavas (Xu, 1994). The upper unit (1·04–1·41 km thick) is widely distributed in the central and northern parts of the Maixiu Basin. The lower part of the upper unit is composed of amphibole-bearing andesites and andesitic pyroclastic rocks, whereas the upper part consists of pyroclastic rocks and clastic sedimentary rocks (Xu, 1994). The andesites described in this study were collected from the lower unit of the Maixiu group. The Maixiu andesites are dark grey with abundant phenocrysts (35–45 vol. %, up to ∼50 vol. %). Petrographically, two distinct groups can be distinguished. One group contains abundant plagioclase (35–40 vol. %), minor clinopyroxene (1–3 vol. %), orthopyroxene (2–5 vol. %) and rare amphibole and biotite, whereas the other group contains less abundant plagioclase (25–35 vol. %) and more abundant clinopyroxene (5–7 vol. %) and orthopyroxene (5–10 vol. %) (Fig. 2a and b). Orthopyroxene is typically 0·2–0·8 mm in size and occurs as subhedral phenocrysts and glomeroporphyritic aggregates (Fig. 2a). Clinopyroxene (0·3–0·7 mm) and plagioclase (0·4–2 mm) are present as euhedral to subhedral phenocrysts. Some plagioclase grains display sieve-textured cores and pristine rims, and they commonly occur as glomeroporphyritic aggregates (Fig. 2b). The groundmass has a hyalopilitic and/or microcrystalline texture with plagioclase as the dominant microphenocryst (Fig. 2a and b). Some samples exhibit minor sericitization, epidotization and chloritization. Fig. 2 View largeDownload slide Representative photomicrographs. (a) and (b) Maixiu (MX) andesite (0929), orthopyroxene and plagioclase glomerocrysts, and plagioclase with sieve-textured core overgrown by a thin rim. (c) Dewulu (DWL) andesite (0993), plagioclase grains show different shapes, compositions and chemical zoning. (d) Saierqingou (SEQG) andesite (09129), orthopyroxene glomerocryst. (e) Saierqingou garnet-bearing dacite (0988), plagioclase grains have different compositions and patchy-zoning. (f) Fangdelongwa (FDLW) garnet-bearing porphyry (0981), plagioclase and quartz grains are zoned. (g) metapelitic enclave in the Fangdelongwa garnet-bearing porphyry (09164). (h) Sangke (SK) leucogranite (09117). (c), (e) and (f) are cathodoluminescence images. Opx, orthopyroxene; Pl, plagioclase; Bio, biotite; Q, quartz; Kf, k-feldspar; Mus, muscovite; Grt, Garnet. Fig. 2 View largeDownload slide Representative photomicrographs. (a) and (b) Maixiu (MX) andesite (0929), orthopyroxene and plagioclase glomerocrysts, and plagioclase with sieve-textured core overgrown by a thin rim. (c) Dewulu (DWL) andesite (0993), plagioclase grains show different shapes, compositions and chemical zoning. (d) Saierqingou (SEQG) andesite (09129), orthopyroxene glomerocryst. (e) Saierqingou garnet-bearing dacite (0988), plagioclase grains have different compositions and patchy-zoning. (f) Fangdelongwa (FDLW) garnet-bearing porphyry (0981), plagioclase and quartz grains are zoned. (g) metapelitic enclave in the Fangdelongwa garnet-bearing porphyry (09164). (h) Sangke (SK) leucogranite (09117). (c), (e) and (f) are cathodoluminescence images. Opx, orthopyroxene; Pl, plagioclase; Bio, biotite; Q, quartz; Kf, k-feldspar; Mus, muscovite; Grt, Garnet. The Dewulu volcanic rocks The Dewulu volcanic rocks, which cover an exposed area of ∼56 km2 and unconformably overlie Lower Permian strata, consist mainly of andesites and dacites with minor pyroclastic rocks and rhyolitic tuffs (Fig. 1c). They are intruded by a porphyritic quartz diorite stock and locally overlain by Jurassic strata. The Dewulu andesites and dacites were sampled in this study. The grey-black andesites are porphyritic, containing abundant plagioclase phenocrysts (35–40 vol. %), and minor pyroxene and amphibole (2–5 vol. %) (Fig. 2c). Plagioclase phenocrysts (0·3–2 mm in length) occur as euhedral long platy crystals and/or irregular porphyroclastic grains, and exhibit variable compositions and complex zoning (e.g. simple, oscillatory, and/or patchy zoning) (Fig. 2c). Small grains of pyroxenes and amphibole (0·2–0·4 mm) have been replaced by chlorite. The matrix has a hyalopilitic texture and consists of microlites of plagioclase, quartz, biotite and opaque minerals (Fig. 2c). The grey dacites display textures similar to those of the andesites, but have different phenocryst assemblages. Their phenocrysts comprise plagioclase (30–35 vol. %), quartz (10–15 vol. %) and minor biotite (2–3 vol. %), with a microlitic matrix consisting of K-feldspar, plagioclase, biotite and magnetite. The Saierqingou volcanic rocks The Saierqingou volcanic rocks, which unconformably overlie Lower Triassic strata, have a total thickness of ∼2·6 km over an area of ∼182 km2 (Fig. 1c). The volcanic succession can be divided into two units from bottom to top. The lower unit is composed of: (1) grey-black porphyritic andesites (∼0·05 km); (2) grey-black, medium-thick bedded siliceous rocks with thin layers of grey-green volcanic breccia (∼0·01 km); (3) yellow-grey rhyolitic tuff and grey-green dacites with minor andesitic breccia (∼0·03 km); and (4) grey-green dacitic breccia and purple red rhyolite. The upper unit consists mainly of purple red rhyolite and porphyries (∼2·05 km), with pyroclastic rocks at the top (∼0·35 km). The porphyritic andesites, garnet-bearing dacites and pyroclastic rocks, and rhyolite were sampled in this study. The Saierqingou grey-black andesites have a porphyritic texture with ∼40 vol. % phenocrysts. Similar to the Maixiu andesites, two groups of andesites can be identified. One group is characterized by abundant plagioclase (30–35 vol. %) and orthopyroxene (5–10 vol. %) (Fig. 2d). The other group is composed mainly of plagioclase (40–45 vol. %) with minor pyroxene, amphibole and biotite (2–3 vol. %). Orthopyroxene grains are generally < 0·5 mm in diameter and some glomeroporphyritic aggregates are also observed (Fig. 2d). Plagioclase (0·3–3 mm in length) mainly forms subhedral to euhedral and lath-shaped crystals, and shows albite twinning or oscillatory zoning (Fig. 2d). The matrix comprises plagioclase and biotite with minor zircon, apatite, and magnetite (Fig. 2d). The grey-green porphyritic garnet-bearing dacites contain ∼40 vol. % of phenocrysts (plagioclase 20–25 vol. %, quartz 10–15 vol. %, biotite 2–5 vol. %, garnet ∼1%) (Fig. 2e). The matrix is composed of plagioclase, K-feldspar and quartz, and minor zircon, apatite, and magnetite. Plagioclase grains (0·2–2 mm in length) show diverse types and some grains have patchy zoning (Fig. 2e). Quartz and biotite are anhedral and have irregular shapes with a diameter of 0·1–0·3 mm. The garnet-bearing pyroclastic rocks are grey-black and grey-green. Their mineral assemblage is comparable to those of the andesites and dacites. The purple red rhyodacites and rhyolites display aphanitic textures and consist mainly of quartz, plagioclase, albite, K-feldspar and biotite, and minor zircon, apatite and magnetite. Hypabyssal intrusive rocks The Fandelongwa garnet-bearing granodioritic porphyry The Fandelongwa garnet-bearing granodioritic porphyry intrudes Lower Triassic strata as a small stock (∼5 km2), which is exposed to the west of Xiahe town (Fig. 1c). The phenocrysts are plagioclase (15–20 vol. %), albite (5–10%), quartz (5–10 vol. %), biotite (7–10 vol. %) and minor garnet (∼1 vol. %) (Fig. 2f), with accessory magnetite, titanite, apatite and zircon. Both plagioclase and quartz have larger sizes (0·5–4 mm) than those in the garnet-bearing dacites. Some plagioclase and quartz grains have been partially resorbed into round or elliptical shapes and have clear core–rim textures (Fig. 2f). Some plagioclase grains also occur in glomeroporphyritic aggregates. Biotite (0·1–0·2 mm) is subhedral to anhedral. Enclaves of wall-rocks and fine-grained diorite and biotite-rich metapelite (Fig. 2g) are locally preserved. They are generally ovoid or ellipsoidal and commonly 2–20 centimetres in size. The Sangke two-mica porphyritic leucogranite The Sangke two-mica, porphyritic leucogranite stock (∼3 km2) crops out to the south of Xiahe town (Fig. 1c). This was emplaced into the Lower Triassic Longwuhe Formation and is covered to the west by Quaternary sediments. K-feldspar (10 vol. %) phenocrysts are generally 0.–2·0 mm in size. The matrix is very fine-grained (0·1–0·2 mm) and composed of quartz (30–35 vol. %), K-feldspar (20–30 vol. %), oligoclase (15–25 vol. %), muscovite (2–3 vol. %), and biotite (3–4 vol. %), with minor zircon, apatite and magnetite (Fig. 2h). Petrography and classification of garnets Different types of garnet are present in the Saierqingou dacites and pyroclastic rocks (Fig. 3), the Fandelongwa porphyries (Fig. 4), as well as in the rare metapelitic xenoliths (Fig. 4) in the Fandelongwa porphyries. According to crystal form, mineral inclusions and zoning pattern, garnets can be classified into the following four types. Fig. 3 View largeDownload slide Garnet in the Saierqingou volcanic rocks. (a) and (b) Type-1 garnets in dacite. (a) resorbed, rounded garnet 0988#1 is surrounded by a plagioclase corona; (b) garnet 0988#2 has a dark core and a pale rim; (c) Type-1 garnet (0984#1) from pyroclastic rock. (d)–(h) Type-2 garnets from dacite (09131). Grains #0 and #1 contain mineral inclusions, and #0 is also surrounded by a cordierite corona (d)–(e). Grains of #2–4 are rounded and cracked and are enclosed by plagioclase (f)–(h). (i) Type-4 garnet (09131#5) is an irregular fragment. Yellow circles represent spots analyzed by EMPA. Zrn, zircon; Ap, apatite; Ru, rutile; Crd, cordierite; MI, melt inclusions; FI, fluid inclusions. Fig. 3 View largeDownload slide Garnet in the Saierqingou volcanic rocks. (a) and (b) Type-1 garnets in dacite. (a) resorbed, rounded garnet 0988#1 is surrounded by a plagioclase corona; (b) garnet 0988#2 has a dark core and a pale rim; (c) Type-1 garnet (0984#1) from pyroclastic rock. (d)–(h) Type-2 garnets from dacite (09131). Grains #0 and #1 contain mineral inclusions, and #0 is also surrounded by a cordierite corona (d)–(e). Grains of #2–4 are rounded and cracked and are enclosed by plagioclase (f)–(h). (i) Type-4 garnet (09131#5) is an irregular fragment. Yellow circles represent spots analyzed by EMPA. Zrn, zircon; Ap, apatite; Ru, rutile; Crd, cordierite; MI, melt inclusions; FI, fluid inclusions. Fig. 4 View largeDownload slide View largeDownload slide Garnet in the Fangdelongwa granodioritic porphyry. Type-1 garnets: (a) resorbed garnet 0981#1; (b) garnet 0981#2 has a dark red core and a light-coloured rim. Type-4 garnets: (c) and (d) garnets 09163#1 and 09163#2 are surrounded by reaction coronae of albite and biotite, and 09163#1 has a dusty core and a clear rim; (e) and (f) garnet 09164#4 has been extensively replaced by biotite and chlorite. (g)–(k) Type-3 garnets occur as clusters in a metapelitic xenolith. (g) metapelitic xenolith in the porphyry is surrounded by glomerocrysts of plagioclase. (h)–(k) garnets (09164#1–3) have dusty cores and clear rims. Yellow circles represent spots analyzed by EMPA and red circles represent spots analyzed by LA-ICP-MS. Pl, plagioclase; Bio, biotite; Chl,chlorite; Sil, sillimanite; Ab,albite. Fig. 4 View largeDownload slide View largeDownload slide Garnet in the Fangdelongwa granodioritic porphyry. Type-1 garnets: (a) resorbed garnet 0981#1; (b) garnet 0981#2 has a dark red core and a light-coloured rim. Type-4 garnets: (c) and (d) garnets 09163#1 and 09163#2 are surrounded by reaction coronae of albite and biotite, and 09163#1 has a dusty core and a clear rim; (e) and (f) garnet 09164#4 has been extensively replaced by biotite and chlorite. (g)–(k) Type-3 garnets occur as clusters in a metapelitic xenolith. (g) metapelitic xenolith in the porphyry is surrounded by glomerocrysts of plagioclase. (h)–(k) garnets (09164#1–3) have dusty cores and clear rims. Yellow circles represent spots analyzed by EMPA and red circles represent spots analyzed by LA-ICP-MS. Pl, plagioclase; Bio, biotite; Chl,chlorite; Sil, sillimanite; Ab,albite. Type-1 garnets, which are subhedral to euhedral and usually 0·3–5 mm in diameter, occur as disseminated crystals in the Saierqingou volcanic rocks (Fig. 3) and the Fandelongwa porphyries (Fig. 4a and b). They are characterized by abundant inclusions (e.g. ilmenite, zircon and apatite), zoning of colour and inclusions and some reaction and resorption textures. A rounded and resorbed garnet 0988#1 is surrounded by a reaction corona consisting of plagioclase (Fig. 3a). Garnet 0988#2 has a dark red core containing inclusions of zircon and apatite and a light red rim with glass inclusions (Fig. 3b). Garnets 0984#1 and 0981#1 have also been resorbed and have dentated margins (Figs 3c and 4a). Garnet 0981#2 is intergrown with plagioclase and biotite and has a dark-red core and a light-colored rim (Fig. 4b). Type-2 garnets, mainly found in the Saierqingou dacites, are anhedral to subhedral, and mostly 0·1–1 mm in diameter and are lighter in colour than Type-1 garnets (Fig. 3d–h). Their cross sections are pristine and inclusion poor with only the occasional zircon, apatite or rutile inclusion (Fig. 3d and e). Most of the resorbed Type-2 garnets are spherical or roughly spherical, and some grains occur as inclusions in plagioclase (Fig. 3f–h). Garnet 09131#0 is surrounded by a reaction corona of cordierite (Fig. 3d). Type-3 garnets are anhedral to subhedral, 0·3 to 3 mm in diameter and occur as clusters in the metapelitic xenoliths from the Fandelongwa porphyries (Figs 2g and 4g–k). A typical metapelitic xenolith is composed of biotite, quartz, albite, garnet, ilmenite and rutile and surrounded by glomerocrysts of plagioclase (Figs 2g and 4g). The cores of the garnets are dusty, containing abundant needle-like sillimanite inclusions, but the outer zones are pristine and inclusion free (Fig. 4h–k). Type-4 garnets are 0·1 to 5 mm in diameter and occur as anhedral to subhedral crystals in the Saierqingou dacites and the Fandelongwa granodioritic porphyries. Garnet 09131#5 forms irregular fragments containing abundant fluid inclusions (Fig. 3i). Garnets 09163#1 (Fig. 4c) and 09163#2 (Fig. 4d) are surrounded by reaction coronae of albite and biotite. Garnet 09163#1 has a dusty and obvious cracked core but a relatively light-coloured rim (Fig. 4c). Garnet 09164#4 has been extensively replaced by biotite and chlorite (Fig. 4e and f). ANALYTICAL METHODS Twenty-nine fresh samples were crushed in a steel crusher and then powdered in an agate mill to a grain size <200 mesh. Major element abundances in whole-rocks were measured at the State Key Laboratory of Biogeology and Environmental Geology, China University of Geosciences, Wuhan, using a Shimadzu X-1800 sequential X-ray fluorescence spectrometer, with an Rh-anode, a total voltage of 50 kV, and a current of 50 mA. Sample powder (0·7 g), LiF (0·6 g), NH4NO3 (0·3 g) and 4 drops of LiBr (1·5%), were fused in a high-frequency furnace for 11 min at ∼1000°C in 95%Pt–5%Au crucibles. The material was swirled repeatedly to ensure complete dissolution and homogenization, and then poured into a mould to form a thin, flat-surfaced disc. The loss-on-ignition (LOI) was measured on dried sample powder by heating in a pre-heated crucible to 105°C for 2 hours. The analytical uncertainty for major elements is generally <5%. The data quality was monitored by analysis of Chinese National standards (Supplementary Data Table S1; supplementary data are available for downloading at http://www.petrology.oxfordjournals.org). Trace element abundances in whole rock samples were measured using an Agilent 7500a ICP-MS at the State Key Laboratory of Geological Processes and Mineral Resources (GPMR), China University of Geosciences, Wuhan. Analytical precision for most elements is better than 5%. Analyses of trace element for the reference standards are given in Supplementary Data Table S2. Analytical procedures, precision and accuracy were the same as described by Liu et al. (2008c). Nineteen samples were selected for analysis of their Sr–Nd isotopic compositions. Whole-rock Sr and Nd isotopic ratios were measured on a Triton thermal ionization mass spectrometer at the GPMR. During the period of analysis, the GBW04411 standard yielded an 87Sr/86Sr value of 0·760227 ± 4 (2σ) and the BCR-2 standard gave a 143Nd/144Nd value of 0·51264 ± 4 (2σ). For details of the Sr and Nd isotopic analytical procedures see Gao et al. (2004). Mineral compositions for garnet and plagioclase were determined at the GPMR, using a JEOL JXA-8100 Electron Probe Micro Analyzer (EPMA). An accelerating voltage of 15 kV, a beam current of 20 nA, a 5 µm focused electron beam and a counting time of 30 s were used to analyze minerals. Major elements of some garnets were also obtained by LA-ICP-MS at the GPMR. Operating conditions and data reduction methods are the same as those described by Liu et al. (2008b). Applying an ablation yield correction factor and using NIST SRM 610 as a reference material for external calibration, analyses of the USGS reference glasses BCR-2 G, BHVO-2 G and BIR-1 G produced results that generally agree with recommended values within 5·5% for major elements (relative standard deviation =1·0–5·5% except for K2O and P2O5, n = 15) (Supplementary Data Table S3). Cathode-luminescence (CL) images, taken at the GPMR, were used to check the internal textures of individual zircon and to guide U–Pb dating and Hf isotope analysis. U–Pb dating and trace element analyses of zircon were carried out synchronously by LA-ICP-MS at the GPMR. A beam diameter of 32 µm was used. Zircon 91500 was used as an external standard. Trace elements (Ti concentration) were calibrated using NIST SRM 610 as an external standard and using ZrO2 = 66·1 wt % as an internal standard (Liu et al., 2010). Zircon Hf isotope analysis was carried out in situ using a Neptune MC-ICP-MS at the GPMR. The analyses were conducted with a beam diameter of 44 μm for larger grains and 32 μm for small grains. The corrections and analytical precision and accuracy are as described by Luo et al. (2015). Initial εHf(t) was calculated relative to the chondritic reservoir with a 176Hf/177Hf ratio of 0·282772 and 176Lu/177Hf of 0·0332 (Blichert-Toft et al., 1997). RESULTS Mineral chemistry Representative mineral composition data are given in Table 1. The chemical compositions of plagioclase and garnet analyzed in this study are given in Supplementary Data Tables S4 and S5. The results are briefly summarized below (Figs 5–7). Table 1 Representative of major element (wt %) for plagioclase, clinopyroxene, orthopyroxene and garnet Mineral plagioclase clinopyroxene orthopyroxene Sample andesite* garnet-beaing dacite (0988) andesite* Type phenocryst matrix phenocryst phenocryst rims cores rims cores Spot-no n=23 n=11 n=18 n=4 n=7 n=14 n=49 SiO2 47·96 50·68 60·88 58·12 59·76 62·13 52·24 52·14 52·68 55·66 56·19 TiO2 0·03 0·05 0·02 0·02 0·04 0·01 0·40 0·40 0·20 0·10 0·05 Al2O3 32·65 31·93 23·86 25·97 25·05 23·21 2·10 2·71 2·65 1·63 2·13 FeO 0·34 0·45 0·09 0·29 0·05 0·03 7·08 8·22 5·22 10·46 7·12 Cr2O3 0·43 0·24 0·74 0·47 0·66 MnO 0·01 0·02 – 0·01 – – 0·15 0·08 0·08 0·12 0·07 NiO 0·03 0·05 0·01 0·03 0·07 MgO 0·11 0·14 0·03 0·04 0·01 0·03 17·71 15·48 17·25 29·85 32·26 CaO 15·69 13·38 6·19 8·42 7·32 5·42 19·76 20·95 21·29 1·87 1·35 Na2O 2·48 3·36 7·48 5·97 6·71 7·64 0·22 0·24 0·19 0·03 0·04 K2O 0·17 0·24 0·78 0·40 0·57 0·85 0·01 0·00 0·00 0·00 0·00 Total 99·47 100·26 99·32 99·24 99·49 99·30 100·13 100·49 100·33 100·23 99·95 End (%) An 77 68 30 43 36 27 Wo 39 43 43 4 3 Ab 22 31 66 55 60 68 En 49 44 49 80 87 Or 1 1 4 2 3 5 Fs 11 13 8 16 11 Mg# 82 80 88 84 89 Mineral Type-1 garnet Type-2 garnet Type-3 garnet Type-4 garnet Sample dactie(0988#2) porphyry(0981#2) dacite(09131#1-4) metapelitic xenolith porphyry(09163#1) porphyry(09164#4) Type cores rims core rims cores rims cores rims cores rims core rims Spot-no n=3 n=3 n=1 n=5 n=5 n=5 n=8 n=8 n=4 n=6 n=1 n=7 SiO2 37·02 36·84 36·20 35·70 37·87 37·90 35·48 35·34 36·43 36·61 37·92 37·79 TiO2 0·24 0·07 0·48 0·43 0·18 0·25 0·06 0·10 0·08 0·19 0·1 0·08 Al2O3 20·38 20·69 22·20 21·74 21·22 21·36 22·33 21·98 21·98 22·22 21·19 21·08 FeO 34·53 34·20 28·80 30·80 31·41 31·35 32·66 30·23 33·41 29·38 28·37 27·05 MnO 1·22 2·80 1·44 2·79 1·31 1·06 1·20 2·92 2·04 1·47 2·55 3·99 MgO 1·11 1·26 4·50 1·97 3·85 4·87 5·53 3·65 4·93 5·02 4·62 3·14 CaO 4·76 3·31 6·13 6·35 3·21 2·36 2·26 5·65 0·82 4·90 4·8 6·36 Na2O 0·03 0·03 0·01 0·01 0·04 0·02 0·01 0·01 0·03 0·01 – 0·03 K2O 0·02 0·00 0·01 0·00 0·01 0·01 0·00 0·00 0·00 0·00 – 0·01 Total 99·28 99·20 99·85 99·86 99·09 99·22 99·56 99·90 99·83 99·82 99·54 99·53 End (%) Pyr 5 5 17 8 15 19 21 14 19 19 18 12 Spe 3 6 3 6 3 2 3 6 4 3 6 9 Gro 13 10 17 18 9 7 6 15 2 13 13 18 Alm 79 78 61 67 70 69 69 64 72 62 62 60 Mineral plagioclase clinopyroxene orthopyroxene Sample andesite* garnet-beaing dacite (0988) andesite* Type phenocryst matrix phenocryst phenocryst rims cores rims cores Spot-no n=23 n=11 n=18 n=4 n=7 n=14 n=49 SiO2 47·96 50·68 60·88 58·12 59·76 62·13 52·24 52·14 52·68 55·66 56·19 TiO2 0·03 0·05 0·02 0·02 0·04 0·01 0·40 0·40 0·20 0·10 0·05 Al2O3 32·65 31·93 23·86 25·97 25·05 23·21 2·10 2·71 2·65 1·63 2·13 FeO 0·34 0·45 0·09 0·29 0·05 0·03 7·08 8·22 5·22 10·46 7·12 Cr2O3 0·43 0·24 0·74 0·47 0·66 MnO 0·01 0·02 – 0·01 – – 0·15 0·08 0·08 0·12 0·07 NiO 0·03 0·05 0·01 0·03 0·07 MgO 0·11 0·14 0·03 0·04 0·01 0·03 17·71 15·48 17·25 29·85 32·26 CaO 15·69 13·38 6·19 8·42 7·32 5·42 19·76 20·95 21·29 1·87 1·35 Na2O 2·48 3·36 7·48 5·97 6·71 7·64 0·22 0·24 0·19 0·03 0·04 K2O 0·17 0·24 0·78 0·40 0·57 0·85 0·01 0·00 0·00 0·00 0·00 Total 99·47 100·26 99·32 99·24 99·49 99·30 100·13 100·49 100·33 100·23 99·95 End (%) An 77 68 30 43 36 27 Wo 39 43 43 4 3 Ab 22 31 66 55 60 68 En 49 44 49 80 87 Or 1 1 4 2 3 5 Fs 11 13 8 16 11 Mg# 82 80 88 84 89 Mineral Type-1 garnet Type-2 garnet Type-3 garnet Type-4 garnet Sample dactie(0988#2) porphyry(0981#2) dacite(09131#1-4) metapelitic xenolith porphyry(09163#1) porphyry(09164#4) Type cores rims core rims cores rims cores rims cores rims core rims Spot-no n=3 n=3 n=1 n=5 n=5 n=5 n=8 n=8 n=4 n=6 n=1 n=7 SiO2 37·02 36·84 36·20 35·70 37·87 37·90 35·48 35·34 36·43 36·61 37·92 37·79 TiO2 0·24 0·07 0·48 0·43 0·18 0·25 0·06 0·10 0·08 0·19 0·1 0·08 Al2O3 20·38 20·69 22·20 21·74 21·22 21·36 22·33 21·98 21·98 22·22 21·19 21·08 FeO 34·53 34·20 28·80 30·80 31·41 31·35 32·66 30·23 33·41 29·38 28·37 27·05 MnO 1·22 2·80 1·44 2·79 1·31 1·06 1·20 2·92 2·04 1·47 2·55 3·99 MgO 1·11 1·26 4·50 1·97 3·85 4·87 5·53 3·65 4·93 5·02 4·62 3·14 CaO 4·76 3·31 6·13 6·35 3·21 2·36 2·26 5·65 0·82 4·90 4·8 6·36 Na2O 0·03 0·03 0·01 0·01 0·04 0·02 0·01 0·01 0·03 0·01 – 0·03 K2O 0·02 0·00 0·01 0·00 0·01 0·01 0·00 0·00 0·00 0·00 – 0·01 Total 99·28 99·20 99·85 99·86 99·09 99·22 99·56 99·90 99·83 99·82 99·54 99·53 End (%) Pyr 5 5 17 8 15 19 21 14 19 19 18 12 Spe 3 6 3 6 3 2 3 6 4 3 6 9 Gro 13 10 17 18 9 7 6 15 2 13 13 18 Alm 79 78 61 67 70 69 69 64 72 62 62 60 * , plagioclase, clinopyroxene and orthopyroxene in andesite from Li et al. (2013) Table 1 Representative of major element (wt %) for plagioclase, clinopyroxene, orthopyroxene and garnet Mineral plagioclase clinopyroxene orthopyroxene Sample andesite* garnet-beaing dacite (0988) andesite* Type phenocryst matrix phenocryst phenocryst rims cores rims cores Spot-no n=23 n=11 n=18 n=4 n=7 n=14 n=49 SiO2 47·96 50·68 60·88 58·12 59·76 62·13 52·24 52·14 52·68 55·66 56·19 TiO2 0·03 0·05 0·02 0·02 0·04 0·01 0·40 0·40 0·20 0·10 0·05 Al2O3 32·65 31·93 23·86 25·97 25·05 23·21 2·10 2·71 2·65 1·63 2·13 FeO 0·34 0·45 0·09 0·29 0·05 0·03 7·08 8·22 5·22 10·46 7·12 Cr2O3 0·43 0·24 0·74 0·47 0·66 MnO 0·01 0·02 – 0·01 – – 0·15 0·08 0·08 0·12 0·07 NiO 0·03 0·05 0·01 0·03 0·07 MgO 0·11 0·14 0·03 0·04 0·01 0·03 17·71 15·48 17·25 29·85 32·26 CaO 15·69 13·38 6·19 8·42 7·32 5·42 19·76 20·95 21·29 1·87 1·35 Na2O 2·48 3·36 7·48 5·97 6·71 7·64 0·22 0·24 0·19 0·03 0·04 K2O 0·17 0·24 0·78 0·40 0·57 0·85 0·01 0·00 0·00 0·00 0·00 Total 99·47 100·26 99·32 99·24 99·49 99·30 100·13 100·49 100·33 100·23 99·95 End (%) An 77 68 30 43 36 27 Wo 39 43 43 4 3 Ab 22 31 66 55 60 68 En 49 44 49 80 87 Or 1 1 4 2 3 5 Fs 11 13 8 16 11 Mg# 82 80 88 84 89 Mineral Type-1 garnet Type-2 garnet Type-3 garnet Type-4 garnet Sample dactie(0988#2) porphyry(0981#2) dacite(09131#1-4) metapelitic xenolith porphyry(09163#1) porphyry(09164#4) Type cores rims core rims cores rims cores rims cores rims core rims Spot-no n=3 n=3 n=1 n=5 n=5 n=5 n=8 n=8 n=4 n=6 n=1 n=7 SiO2 37·02 36·84 36·20 35·70 37·87 37·90 35·48 35·34 36·43 36·61 37·92 37·79 TiO2 0·24 0·07 0·48 0·43 0·18 0·25 0·06 0·10 0·08 0·19 0·1 0·08 Al2O3 20·38 20·69 22·20 21·74 21·22 21·36 22·33 21·98 21·98 22·22 21·19 21·08 FeO 34·53 34·20 28·80 30·80 31·41 31·35 32·66 30·23 33·41 29·38 28·37 27·05 MnO 1·22 2·80 1·44 2·79 1·31 1·06 1·20 2·92 2·04 1·47 2·55 3·99 MgO 1·11 1·26 4·50 1·97 3·85 4·87 5·53 3·65 4·93 5·02 4·62 3·14 CaO 4·76 3·31 6·13 6·35 3·21 2·36 2·26 5·65 0·82 4·90 4·8 6·36 Na2O 0·03 0·03 0·01 0·01 0·04 0·02 0·01 0·01 0·03 0·01 – 0·03 K2O 0·02 0·00 0·01 0·00 0·01 0·01 0·00 0·00 0·00 0·00 – 0·01 Total 99·28 99·20 99·85 99·86 99·09 99·22 99·56 99·90 99·83 99·82 99·54 99·53 End (%) Pyr 5 5 17 8 15 19 21 14 19 19 18 12 Spe 3 6 3 6 3 2 3 6 4 3 6 9 Gro 13 10 17 18 9 7 6 15 2 13 13 18 Alm 79 78 61 67 70 69 69 64 72 62 62 60 Mineral plagioclase clinopyroxene orthopyroxene Sample andesite* garnet-beaing dacite (0988) andesite* Type phenocryst matrix phenocryst phenocryst rims cores rims cores Spot-no n=23 n=11 n=18 n=4 n=7 n=14 n=49 SiO2 47·96 50·68 60·88 58·12 59·76 62·13 52·24 52·14 52·68 55·66 56·19 TiO2 0·03 0·05 0·02 0·02 0·04 0·01 0·40 0·40 0·20 0·10 0·05 Al2O3 32·65 31·93 23·86 25·97 25·05 23·21 2·10 2·71 2·65 1·63 2·13 FeO 0·34 0·45 0·09 0·29 0·05 0·03 7·08 8·22 5·22 10·46 7·12 Cr2O3 0·43 0·24 0·74 0·47 0·66 MnO 0·01 0·02 – 0·01 – – 0·15 0·08 0·08 0·12 0·07 NiO 0·03 0·05 0·01 0·03 0·07 MgO 0·11 0·14 0·03 0·04 0·01 0·03 17·71 15·48 17·25 29·85 32·26 CaO 15·69 13·38 6·19 8·42 7·32 5·42 19·76 20·95 21·29 1·87 1·35 Na2O 2·48 3·36 7·48 5·97 6·71 7·64 0·22 0·24 0·19 0·03 0·04 K2O 0·17 0·24 0·78 0·40 0·57 0·85 0·01 0·00 0·00 0·00 0·00 Total 99·47 100·26 99·32 99·24 99·49 99·30 100·13 100·49 100·33 100·23 99·95 End (%) An 77 68 30 43 36 27 Wo 39 43 43 4 3 Ab 22 31 66 55 60 68 En 49 44 49 80 87 Or 1 1 4 2 3 5 Fs 11 13 8 16 11 Mg# 82 80 88 84 89 Mineral Type-1 garnet Type-2 garnet Type-3 garnet Type-4 garnet Sample dactie(0988#2) porphyry(0981#2) dacite(09131#1-4) metapelitic xenolith porphyry(09163#1) porphyry(09164#4) Type cores rims core rims cores rims cores rims cores rims core rims Spot-no n=3 n=3 n=1 n=5 n=5 n=5 n=8 n=8 n=4 n=6 n=1 n=7 SiO2 37·02 36·84 36·20 35·70 37·87 37·90 35·48 35·34 36·43 36·61 37·92 37·79 TiO2 0·24 0·07 0·48 0·43 0·18 0·25 0·06 0·10 0·08 0·19 0·1 0·08 Al2O3 20·38 20·69 22·20 21·74 21·22 21·36 22·33 21·98 21·98 22·22 21·19 21·08 FeO 34·53 34·20 28·80 30·80 31·41 31·35 32·66 30·23 33·41 29·38 28·37 27·05 MnO 1·22 2·80 1·44 2·79 1·31 1·06 1·20 2·92 2·04 1·47 2·55 3·99 MgO 1·11 1·26 4·50 1·97 3·85 4·87 5·53 3·65 4·93 5·02 4·62 3·14 CaO 4·76 3·31 6·13 6·35 3·21 2·36 2·26 5·65 0·82 4·90 4·8 6·36 Na2O 0·03 0·03 0·01 0·01 0·04 0·02 0·01 0·01 0·03 0·01 – 0·03 K2O 0·02 0·00 0·01 0·00 0·01 0·01 0·00 0·00 0·00 0·00 – 0·01 Total 99·28 99·20 99·85 99·86 99·09 99·22 99·56 99·90 99·83 99·82 99·54 99·53 End (%) Pyr 5 5 17 8 15 19 21 14 19 19 18 12 Spe 3 6 3 6 3 2 3 6 4 3 6 9 Gro 13 10 17 18 9 7 6 15 2 13 13 18 Alm 79 78 61 67 70 69 69 64 72 62 62 60 * , plagioclase, clinopyroxene and orthopyroxene in andesite from Li et al. (2013) Fig. 5 View largeDownload slide Summary of mineral composition ranges analyzed by EMPA. (a) Clinopyroxene Mg# and (b) orthopyroxene Mg# in andesites described by Li et al. (2013). (c) Plagioclase An content in andesites described by Wang & Ding (1990) and Li et al. (2013) and in the garnet-bearing dacites and porphyries described in this study. Fig. 5 View largeDownload slide Summary of mineral composition ranges analyzed by EMPA. (a) Clinopyroxene Mg# and (b) orthopyroxene Mg# in andesites described by Li et al. (2013). (c) Plagioclase An content in andesites described by Wang & Ding (1990) and Li et al. (2013) and in the garnet-bearing dacites and porphyries described in this study. Fig. 6 View largeDownload slide End-member composition of garnet in rocks described in this study. (a) grossular (Gro) v. spessartine (Spe), and (b) Pyrope (Pyr) v. almandine (Alm). Fig. 6 View largeDownload slide End-member composition of garnet in rocks described in this study. (a) grossular (Gro) v. spessartine (Spe), and (b) Pyrope (Pyr) v. almandine (Alm). Fig. 7 View largeDownload slide Compositional profiles for garnet described in this study. Type-1: (a) 0988#1, (b) 0988#2, and (c) 0981#2. Type-3: (d) 09164#1. Type-4: (e) 09163#1 and (f) 09164#4. Alm,almandine; Pyr,pyrope; Gro,grossular; Spe,spessartine. Fig. 7 View largeDownload slide Compositional profiles for garnet described in this study. Type-1: (a) 0988#1, (b) 0988#2, and (c) 0981#2. Type-3: (d) 09164#1. Type-4: (e) 09163#1 and (f) 09164#4. Alm,almandine; Pyr,pyrope; Gro,grossular; Spe,spessartine. Pyroxenes and plagioclase Clinopyroxene from the Maixiu andesites are mainly augite in composition and, although most of them display no compositional zoning, rare grains show reverse zoning (Fig. 5a) (Li et al., 2013). Orthopyroxene grains usually display normal compositional zoning with high Mg# cores and mantles and thin, low Mg# rims (Table 1) (Fig. 5b) (Li et al., 2013). Plagioclase phenocrysts from the Maixiu high-Mg andesites are mainly labradorite and bytownite with compositions An68-83, whereas matrix plagioclases exhibit relatively low An60–81 (Table 1) (Fig. 5c) (Li et al., 2013). Plagioclase phenocrysts from the garnet-bearing andesites, with normal and reverse zoning, are andesine and labradorite with An48–61, whereas one matrix plagioclase is oligoclase with An28 (Fig. 5c) (Wang & Ding, 1990). Plagioclase phenocrysts from the garnet-bearing Saierqingou dacites and Fandelongwa porphyries have similar compositions and are oligoclase–andesine with An27–43 (Table 1) (Fig. 5c). Garnet Type-1 garnets show slightly different characteristics in different lithologies. In the Fandelongwa porphyries, garnet 0981#1 (Fig. 4a) and the core of garnet 0981#2 (Fig. 4b) have a homogeneous composition of Alm55–62Pyr17–25Gro12–17Spe3 (Figs 6 and 7c). Garnet 0981#2 (Fig. 4b) shows increasing grossular (Gro) and spessartine (Spe), and decreasing pyrope (Pyr) from core to rim, whereas the almandine (Alm) content initially increases and then decreases from core to rim (Figs 6 and 7c). The rim has a composition of Alm65Pyr7Gro18Spe7. Type-1 garnets 0988#1 (Fig. 3a) and 0988#2 (Fig. 3b) hosted by the Saierqingou dacites have compositions in the range of Alm73–79Pyr5–10Gro12–14Spe2–4 (Fig. 6). Garnet 0988#2 (Fig. 3b) displays increasing Spe and decreasing Gro content from core to rim (Figs 6 and 7b), which is different from the trend observed from 0981#2. Type-2 garnets (09131#0–4) (Fig. 3d–h) show a wide composition range of Alm62–75Pyr10–30Gro3–12Spe1–4 but are unzoned (Fig. 6). The main difference between the Type-1 and Type-2 garnets is that the former has lower Gro (3–12 mol %) and relatively high Pyr (10–30 mol %) contents. Type-3 garnets show a positive correlation between Alm and Pyr contents, which is opposite to that in the Type-1 garnets (Fig. 6). Type-3 garnets have uniform compositional zoning. From cores to rims, Gro and Spe contents increase, and Alm and Pyr contents decrease (Figs 6 and 7d). The core compositions are in the range Alm65–72Pyr18–24Gro4–10Spe1–3, whereas the rims compositions are Alm63–66Pyr11–18Gro12–16Spe5–7. Type-4 garnets vary widely in composition but the variation is not systematic. Garnet 09131#5 is similar in composition to Type-2 garnets (Fig. 6). Garnet 09163#1 has obvious chemical zoning defined by increasing Gro and decreasing Alm and Spe from core to rim. The core has a composition in the range Alm70–73Pyr15–22Gro1–3Spe4–5, but the rim has a composition similar to Type-1 garnets (Figs 6 and 7e). The composition of the core of garnet 09164#4 is similar to those of the rims of the Type-2 garnets, but Spe and Gro are higher in the rims (Figs 6 and 7f). Garnet 09163#2 has the lowest Alm and highest Pyr, with a composition of Alm53–56Pyr30–35Gro9–11Spe1–2 (Fig. 6). Geochronology Five samples were selected for LA-ICP-MS zircon U–Pb dating and the results are listed in Supplementary Data Table S6. As shown by representative zircon CL images (Supplementary Data Fig. S1), zircons from all the samples are short to long prismatic and 80 to 300 μm in length. Most of them show clear magmatic oscillatory zoning and have low U (64–855 ppm) and Th (46–3333 ppm) contents and high Th/U ratios (0·1–0·8). For the Dewulu dacite (0991), one analysis yields an inherited 206Pb/238 U age of 266 Ma. The remaining eighteen analyses yield 206Pb/238 U ages between 234 and 246 Ma, with a weighted mean of 239 ± 2 Ma (2σ; MSWD = 2) (Fig. 8a). Zircons from two Saierqingou samples (0984 and 09133) are similar (Supplementary Data Fig. S1). Twenty-one analyses from 0984 (a garnet-bearing pyroclastic rock) yield 206Pb/238 U ages ranging from 232 to 252 Ma, with a weighted mean of 243 ± 3 Ma (2σ; MSWD = 3·3) (Fig. 8b). Fourteen analyses of 09133 (andesite) yielded 206Pb/238 U ages between 237 and 246 Ma, with a weighted mean of 241 ± 2 Ma (2σ; MSWD = 2·5) (Fig. 8c). Four inherited grains from the two samples give old ages of 259 Ma, 270 Ma, 362 Ma and 1917 Ma and one spot gives a younger age of 227 Ma (Fig. 8b and c); these anomalous ages probably reflect inheritance and lead loss, respectively. Fig. 8 View largeDownload slide Zircon U–Pb Concordia diagrams. (a) Dewulu (DWL) dacite (0991); (b) Saierqingou (SEQG) garnet-bearing pyroclastic rock (0984); (c) Saierqingou andesite (09133); (d) Fangdelongwa (FDLW) garnet-bearing granodioritic porphyry (0981); (e) and (f) Sangke (SK) leucogranite (09117). Shaded and dashed ellipses are not included in the age calculation. Ellipses represent 1-sigma uncertainty for individual analyses. Fig. 8 View largeDownload slide Zircon U–Pb Concordia diagrams. (a) Dewulu (DWL) dacite (0991); (b) Saierqingou (SEQG) garnet-bearing pyroclastic rock (0984); (c) Saierqingou andesite (09133); (d) Fangdelongwa (FDLW) garnet-bearing granodioritic porphyry (0981); (e) and (f) Sangke (SK) leucogranite (09117). Shaded and dashed ellipses are not included in the age calculation. Ellipses represent 1-sigma uncertainty for individual analyses. Thirteen analyses from 0981 (Fandelongwa garnet-bearing porphyry) gave 206Pb/238 U ages between 237 and 248 Ma, with a weighted mean of 243 ± 2 Ma (2σ; MSWD = 0·94) (Fig. 8d). Ten spots gave slightly older 206Pb/238 U ages (253–283 Ma); these zircons could be xenocrysts. Five spots gave slightly younger 206Pb/238 U ages (232–234 Ma), probably due to lead loss. Zircons from 09117 (Sangke leucogranite) exhibit three types of internal structures: oscillatory zoning (I); light (II); and dark (III) luminescence (Supplementary Data Fig. S1). The type I–II zircons have low U and Th contents and high Th/U ratios. Fourteen analyses yielded 206Pb/238 U ages between 237 and 248 Ma, with a weighted mean of 242 ± 3 Ma (2σ; MSWD = 2·3) (Fig. 8e and f). The dark zircons (III) have high and variable U (764–1213 ppm) and Th (9781–30658 ppm) abundances and low Th/U ratios (0·03–0·10). These grains gave 206Pb/238 U ages of 192 to 221 Ma (Fig. 8f), which could be interpreted as lead loss due to high U radiation damage. The inherited cores give old ages of 285–306 Ma, 429–433 Ma and ∼743 Ma (Fig. 8e). A laser fusion 40Ar/39Ar age of 234± 3 Ma was obtained for the matrix glass of the Maixiu andesite (Li et al., 2013). This age may be too young, because the Ar-Ar isotopic system is easily disturbed by late thermal events. The Duowa granodiorite, which intruded into the northern part of the Maixiu group (Fig. 1c), has been dated at 240–241 Ma by LA-ICP-MS zircon U–Pb dating (Li et al., 2015). This age may represent a lower limit for the Maixiu volcanic rocks. Thus, we suggest that the crystallization ages of the igneous rocks of this study are ∼239–244 Ma. Geochemistry Alteration effects Whole-rock major element abundances have been recalculated to 100% volatile-free (Table 2). The petrography and variable loss on ignition (LOI) (0·76–6·86 wt %) of our samples indicate that some samples may have undergone varying degrees of alteration. Selected elements were plotted against loss-on-ignition LOI (Supplementary Data Fig. S2) for evaluating the alteration influences. The Saierqingou volcanic rocks, Fandelongwa porphyries and Sangke leucogranites have relatively low LOI (0·76–2·90 wt %) and neither the mobile nor the immobile elements exhibit any correlation with LOI (Supplementary Data Fig. S2), indicating that the influence of alteration is negligible. The LOI for the Maixiu (1·99–6·86 wt %) and Dewulu (3·44–6·54 wt %) volcanic rocks is relatively high and variable. Their HFSE, REE, Mg, Al and Cr do not show any correlation with LOI (Supplementary Data Fig. S2), implying that these elements are essentially unaffected by alteration. The samples with LOI > ∼4·5 wt % from these two volcanic rock suites show an enrichment in Na and depletions in K, Rb, Ba and Sr relative to the samples with LOI < ∼4·5 wt % (Supplementary Data Fig. S2), which we interpret to show that these mobile elements have been gained or removed during alteration. Thus, the variations in mobile elements in samples with LOI > ∼4·5 wt % are not used in the discussion of magmatic processes. Table 2 Major (wt %) and trace element (ppm) data for the magmatic rocks in this study Sample Name 0929 WQ553 WQ554 WQ555 WQ557 0991 0992 0993 0993R 0994 0995 Maixiu Dewulu Lithology andesite dacite andesite Age (Ma) 239 SiO2 57·95 58·41 57·63 58·11 58·80 65·19 61·93 60·20 60·23 62·22 62·29 TiO2 0·81 0·81 0·76 0·66 0·74 0·58 0·68 0·76 0·77 0·71 0·75 Al2O3 19·37 18·61 16·19 14·33 16·84 16·60 17·31 18·06 18·05 17·76 18·53 Fe2O3t 5·78 6·04 7·77 7·99 7·60 4·66 6·00 6·30 6·30 5·65 5·34 MnO 0·09 0·09 0·09 0·16 0·12 0·06 0·09 0·10 0·10 0·10 0·07 MgO 2·39 2·14 5·00 6·00 5·18 2·70 3·66 3·37 3·36 3·38 3·10 CaO 8·74 9·30 8·55 7·35 6·75 4·45 5·52 6·53 6·53 5·95 4·93 Na2O 2·62 2·40 1·97 3·89 2·66 3·54 2·44 2·56 2·54 2·28 3·07 K2O 2·13 2·04 1·92 1·40 1·19 2·11 2·23 1·99 1·99 1·85 1·78 P2O5 0·11 0·15 0·12 0·12 0·13 0·10 0·13 0·13 0·13 0·11 0·14 LOI 4·33 4·25 3·44 6·54 5·07 1·99 2·76 2·86 2·84 6·86 4·09 Total 100·33 100·04 100·08 100·07 100·15 99·6 99·56 99·53 99·53 99·98 99·71 K2O/Na2O 0·81 0·85 0·98 0·36 0·45 0·60 0·92 0·78 0·79 0·81 0·58 A/CNK 0·86 0·81 0·78 0·67 0·94 1·02 1·05 0·99 1·99 1·07 1·16 Mg# 45·0 41·2 56·0 59·8 57·5 53·5 54·7 51·5 51·4 54·3 53·4 Cr 152 195 641 584 278 105 76·3 69·1 224 66·7 Co 21·4 31·0 46·8 46·5 38·1 13·0 35·6 46·2 30·2 31·1 Ni 55·0 79·2 185 162 99·8 30·2 13·3 17·4 29·7 15·2 Rb 97·2 115 97·9 83·1 60·2 87·5 117 60·9 76·9 83·0 Sr 264 347 481 184 363 376 241 301 203 297 Y 23·0 29·7 27·7 25·2 28·8 14·5 22·4 20·4 19·7 15·3 Zr 143 203 174 153 176 171 203 177 171 224 Nb 7·68 10·5 8·75 7·67 9·12 13·0 17·6 18·1 13·6 18·2 Ba 371 486 496 114 524 481 389 370 441 430 La 25·2 35·6 29·0 26·3 31·8 30·9 34·1 28·9 28·5 30·5 Ce 50·8 72·9 61·0 54·9 64·7 60·4 65·5 56·8 56·5 61·7 Pr 5·95 8·26 6·94 6·19 7·38 6·61 7·14 6·34 6·20 6·87 Nd 22·7 32·3 27·2 24·5 29·2 24·3 26·2 23·4 22·7 25·4 Sm 4·75 6·22 5·41 4·84 5·84 4·71 5·06 4·68 4·51 4·79 Eu 1·13 1·46 1·25 1·02 1·30 1·02 1·11 1·14 1·04 1·21 Gd 4·39 6·05 5·45 4·82 5·43 3·91 4·49 4·24 4·16 4·05 Tb 0·69 0·92 0·82 0·75 0·86 0·54 0·68 0·64 0·62 0·57 Dy 4·13 5·46 4·80 4·40 5·14 2·68 3·92 3·80 3·54 2·94 Ho 0·84 1·10 1·01 0·89 1·03 0·50 0·78 0·74 0·69 0·55 Er 2·30 3·17 2·89 2·55 2·94 1·44 2·20 2·07 1·97 1·52 Tm 0·34 0·45 0·43 0·40 0·43 0·20 0·32 0·28 0·30 0·21 Yb 2·24 2·97 2·79 2·49 2·86 1·26 2·05 1·92 1·93 1·33 Lu 0·32 0·44 0·40 0·37 0·43 0·20 0·34 0·30 0·28 0·21 Hf 3·85 5·87 5·07 4·42 5·13 4·77 5·20 4·54 4·61 5·87 Ta 0·55 0·69 0·55 0·49 0·59 0·97 1·29 1·29 0·99 1·29 Pb 27·0 30·7 22·4 14·5 19·1 41·3 22·1 17·7 12·5 20·7 Th 8·92 11·8 9·44 8·21 9·48 10·4 9·54 8·13 8·62 9·84 U 2·12 2·69 2·20 1·92 2·19 2·55 1·90 1·84 1·69 1·93 Eu/Eu* 0·75 0·72 0·69 0·64 0·69 0·70 0·70 0·77 0·72 0·82 LaN/YbN 8·05 8·60 7·46 7·60 7·98 17·6 12·0 10·8 10·6 16·4 Sample Name 0929 WQ553 WQ554 WQ555 WQ557 0991 0992 0993 0993R 0994 0995 Maixiu Dewulu Lithology andesite dacite andesite Age (Ma) 239 SiO2 57·95 58·41 57·63 58·11 58·80 65·19 61·93 60·20 60·23 62·22 62·29 TiO2 0·81 0·81 0·76 0·66 0·74 0·58 0·68 0·76 0·77 0·71 0·75 Al2O3 19·37 18·61 16·19 14·33 16·84 16·60 17·31 18·06 18·05 17·76 18·53 Fe2O3t 5·78 6·04 7·77 7·99 7·60 4·66 6·00 6·30 6·30 5·65 5·34 MnO 0·09 0·09 0·09 0·16 0·12 0·06 0·09 0·10 0·10 0·10 0·07 MgO 2·39 2·14 5·00 6·00 5·18 2·70 3·66 3·37 3·36 3·38 3·10 CaO 8·74 9·30 8·55 7·35 6·75 4·45 5·52 6·53 6·53 5·95 4·93 Na2O 2·62 2·40 1·97 3·89 2·66 3·54 2·44 2·56 2·54 2·28 3·07 K2O 2·13 2·04 1·92 1·40 1·19 2·11 2·23 1·99 1·99 1·85 1·78 P2O5 0·11 0·15 0·12 0·12 0·13 0·10 0·13 0·13 0·13 0·11 0·14 LOI 4·33 4·25 3·44 6·54 5·07 1·99 2·76 2·86 2·84 6·86 4·09 Total 100·33 100·04 100·08 100·07 100·15 99·6 99·56 99·53 99·53 99·98 99·71 K2O/Na2O 0·81 0·85 0·98 0·36 0·45 0·60 0·92 0·78 0·79 0·81 0·58 A/CNK 0·86 0·81 0·78 0·67 0·94 1·02 1·05 0·99 1·99 1·07 1·16 Mg# 45·0 41·2 56·0 59·8 57·5 53·5 54·7 51·5 51·4 54·3 53·4 Cr 152 195 641 584 278 105 76·3 69·1 224 66·7 Co 21·4 31·0 46·8 46·5 38·1 13·0 35·6 46·2 30·2 31·1 Ni 55·0 79·2 185 162 99·8 30·2 13·3 17·4 29·7 15·2 Rb 97·2 115 97·9 83·1 60·2 87·5 117 60·9 76·9 83·0 Sr 264 347 481 184 363 376 241 301 203 297 Y 23·0 29·7 27·7 25·2 28·8 14·5 22·4 20·4 19·7 15·3 Zr 143 203 174 153 176 171 203 177 171 224 Nb 7·68 10·5 8·75 7·67 9·12 13·0 17·6 18·1 13·6 18·2 Ba 371 486 496 114 524 481 389 370 441 430 La 25·2 35·6 29·0 26·3 31·8 30·9 34·1 28·9 28·5 30·5 Ce 50·8 72·9 61·0 54·9 64·7 60·4 65·5 56·8 56·5 61·7 Pr 5·95 8·26 6·94 6·19 7·38 6·61 7·14 6·34 6·20 6·87 Nd 22·7 32·3 27·2 24·5 29·2 24·3 26·2 23·4 22·7 25·4 Sm 4·75 6·22 5·41 4·84 5·84 4·71 5·06 4·68 4·51 4·79 Eu 1·13 1·46 1·25 1·02 1·30 1·02 1·11 1·14 1·04 1·21 Gd 4·39 6·05 5·45 4·82 5·43 3·91 4·49 4·24 4·16 4·05 Tb 0·69 0·92 0·82 0·75 0·86 0·54 0·68 0·64 0·62 0·57 Dy 4·13 5·46 4·80 4·40 5·14 2·68 3·92 3·80 3·54 2·94 Ho 0·84 1·10 1·01 0·89 1·03 0·50 0·78 0·74 0·69 0·55 Er 2·30 3·17 2·89 2·55 2·94 1·44 2·20 2·07 1·97 1·52 Tm 0·34 0·45 0·43 0·40 0·43 0·20 0·32 0·28 0·30 0·21 Yb 2·24 2·97 2·79 2·49 2·86 1·26 2·05 1·92 1·93 1·33 Lu 0·32 0·44 0·40 0·37 0·43 0·20 0·34 0·30 0·28 0·21 Hf 3·85 5·87 5·07 4·42 5·13 4·77 5·20 4·54 4·61 5·87 Ta 0·55 0·69 0·55 0·49 0·59 0·97 1·29 1·29 0·99 1·29 Pb 27·0 30·7 22·4 14·5 19·1 41·3 22·1 17·7 12·5 20·7 Th 8·92 11·8 9·44 8·21 9·48 10·4 9·54 8·13 8·62 9·84 U 2·12 2·69 2·20 1·92 2·19 2·55 1·90 1·84 1·69 1·93 Eu/Eu* 0·75 0·72 0·69 0·64 0·69 0·70 0·70 0·77 0·72 0·82 LaN/YbN 8·05 8·60 7·46 7·60 7·98 17·6 12·0 10·8 10·6 16·4 Table 2 Major (wt %) and trace element (ppm) data for the magmatic rocks in this study Sample Name 0929 WQ553 WQ554 WQ555 WQ557 0991 0992 0993 0993R 0994 0995 Maixiu Dewulu Lithology andesite dacite andesite Age (Ma) 239 SiO2 57·95 58·41 57·63 58·11 58·80 65·19 61·93 60·20 60·23 62·22 62·29 TiO2 0·81 0·81 0·76 0·66 0·74 0·58 0·68 0·76 0·77 0·71 0·75 Al2O3 19·37 18·61 16·19 14·33 16·84 16·60 17·31 18·06 18·05 17·76 18·53 Fe2O3t 5·78 6·04 7·77 7·99 7·60 4·66 6·00 6·30 6·30 5·65 5·34 MnO 0·09 0·09 0·09 0·16 0·12 0·06 0·09 0·10 0·10 0·10 0·07 MgO 2·39 2·14 5·00 6·00 5·18 2·70 3·66 3·37 3·36 3·38 3·10 CaO 8·74 9·30 8·55 7·35 6·75 4·45 5·52 6·53 6·53 5·95 4·93 Na2O 2·62 2·40 1·97 3·89 2·66 3·54 2·44 2·56 2·54 2·28 3·07 K2O 2·13 2·04 1·92 1·40 1·19 2·11 2·23 1·99 1·99 1·85 1·78 P2O5 0·11 0·15 0·12 0·12 0·13 0·10 0·13 0·13 0·13 0·11 0·14 LOI 4·33 4·25 3·44 6·54 5·07 1·99 2·76 2·86 2·84 6·86 4·09 Total 100·33 100·04 100·08 100·07 100·15 99·6 99·56 99·53 99·53 99·98 99·71 K2O/Na2O 0·81 0·85 0·98 0·36 0·45 0·60 0·92 0·78 0·79 0·81 0·58 A/CNK 0·86 0·81 0·78 0·67 0·94 1·02 1·05 0·99 1·99 1·07 1·16 Mg# 45·0 41·2 56·0 59·8 57·5 53·5 54·7 51·5 51·4 54·3 53·4 Cr 152 195 641 584 278 105 76·3 69·1 224 66·7 Co 21·4 31·0 46·8 46·5 38·1 13·0 35·6 46·2 30·2 31·1 Ni 55·0 79·2 185 162 99·8 30·2 13·3 17·4 29·7 15·2 Rb 97·2 115 97·9 83·1 60·2 87·5 117 60·9 76·9 83·0 Sr 264 347 481 184 363 376 241 301 203 297 Y 23·0 29·7 27·7 25·2 28·8 14·5 22·4 20·4 19·7 15·3 Zr 143 203 174 153 176 171 203 177 171 224 Nb 7·68 10·5 8·75 7·67 9·12 13·0 17·6 18·1 13·6 18·2 Ba 371 486 496 114 524 481 389 370 441 430 La 25·2 35·6 29·0 26·3 31·8 30·9 34·1 28·9 28·5 30·5 Ce 50·8 72·9 61·0 54·9 64·7 60·4 65·5 56·8 56·5 61·7 Pr 5·95 8·26 6·94 6·19 7·38 6·61 7·14 6·34 6·20 6·87 Nd 22·7 32·3 27·2 24·5 29·2 24·3 26·2 23·4 22·7 25·4 Sm 4·75 6·22 5·41 4·84 5·84 4·71 5·06 4·68 4·51 4·79 Eu 1·13 1·46 1·25 1·02 1·30 1·02 1·11 1·14 1·04 1·21 Gd 4·39 6·05 5·45 4·82 5·43 3·91 4·49 4·24 4·16 4·05 Tb 0·69 0·92 0·82 0·75 0·86 0·54 0·68 0·64 0·62 0·57 Dy 4·13 5·46 4·80 4·40 5·14 2·68 3·92 3·80 3·54 2·94 Ho 0·84 1·10 1·01 0·89 1·03 0·50 0·78 0·74 0·69 0·55 Er 2·30 3·17 2·89 2·55 2·94 1·44 2·20 2·07 1·97 1·52 Tm 0·34 0·45 0·43 0·40 0·43 0·20 0·32 0·28 0·30 0·21 Yb 2·24 2·97 2·79 2·49 2·86 1·26 2·05 1·92 1·93 1·33 Lu 0·32 0·44 0·40 0·37 0·43 0·20 0·34 0·30 0·28 0·21 Hf 3·85 5·87 5·07 4·42 5·13 4·77 5·20 4·54 4·61 5·87 Ta 0·55 0·69 0·55 0·49 0·59 0·97 1·29 1·29 0·99 1·29 Pb 27·0 30·7 22·4 14·5 19·1 41·3 22·1 17·7 12·5 20·7 Th 8·92 11·8 9·44 8·21 9·48 10·4 9·54 8·13 8·62 9·84 U 2·12 2·69 2·20 1·92 2·19 2·55 1·90 1·84 1·69 1·93 Eu/Eu* 0·75 0·72 0·69 0·64 0·69 0·70 0·70 0·77 0·72 0·82 LaN/YbN 8·05 8·60 7·46 7·60 7·98 17·6 12·0 10·8 10·6 16·4 Sample Name 0929 WQ553 WQ554 WQ555 WQ557 0991 0992 0993 0993R 0994 0995 Maixiu Dewulu Lithology andesite dacite andesite Age (Ma) 239 SiO2 57·95 58·41 57·63 58·11 58·80 65·19 61·93 60·20 60·23 62·22 62·29 TiO2 0·81 0·81 0·76 0·66 0·74 0·58 0·68 0·76 0·77 0·71 0·75 Al2O3 19·37 18·61 16·19 14·33 16·84 16·60 17·31 18·06 18·05 17·76 18·53 Fe2O3t 5·78 6·04 7·77 7·99 7·60 4·66 6·00 6·30 6·30 5·65 5·34 MnO 0·09 0·09 0·09 0·16 0·12 0·06 0·09 0·10 0·10 0·10 0·07 MgO 2·39 2·14 5·00 6·00 5·18 2·70 3·66 3·37 3·36 3·38 3·10 CaO 8·74 9·30 8·55 7·35 6·75 4·45 5·52 6·53 6·53 5·95 4·93 Na2O 2·62 2·40 1·97 3·89 2·66 3·54 2·44 2·56 2·54 2·28 3·07 K2O 2·13 2·04 1·92 1·40 1·19 2·11 2·23 1·99 1·99 1·85 1·78 P2O5 0·11 0·15 0·12 0·12 0·13 0·10 0·13 0·13 0·13 0·11 0·14 LOI 4·33 4·25 3·44 6·54 5·07 1·99 2·76 2·86 2·84 6·86 4·09 Total 100·33 100·04 100·08 100·07 100·15 99·6 99·56 99·53 99·53 99·98 99·71 K2O/Na2O 0·81 0·85 0·98 0·36 0·45 0·60 0·92 0·78 0·79 0·81 0·58 A/CNK 0·86 0·81 0·78 0·67 0·94 1·02 1·05 0·99 1·99 1·07 1·16 Mg# 45·0 41·2 56·0 59·8 57·5 53·5 54·7 51·5 51·4 54·3 53·4 Cr 152 195 641 584 278 105 76·3 69·1 224 66·7 Co 21·4 31·0 46·8 46·5 38·1 13·0 35·6 46·2 30·2 31·1 Ni 55·0 79·2 185 162 99·8 30·2 13·3 17·4 29·7 15·2 Rb 97·2 115 97·9 83·1 60·2 87·5 117 60·9 76·9 83·0 Sr 264 347 481 184 363 376 241 301 203 297 Y 23·0 29·7 27·7 25·2 28·8 14·5 22·4 20·4 19·7 15·3 Zr 143 203 174 153 176 171 203 177 171 224 Nb 7·68 10·5 8·75 7·67 9·12 13·0 17·6 18·1 13·6 18·2 Ba 371 486 496 114 524 481 389 370 441 430 La 25·2 35·6 29·0 26·3 31·8 30·9 34·1 28·9 28·5 30·5 Ce 50·8 72·9 61·0 54·9 64·7 60·4 65·5 56·8 56·5 61·7 Pr 5·95 8·26 6·94 6·19 7·38 6·61 7·14 6·34 6·20 6·87 Nd 22·7 32·3 27·2 24·5 29·2 24·3 26·2 23·4 22·7 25·4 Sm 4·75 6·22 5·41 4·84 5·84 4·71 5·06 4·68 4·51 4·79 Eu 1·13 1·46 1·25 1·02 1·30 1·02 1·11 1·14 1·04 1·21 Gd 4·39 6·05 5·45 4·82 5·43 3·91 4·49 4·24 4·16 4·05 Tb 0·69 0·92 0·82 0·75 0·86 0·54 0·68 0·64 0·62 0·57 Dy 4·13 5·46 4·80 4·40 5·14 2·68 3·92 3·80 3·54 2·94 Ho 0·84 1·10 1·01 0·89 1·03 0·50 0·78 0·74 0·69 0·55 Er 2·30 3·17 2·89 2·55 2·94 1·44 2·20 2·07 1·97 1·52 Tm 0·34 0·45 0·43 0·40 0·43 0·20 0·32 0·28 0·30 0·21 Yb 2·24 2·97 2·79 2·49 2·86 1·26 2·05 1·92 1·93 1·33 Lu 0·32 0·44 0·40 0·37 0·43 0·20 0·34 0·30 0·28 0·21 Hf 3·85 5·87 5·07 4·42 5·13 4·77 5·20 4·54 4·61 5·87 Ta 0·55 0·69 0·55 0·49 0·59 0·97 1·29 1·29 0·99 1·29 Pb 27·0 30·7 22·4 14·5 19·1 41·3 22·1 17·7 12·5 20·7 Th 8·92 11·8 9·44 8·21 9·48 10·4 9·54 8·13 8·62 9·84 U 2·12 2·69 2·20 1·92 2·19 2·55 1·90 1·84 1·69 1·93 Eu/Eu* 0·75 0·72 0·69 0·64 0·69 0·70 0·70 0·77 0·72 0·82 LaN/YbN 8·05 8·60 7·46 7·60 7·98 17·6 12·0 10·8 10·6 16·4 Whole-rock major and trace elements The Maixiu andesites (SiO2 = 57·49–60·43 wt %) are metaluminous and medium- and high-K calc-alkaline (Fig. 9) (Li et al., 2013 and this study). Based on MgO and Al2O3 contents, the Maixiu andesites can be divided into two groups: high-Mg andesites and low-Mg andesites. The high-Mg andesites are characterized by high MgO (5·00–8·49 wt %), Mg# (56–69), Cr (253–797 ppm) and Ni (65–185 ppm) and low Al2O3 (14·3–17·1 wt %) and TiO2 (0·63–0·76 wt %) contents, whereas the low-Mg andesites have relatively low MgO (2·14–4·87 wt %), Mg# (41–61), Cr (133–491 ppm) and Ni (55–105 ppm) and high Al2O3 (18·3–19·4 wt %) and TiO2 (0·75–0·86 wt %) contents (Fig. 10). Fig. 9 View largeDownload slide Plots of (a) SiO2 content v. Zr/TiO2*0·0001 (Winchester & Floyd, 1977), (b) Na2O + K2O v. SiO2 content (Le Bas et al., 1986), (c) K2O v. SiO2 content (Peccerillo & Taylor, 1976) and (d) A/NK [molar Al2O3/(Na2O + K2O)] v. A/CNK [molar Al2O3/(CaO + Na2O + K2O)] (Maniar & Piccoli, 1989) for the igneous rocks described in this study. Grt, Garnet. Fig. 9 View largeDownload slide Plots of (a) SiO2 content v. Zr/TiO2*0·0001 (Winchester & Floyd, 1977), (b) Na2O + K2O v. SiO2 content (Le Bas et al., 1986), (c) K2O v. SiO2 content (Peccerillo & Taylor, 1976) and (d) A/NK [molar Al2O3/(Na2O + K2O)] v. A/CNK [molar Al2O3/(CaO + Na2O + K2O)] (Maniar & Piccoli, 1989) for the igneous rocks described in this study. Grt, Garnet. Fig. 10 View largeDownload slide Variation of selected element abundances and ratios v. SiO2 content for the igneous rocks described in this study: (a) MgO; (b) Al2O3; (c) P2O5; (d) Mg#; (e) K2O/Na2O; (f) A/CNK [molar Al2O3/(CaO + Na2O + K2O)]; (g) Zr; (h) Cr; (i) Ni; (j) Th; (k) La; (l) Yb; (m) Dy/Yb; (n) Sr; (o) Eu/Eu*. Expermental andesitic liquid lines of descent from Alonso-Perez et al. (2009). Experiments on dehydration melting of metasedimentary rocks from Montel & Vielzeuf (1997), Patiño Douce & Harris (1998) and Patiño Douce (1999). North Himalayan leucogranite data from Zhang et al. (2004), Aoya et al. (2005), King et al. (2011), Guo et al. (2012), Gao & Zeng (2014) and Huang et al. (2014). Grt, Garnet. Fig. 10 View largeDownload slide Variation of selected element abundances and ratios v. SiO2 content for the igneous rocks described in this study: (a) MgO; (b) Al2O3; (c) P2O5; (d) Mg#; (e) K2O/Na2O; (f) A/CNK [molar Al2O3/(CaO + Na2O + K2O)]; (g) Zr; (h) Cr; (i) Ni; (j) Th; (k) La; (l) Yb; (m) Dy/Yb; (n) Sr; (o) Eu/Eu*. Expermental andesitic liquid lines of descent from Alonso-Perez et al. (2009). Experiments on dehydration melting of metasedimentary rocks from Montel & Vielzeuf (1997), Patiño Douce & Harris (1998) and Patiño Douce (1999). North Himalayan leucogranite data from Zhang et al. (2004), Aoya et al. (2005), King et al. (2011), Guo et al. (2012), Gao & Zeng (2014) and Huang et al. (2014). Grt, Garnet. The Saierqingou volcanic rocks have variable SiO2 abundances of 58·73–71·49 wt %. With increasing SiO2, their K2O contents and A/CNK [molar ratio of Al2O3/(CaO + Na2O + K2O)] increase, ranging from metaluminous and medium-K calc-alkaline to peraluminous and high-K calc-alkaline (Fig. 9). The Saierqingou andesites are compositionally comparable to the Maixiu andesites with variable Mg# (52–66), Al2O3 (15·95–18·5 wt %) and Cr (65–379 ppm) and Ni (15–76 ppm) contents (Fig. 10). The Saierqingou felsic volcanic rocks have relatively low contents of Al2O3 (14·45–16·52 wt %), MgO (0·38–1·57 wt %), Cr (2–77 ppm) and Ni (1–14 ppm) with K2O/Na2O ratios = 0·83–0·93 (Fig. 10). The Dewulu andesites and dacites are slightly peraluminous and medium-K calc-alkaline, whereas the Fandelongwa garnet-bearing porphyries are peraluminous and high-K calc-alkaline (Fig. 9). In Harker variation diagrams (Fig. 10), the Dewulu volcanic rocks and Fandelongwa garnet-bearing porphyries show coherent chemical trends to the Saierqingou volcanic rocks. With increasing SiO2 content, their TiO2, CaO, MgO (Mg#), Fe2O3t, Cr, and Ni contents decrease, whereas Dy/Yb increases and P2O5 abundance is constant (Fig. 10). Interestingly, Al2O3 and Sr contents and Eu/Eu* [normalized Eu/((Nd + Gd)/2)] firstly increase and then rapidly decrease towards the most felsic end-member; Zr, Th and La (LREE) abundances are relatively constant at SiO2 < ∼68% and then increase rapidly, and Yb (HREE) contents firstly decrease at SiO2 < ∼68% and then increase (Fig. 10). The Sangke leucogranites with the highest SiO2 contents (73·9–74·1%) are strongly peraluminous and high-K calc-alkaline (Fig. 9). They have the lowest MgO, CaO, Cr, Ni, Zr, and Y(HREE) contents of all the rocks analyzed in this study (Fig. 10). In normalized trace element diagrams (Fig. 11), all these rocks show relative enrichment of Rb, Th, U, K, and LREE, and depletion in HFSE. The Saierqingou felsic rocks are the most enriched in Zr, Hf, Th, and REE, while the Sangke leucogranites show the lowest normalized LREE abundances and the most pronounced negative anomalies of P, Zr, Hf and Ti. Chondrite-normalized REE patterns (Fig. 11) of all these rocks are characterized by enrichment of LREE relative to HREE and variable negative Eu anomalies. The Dewulu volcanic rocks display more pronounced fractionated REE patterns ((La/Yb)N = 10·6–17·6) and greater HREE depletion than the Maixiu volcanic rocks ((La/Yb)N = 7·5–8·6). They have moderate to weak negative Eu anomalies (Eu/Eu* = 0·70–0·82) (Fig. 11). With increasing SiO2 contents, the Saierqingou volcanic rocks display progressively more fractionated REE patterns ((La/Yb)N = 10·2–27·2) and more negative Eu anomalies (Eu/Eu* = 0·37–0·75). The Fandelongwa garnet-bearing porphyries have REE patterns similar to those of the Saierqingou garnet-bearing dacites (Fig. 11). The Sangke leucogranites show the highest LaN/YbN ratios (300–486) and pronounced negative Eu anomalies (Eu/Eu* = 0·42–0·52) (Fig. 11). Fig. 11 View largeDownload slide Primitive mantle (PM) normalized trace element patterns (a), (c), (e) and chondrite-normalized rare earth element patterns (b), (d), (f) for the igneous rocks described in this study. Chondrite and PM values are from Sun & McDonough (1989). Data sources: Archean high-Mg sanukitoids from Smithies & Champion (2000); Setouchi sanukitoid from Tatsumi (2006); North Himalayan leucogranites from Zhang et al. (2004), Aoya et al. (2005), King et al. (2011), Guo et al. (2012), Gao & Zeng (2014) and Huang et al. (2014): Type I, muscovite dehydration melting with minor or no garnet in the source; Type II, muscovite dehydration melting with abundant residual garnet; Type III, fluid-fluxed melting of muscovite; Type IV, fluid-fluxed melting of biotite. Fig. 11 View largeDownload slide Primitive mantle (PM) normalized trace element patterns (a), (c), (e) and chondrite-normalized rare earth element patterns (b), (d), (f) for the igneous rocks described in this study. Chondrite and PM values are from Sun & McDonough (1989). Data sources: Archean high-Mg sanukitoids from Smithies & Champion (2000); Setouchi sanukitoid from Tatsumi (2006); North Himalayan leucogranites from Zhang et al. (2004), Aoya et al. (2005), King et al. (2011), Guo et al. (2012), Gao & Zeng (2014) and Huang et al. (2014): Type I, muscovite dehydration melting with minor or no garnet in the source; Type II, muscovite dehydration melting with abundant residual garnet; Type III, fluid-fluxed melting of muscovite; Type IV, fluid-fluxed melting of biotite. Isotope geochemistry For all samples, initial 87Sr/86Sr isotopic ratios (ISr) and εNd(t) values (Table 3) and zircon εHf(t) values (Supplementary Data Table S7) were calculated at the estimated magma crystallization ages. Table 3 Data for whole-rock Sr and Nd isotopes Sample Lithology 87Rb/86Sr 87Sr/86Sr 2σ Isr 147Sm/144Nd 143Nd/144Nd 2σ εNd(t) T2DM/Ga Maixiu volcanic rocks 0929 andesite 1·069 0·713213 5 0·70956 0·126 0·512076 2 −8·8 1·73 Dewulu volcanic rocks 0991 dacite 0·675 0·710641 5 0·7083 0·117 0·512095 5 −8·2 1·67 0992 andesite 1·401 0·712796 4 0·7080 0·117 0·512056 3 −8·9 1·74 0993 andesite 0·587 0·709317 5 0·7073 0·121 0·512154 3 −7·1 1·59 0994 andesite 1·100 0·712091 4 0·7083 0·120 0·512068 3 −8·8 1·72 Saierqingou volcanic rocks 0986 rhyodacite 1·245 0·713313 4 0·7091 0·118 0·512012 3 −9·8 1·81 0987 rhyolite 2·751 0·719548 5 0·7102 0·095 0·51194 2 −10·5 1·87 0988 Grt-bearing 1·245 0·713917 4 0·7097 0·110 0·51198 3 −10·2 1·84 09131 dacite 1·440 0·714521 6 0·7096 0·108 0·512068 8 −8·4 1·69 09129 andesite 1·167 0·711728 6 0·7077 0·122 0·512081 4 −8·6 1·71 09130 andesite 1·102 0·711712 6 0·7080 0·116 0·512068 5 −8·7 1·72 09133 andesite 0·610 0·710347 7 0·7083 0·114 0·512058 1 −8·8 1·72 Fangdelongwa Grt-bearing porphyries 0980 granodiorite 1·196 0·713286 6 0·7092 0·113 0·512041 2 −9·1 1·75 0981 granodiorite 1·343 0·713554 4 0·7090 0·116 0·51202 3 −9·6 1·79 0982 granodiorite 1·270 0·713347 4 0·7090 0·118 0·512017 7 −9·7 1·80 Sangke leucogranites 09117 leucogranite 6·175 0·732985 7 0·7119 0·159 0·5119793 8 −12·2 1·98 09119 leucogranite 9·681 0·744820 7 0·7118 0·181 0·511992 6 −12·1 1·97 09121 leucogranite 7·010 0·734200 7 0·7103 0·168 0·511980 8 −12·0 1·97 09122 leucogranite 7·038 0·735416 5 0·7114 0·175 0·511966 10 −12·4 2·00 Sample Lithology 87Rb/86Sr 87Sr/86Sr 2σ Isr 147Sm/144Nd 143Nd/144Nd 2σ εNd(t) T2DM/Ga Maixiu volcanic rocks 0929 andesite 1·069 0·713213 5 0·70956 0·126 0·512076 2 −8·8 1·73 Dewulu volcanic rocks 0991 dacite 0·675 0·710641 5 0·7083 0·117 0·512095 5 −8·2 1·67 0992 andesite 1·401 0·712796 4 0·7080 0·117 0·512056 3 −8·9 1·74 0993 andesite 0·587 0·709317 5 0·7073 0·121 0·512154 3 −7·1 1·59 0994 andesite 1·100 0·712091 4 0·7083 0·120 0·512068 3 −8·8 1·72 Saierqingou volcanic rocks 0986 rhyodacite 1·245 0·713313 4 0·7091 0·118 0·512012 3 −9·8 1·81 0987 rhyolite 2·751 0·719548 5 0·7102 0·095 0·51194 2 −10·5 1·87 0988 Grt-bearing 1·245 0·713917 4 0·7097 0·110 0·51198 3 −10·2 1·84 09131 dacite 1·440 0·714521 6 0·7096 0·108 0·512068 8 −8·4 1·69 09129 andesite 1·167 0·711728 6 0·7077 0·122 0·512081 4 −8·6 1·71 09130 andesite 1·102 0·711712 6 0·7080 0·116 0·512068 5 −8·7 1·72 09133 andesite 0·610 0·710347 7 0·7083 0·114 0·512058 1 −8·8 1·72 Fangdelongwa Grt-bearing porphyries 0980 granodiorite 1·196 0·713286 6 0·7092 0·113 0·512041 2 −9·1 1·75 0981 granodiorite 1·343 0·713554 4 0·7090 0·116 0·51202 3 −9·6 1·79 0982 granodiorite 1·270 0·713347 4 0·7090 0·118 0·512017 7 −9·7 1·80 Sangke leucogranites 09117 leucogranite 6·175 0·732985 7 0·7119 0·159 0·5119793 8 −12·2 1·98 09119 leucogranite 9·681 0·744820 7 0·7118 0·181 0·511992 6 −12·1 1·97 09121 leucogranite 7·010 0·734200 7 0·7103 0·168 0·511980 8 −12·0 1·97 09122 leucogranite 7·038 0·735416 5 0·7114 0·175 0·511966 10 −12·4 2·00 Notes: 87Rb/86Sr and 147Sm/144Nd ratios are calculated using Rb, Sr, Sm and Nd contents (Table 2), measured by ICP-MS. Table 3 Data for whole-rock Sr and Nd isotopes Sample Lithology 87Rb/86Sr 87Sr/86Sr 2σ Isr 147Sm/144Nd 143Nd/144Nd 2σ εNd(t) T2DM/Ga Maixiu volcanic rocks 0929 andesite 1·069 0·713213 5 0·70956 0·126 0·512076 2 −8·8 1·73 Dewulu volcanic rocks 0991 dacite 0·675 0·710641 5 0·7083 0·117 0·512095 5 −8·2 1·67 0992 andesite 1·401 0·712796 4 0·7080 0·117 0·512056 3 −8·9 1·74 0993 andesite 0·587 0·709317 5 0·7073 0·121 0·512154 3 −7·1 1·59 0994 andesite 1·100 0·712091 4 0·7083 0·120 0·512068 3 −8·8 1·72 Saierqingou volcanic rocks 0986 rhyodacite 1·245 0·713313 4 0·7091 0·118 0·512012 3 −9·8 1·81 0987 rhyolite 2·751 0·719548 5 0·7102 0·095 0·51194 2 −10·5 1·87 0988 Grt-bearing 1·245 0·713917 4 0·7097 0·110 0·51198 3 −10·2 1·84 09131 dacite 1·440 0·714521 6 0·7096 0·108 0·512068 8 −8·4 1·69 09129 andesite 1·167 0·711728 6 0·7077 0·122 0·512081 4 −8·6 1·71 09130 andesite 1·102 0·711712 6 0·7080 0·116 0·512068 5 −8·7 1·72 09133 andesite 0·610 0·710347 7 0·7083 0·114 0·512058 1 −8·8 1·72 Fangdelongwa Grt-bearing porphyries 0980 granodiorite 1·196 0·713286 6 0·7092 0·113 0·512041 2 −9·1 1·75 0981 granodiorite 1·343 0·713554 4 0·7090 0·116 0·51202 3 −9·6 1·79 0982 granodiorite 1·270 0·713347 4 0·7090 0·118 0·512017 7 −9·7 1·80 Sangke leucogranites 09117 leucogranite 6·175 0·732985 7 0·7119 0·159 0·5119793 8 −12·2 1·98 09119 leucogranite 9·681 0·744820 7 0·7118 0·181 0·511992 6 −12·1 1·97 09121 leucogranite 7·010 0·734200 7 0·7103 0·168 0·511980 8 −12·0 1·97 09122 leucogranite 7·038 0·735416 5 0·7114 0·175 0·511966 10 −12·4 2·00 Sample Lithology 87Rb/86Sr 87Sr/86Sr 2σ Isr 147Sm/144Nd 143Nd/144Nd 2σ εNd(t) T2DM/Ga Maixiu volcanic rocks 0929 andesite 1·069 0·713213 5 0·70956 0·126 0·512076 2 −8·8 1·73 Dewulu volcanic rocks 0991 dacite 0·675 0·710641 5 0·7083 0·117 0·512095 5 −8·2 1·67 0992 andesite 1·401 0·712796 4 0·7080 0·117 0·512056 3 −8·9 1·74 0993 andesite 0·587 0·709317 5 0·7073 0·121 0·512154 3 −7·1 1·59 0994 andesite 1·100 0·712091 4 0·7083 0·120 0·512068 3 −8·8 1·72 Saierqingou volcanic rocks 0986 rhyodacite 1·245 0·713313 4 0·7091 0·118 0·512012 3 −9·8 1·81 0987 rhyolite 2·751 0·719548 5 0·7102 0·095 0·51194 2 −10·5 1·87 0988 Grt-bearing 1·245 0·713917 4 0·7097 0·110 0·51198 3 −10·2 1·84 09131 dacite 1·440 0·714521 6 0·7096 0·108 0·512068 8 −8·4 1·69 09129 andesite 1·167 0·711728 6 0·7077 0·122 0·512081 4 −8·6 1·71 09130 andesite 1·102 0·711712 6 0·7080 0·116 0·512068 5 −8·7 1·72 09133 andesite 0·610 0·710347 7 0·7083 0·114 0·512058 1 −8·8 1·72 Fangdelongwa Grt-bearing porphyries 0980 granodiorite 1·196 0·713286 6 0·7092 0·113 0·512041 2 −9·1 1·75 0981 granodiorite 1·343 0·713554 4 0·7090 0·116 0·51202 3 −9·6 1·79 0982 granodiorite 1·270 0·713347 4 0·7090 0·118 0·512017 7 −9·7 1·80 Sangke leucogranites 09117 leucogranite 6·175 0·732985 7 0·7119 0·159 0·5119793 8 −12·2 1·98 09119 leucogranite 9·681 0·744820 7 0·7118 0·181 0·511992 6 −12·1 1·97 09121 leucogranite 7·010 0·734200 7 0·7103 0·168 0·511980 8 −12·0 1·97 09122 leucogranite 7·038 0·735416 5 0·7114 0·175 0·511966 10 −12·4 2·00 Notes: 87Rb/86Sr and 147Sm/144Nd ratios are calculated using Rb, Sr, Sm and Nd contents (Table 2), measured by ICP-MS. The Maixiu andesites show large variation in ISr (0·7079 to 0·7123), but a limited range of εNd(t) (-9·3 to -8·0) (Fig. 12a) (Li et al., 2013 and this study). Rubidium and Sr contents and Rb/Sr ratios in the Maixiu andesites may have been disturbed by alteration (Supplementary Data Fig. S2), and, therefore, their Sr isotope compositions may also have been modified and have no petrogenetic significance. However, the Maixiu andesites have εNd(t) values similar to those of the Dewulu and Saierqingou andesites, indicating that the Nd isotopes were not affected by alteration. The Saierqingou volcanic rocks have ISr = 0·7077–0·7102 and εNd(t) = -10·5 to -8·4 (Fig. 12a). The Dewulu volcanic rocks have ISr = 0·7073–0·7083 and εNd(t) = -8·9 to -7·1 (Fig. 12a). The Fandelongwa garnet-bearing porphyrites have ISr = 0·7090–0·7092 and εNd(t) = -9·7 to -9·1 (Fig. 12a). The Sangke leucogranites have the highest ISr of 0·7103–0·7119 and the lowest εNd(t) of -12·4 to -12·0 (Fig. 12a). Fig. 12 View largeDownload slide (a) Plot of εNd(t) v. ISr and (b) zircon εHf(t) v. the crystallization age for the igneous rocks described in this study. Data sources: Middle Triassic I-type granites from Zhang et al. (2007) and Luo et al. (2012, 2015); mafic rocks from Luo et al. (2012) and Li et al., (2015); Mian-Lue and A’nimaque N-MORB ophiolites from Xu et al. (2002) and Guo et al. (2007). Data for Devonian and Carboniferous sedimentary rocks from Liu et al. (2008b); Xiekeng plutons from Guo et al. (2012); Meiwu I-type granitoid batholith from Luo et al. (2015). Grt, Garnet. Fig. 12 View largeDownload slide (a) Plot of εNd(t) v. ISr and (b) zircon εHf(t) v. the crystallization age for the igneous rocks described in this study. Data sources: Middle Triassic I-type granites from Zhang et al. (2007) and Luo et al. (2012, 2015); mafic rocks from Luo et al. (2012) and Li et al., (2015); Mian-Lue and A’nimaque N-MORB ophiolites from Xu et al. (2002) and Guo et al. (2007). Data for Devonian and Carboniferous sedimentary rocks from Liu et al. (2008b); Xiekeng plutons from Guo et al. (2012); Meiwu I-type granitoid batholith from Luo et al. (2015). Grt, Garnet. The Saierqingou andesite (09133) and garnet-bearing pyroclastic rock (0984) exhibit zircon εHf(t) of -7·5 to -5·8 (mean = -6·8 ± 0·3) and -7·6 to -5·0 (mean = -6·3 ± 0·4) (Fig. 12b), respectively. The Dewulu garnet-bearing dacite (0991) has zircon εHf(t) = -9·3 to -5·7 (mean = -7·7 ± 0·6) (Fig. 12b). The Fandelongwa garnet-bearing porphyry (0981) has zircon εHf(t) of -13·3 to -9·5 (mean = -11·0 ± 0·6) (Fig. 12b). The Sangke leucogranite (09117) has the lowest zircon εHf(t) of -14·0 to -9·1 (mean = -11·3 ± 0·8) (Fig. 12b). DISCUSSION The diversity of igneous rock types described in this study belong to a coeval and spatially associated suite. Wang & Ding (1990) have reported garnet-bearing andesites from the Dewulu and Maixiu volcanic suites and, although we did not find garnet in the andesitic samples used in this study, their Yb contents and Dy/Yb ratios (Fig. 10l and m) indicate cryptic garnet fractionation at depth. There may be a petrogenetic link between the MSD (Maixiu, Saierqingou and Dewulu) andesites and the peraluminous garnet-bearing Saierqingou dacites and Fandelongwa porphyries. The leucogranites show extreme depletion of HREE (Fig. 11f), consistent with abundant residual garnet in their sources. The various types of garnets in this study may be recording the interaction between the diversity of magmas represented by the igneous rocks. Our objective is to use the petrographic, geochemical and mineral chemical information to establish the petrogenetic links between the igneous rocks and thereby to reconstruct the whole magmatic system in which they were generated and evolved. Furthermore, variation in the composition of crystal populations in the igneous rocks affects and controls the bulk-rock composition and thus we seek to examine the relationship of the different types of crystal populations to their host-rocks. Petrogenesis of the high-Mg andesites The Maixiu, Saierqingou and Dewulu andesites, hereinafter referred to as the MSD high-Mg andesites, have high MgO (Mg#), Cr and Ni contents. They share some geochemical affinities with Cenozoic high-Mg andesites (HMAs) (Tatsumi, 2006) and Archean high-Mg sanukitoids (Smithies & Champion, 2000) (Fig. 11a and b). Several mechanisms have been proposed to account for the petrogenesis of the HMAs and sanukitoids, including: (1) partial melting of hydrated mantle peridotite (Hirose, 1997); (2) partial melting of metasomatized mantle modified by slab-derived melts (Shirey & Hanson, 1984; Smithies & Champion, 2000); (3) assimilation of mantle peridotite by melts from subducting oceanic sediments (Tatsumi, 2006); and (4) interaction between mantle peridotite and melts derived from delaminated lower crust (Gao et al., 2004). All of these models emphasize that the HMAs and sanukitoids can represent near-primary melts and acquire their primitive features at mantle depths. The Maixiu andesites were formerly proposed to be generated by the interaction between subducted sediment-derived melts and peridotites at mantle depths (Li et al. 2013). However, these bulk-rock geochemical affinities may not be intrinsic features of primary melt (Kamei et al., 2004; Zhang et al., 2013; Chiaradia et al., 2014). The petrography, mineral chemistry and zoning, and whole-rock geochemistry are more consistent with a model whereby the MSD high-Mg andesites are produced by complex open-system processes at crustal levels. Most clinopyroxenes from the Maixiu andesites are unzoned with low Mg# (73–84), although some grains display reverse zoning with low Mg# (∼77–81) interiors surrounded by a narrow rim having high Mg# (87–91) (Fig. 5a) (Li et al., 2013). The orthopyroxenes from the Maixiu andesites commonly have uniform high Mg# (86–92) mantle and core compositions, with slightly increasing Mg# from core to mantle, and thin, low Mg# (82–85) rims (Fig. 5b) (Li et al., 2013). The reverse trend of Mg# in clinopyroxene and orthopyroxene is interpreted to indicate recharge of an evolved crustal magma reservoir by basaltic magma pulses. The plagioclase phenocrysts preserve sieve-textured cores with pristine outer rims (Fig. 2b) and they display various shapes and complex zoning (Fig. 2c); all these features are best interpreted to reflect the influence of magma mixing and mingling processes. Using published KdFe–Mg values for clinopyroxene (0·28 ± 0·07) and orthopyroxene (0·29 ± 0·06) (Putirka, 2008b), we conclude that high-Mg# clinopyroxene and orthopyroxene crystallized from more primitive basaltic magma (Mg# = 62–75), whereas low-Mg# clinopyroxene and orthopyroxene were in equilibrium with more evolved melts (Mg# = 43–61). The plagioclase population has two compositional peaks in An content (Fig. 5c) (Li et al., 2013). High-An (78–83) plagioclase phenocrysts may have crystallized from relatively mafic magmas, whereas some low-An (68–78) plagioclase phenocrysts and matrix plagioclases (An = 60–78) (Fig. 5c) probably crystallized from more evolved melts. Plagioclase phenocrysts with lower An values in garnet-bearing andesites (Wang & Ding, 1990) may also have crystallized from more evolved magmas. Thus, the reversed zoned and high-Mg# clinopyroxene and orthopyroxene, high-An plagioclase and the presence of these in glomerocrysts are best interpreted as antecrysts or dispersed cumulates that crystallized from more primitive basaltic magmas. The low-Mg# orthopyroxene thin rims and clinopyroxene, and low-An matrix plagioclase could be orthocrysts that were in equilibrium with their host magmas. The MSD high-Mg andesites have higher MgO (Mg#), Cr and Ni contents than experimental melts produced by melting of mafic crust (Rapp & Watson, 1995; Johannes & Holtz, 1996; Sisson et al., 2005) (Fig. 13a), ruling out the possibility that the magmas that they represent were solely partial melts of mafic crust. In addition, the MSD high-Mg andesites define geochemical variations that deviate from the experimental liquid lines of descent (Sisson & Grove, 1993; Müntener et al., 2001; Grove et al., 2003; Alonso-Perez et al., 2009) (curves FC1–5; Fig. 13), providing further evidence that they were not formed exclusively by fractional crystallization of basaltic magmas with or without crustal contamination. Fig. 13 View largeDownload slide Bulk-rock compositions plotted in: (a) Mg# v. SiO2 and (b) Cr v. SiO2 diagrams. Grt, Garnet. Data sources: garnet-bearing volcanic rocks in the Northland Arc, New Zealand from Bach et al. (2012); metabasaltic and eclogite experimental melts (0·7–3 GPa) from Johannes & Holtz (1996), Rapp & Watson (1995) and Sisson et al. (2005). Black lines (FC1–6) represent the experimental liquid lines of decent for hydrous basalts (1, 2 and 4) and andesites (3, 5 and 6), 1 = 0·1–0·2 GPa (Grove et al., 2003); 2–3 = 0·1–0·2 GPa (Sisson & Grove, 1993); 4–5 = 1·2 GPa (Müntener et al., 2001); 6 = 0·8–1·2 GPa (Alonso-Perez et al., 2009). The Cr contents for FC2–4 in (b) are calculated by applying a fractional crystallization model using the phase proportions obtained from the experimental results. The partition coefficients are from Rollinson (1993). The dashed lines with white circles are mixing trajectories. The mafic end-member used in the models is a primitive Permian basalt from the West Qinling Orogen (Kou et al., 2007) and the felsic end-member is from this study. Pyroxene compositions used in the models are from Li et al. (2013). Fig. 13 View largeDownload slide Bulk-rock compositions plotted in: (a) Mg# v. SiO2 and (b) Cr v. SiO2 diagrams. Grt, Garnet. Data sources: garnet-bearing volcanic rocks in the Northland Arc, New Zealand from Bach et al. (2012); metabasaltic and eclogite experimental melts (0·7–3 GPa) from Johannes & Holtz (1996), Rapp & Watson (1995) and Sisson et al. (2005). Black lines (FC1–6) represent the experimental liquid lines of decent for hydrous basalts (1, 2 and 4) and andesites (3, 5 and 6), 1 = 0·1–0·2 GPa (Grove et al., 2003); 2–3 = 0·1–0·2 GPa (Sisson & Grove, 1993); 4–5 = 1·2 GPa (Müntener et al., 2001); 6 = 0·8–1·2 GPa (Alonso-Perez et al., 2009). The Cr contents for FC2–4 in (b) are calculated by applying a fractional crystallization model using the phase proportions obtained from the experimental results. The partition coefficients are from Rollinson (1993). The dashed lines with white circles are mixing trajectories. The mafic end-member used in the models is a primitive Permian basalt from the West Qinling Orogen (Kou et al., 2007) and the felsic end-member is from this study. Pyroxene compositions used in the models are from Li et al. (2013). The MSD high-Mg andesites have radiogenic ISr and negative εNd(t) and plot between the fields of the coeval mafic-ultramafic rocks (Luo et al., 2012; Li et al., 2014) and the Sangke leucogranites or Paleozoic sediments (Liu et al., 2008a; Luo et al., 2015) from the West Qinling Orogen (Fig. 12a). The Saierqingou andesite (09133) also has zircon εHf(t) values between the mafic rocks and the Sangke leucogranites (Fig. 12b). These isotopic features are consistent with mixing of contrasting basaltic and felsic magmas. To better evaluate the role of magma mixing, a simple two component mixing model (DePaolo & Wasserburg, 1979) has been applied based on the Sr and Nd isotope data and Sr contents (Fig. 14). The mafic end-member could be represented by the coeval mafic-ultramafic rocks, which have been suggested to be derived from an enriched lithospheric mantle source (Luo et al., 2012, Li et al., 2014). The potential felsic end-member could be the Sangke leucogranite or Paleozoic sediment (Liu et al., 2008a). However, the low-grade metamorphosed Paleozoic sediments do not contain any evidence for anatectic melting. Thus, in the mixing models, the felsic end-member is assumed to be represented by the Sangke leucogranites (samples 09117 and 09122). The detailed parameters used in the modeling are given in Table 4. The results of mixing modeling (curves B-MIX-1 and B-MIX-2 in Fig. 14) are interpreted to indicate that the MSD high-Mg andesites could have been formed by magma mixing involving mantle-derived basaltic magmas and crustal felsic melts (Fig. 14). Table 4 Parameters used in the mixing and AFC models for the magmatic rocks in this study End member Sr (ppm) Nd (ppm) 87Sr/86Sr 143Nd/144Nd End member Sr (ppm) Nd (ppm) 87Sr/86Sr 143Nd/144Nd Mixing model: B-MIX-1 Mixing model: B-MIX-2 Basaltic magma (B1) 250 10 0·7060 0·5123 Basaltic magma (B2) 250 10 0·7074 0·5123 crustal melt (L1) 106 20 0·7103 0·5117 crustal melt (L2) 120 20 0·7119 0·5117 AFC model: A-AFC-Lower crust, r = 0·5 AFC model: A-AFC-Upper crust, r = 0·5 Cpx = 0·35, Opx = 0·40, Pl = 0·05, Amp = 0·05, Bio = 0·10, Grt = 0·05 Pl = 0·40, Amp= 0·05, Bio = 0·05, Q = 0·2, Kfs = 0·3 Andesitic magma 190 25 0·7077 0·5119 Dacitic magma 385 31 0·7086 0·5118 Lower crust (L1) 106 20 0·7114 0·5117 Upper crust (U) 150 40 0·7160 0·5117 Partition coefficients (D) used in AFC models Host magmas Opx Cpx Grt Pl Amp Bio Qz Kfs DSr Andesitic magma 0·04 0·06 0·015 2·82 0·30 0·012 DNd 0·028 0·18 0·530 0·149 1·2 0·290 DSr Dacitic magma 4·4 0·30 0·012 0·020 3·76 DNd 0·149 4·26 2·230 0·014 0·024 End member Sr (ppm) Nd (ppm) 87Sr/86Sr 143Nd/144Nd End member Sr (ppm) Nd (ppm) 87Sr/86Sr 143Nd/144Nd Mixing model: B-MIX-1 Mixing model: B-MIX-2 Basaltic magma (B1) 250 10 0·7060 0·5123 Basaltic magma (B2) 250 10 0·7074 0·5123 crustal melt (L1) 106 20 0·7103 0·5117 crustal melt (L2) 120 20 0·7119 0·5117 AFC model: A-AFC-Lower crust, r = 0·5 AFC model: A-AFC-Upper crust, r = 0·5 Cpx = 0·35, Opx = 0·40, Pl = 0·05, Amp = 0·05, Bio = 0·10, Grt = 0·05 Pl = 0·40, Amp= 0·05, Bio = 0·05, Q = 0·2, Kfs = 0·3 Andesitic magma 190 25 0·7077 0·5119 Dacitic magma 385 31 0·7086 0·5118 Lower crust (L1) 106 20 0·7114 0·5117 Upper crust (U) 150 40 0·7160 0·5117 Partition coefficients (D) used in AFC models Host magmas Opx Cpx Grt Pl Amp Bio Qz Kfs DSr Andesitic magma 0·04 0·06 0·015 2·82 0·30 0·012 DNd 0·028 0·18 0·530 0·149 1·2 0·290 DSr Dacitic magma 4·4 0·30 0·012 0·020 3·76 DNd 0·149 4·26 2·230 0·014 0·024 Mixing model is after DePaolo & Wasserburg (1979). AFC, assimilation and fractional crystallization (Depaolo, 1981); r, ratio of the rate of assimilation to the rate of fractional crytsallisation. Cpx, Clinopyroxene; Opx, Orthopyroxene; Grt, Garnet; Pl, Plagioclase; Amp, Amphibole; Bio, Biotite; Qz, Quartz; Kfs, K-feldspar. Balsaltic magmas are from Luo et al. (2012) and Li et al. (2015). Crustal melt and Lower crust assimilation is represented by the leucogranites in this study. Upper crust assimilation is from Liu et al. (2008a). Mineral assemblage in AFC models has referred to the experimental results by Alonso-Perez et al. (2009) and petrography observation in this study. Partition coefficients from Rollinson, (1993) and GERM Partition Coefficient (Kd) Database (http://earthref.org/KDD/). Table 4 Parameters used in the mixing and AFC models for the magmatic rocks in this study End member Sr (ppm) Nd (ppm) 87Sr/86Sr 143Nd/144Nd End member Sr (ppm) Nd (ppm) 87Sr/86Sr 143Nd/144Nd Mixing model: B-MIX-1 Mixing model: B-MIX-2 Basaltic magma (B1) 250 10 0·7060 0·5123 Basaltic magma (B2) 250 10 0·7074 0·5123 crustal melt (L1) 106 20 0·7103 0·5117 crustal melt (L2) 120 20 0·7119 0·5117 AFC model: A-AFC-Lower crust, r = 0·5 AFC model: A-AFC-Upper crust, r = 0·5 Cpx = 0·35, Opx = 0·40, Pl = 0·05, Amp = 0·05, Bio = 0·10, Grt = 0·05 Pl = 0·40, Amp= 0·05, Bio = 0·05, Q = 0·2, Kfs = 0·3 Andesitic magma 190 25 0·7077 0·5119 Dacitic magma 385 31 0·7086 0·5118 Lower crust (L1) 106 20 0·7114 0·5117 Upper crust (U) 150 40 0·7160 0·5117 Partition coefficients (D) used in AFC models Host magmas Opx Cpx Grt Pl Amp Bio Qz Kfs DSr Andesitic magma 0·04 0·06 0·015 2·82 0·30 0·012 DNd 0·028 0·18 0·530 0·149 1·2 0·290 DSr Dacitic magma 4·4 0·30 0·012 0·020 3·76 DNd 0·149 4·26 2·230 0·014 0·024 End member Sr (ppm) Nd (ppm) 87Sr/86Sr 143Nd/144Nd End member Sr (ppm) Nd (ppm) 87Sr/86Sr 143Nd/144Nd Mixing model: B-MIX-1 Mixing model: B-MIX-2 Basaltic magma (B1) 250 10 0·7060 0·5123 Basaltic magma (B2) 250 10 0·7074 0·5123 crustal melt (L1) 106 20 0·7103 0·5117 crustal melt (L2) 120 20 0·7119 0·5117 AFC model: A-AFC-Lower crust, r = 0·5 AFC model: A-AFC-Upper crust, r = 0·5 Cpx = 0·35, Opx = 0·40, Pl = 0·05, Amp = 0·05, Bio = 0·10, Grt = 0·05 Pl = 0·40, Amp= 0·05, Bio = 0·05, Q = 0·2, Kfs = 0·3 Andesitic magma 190 25 0·7077 0·5119 Dacitic magma 385 31 0·7086 0·5118 Lower crust (L1) 106 20 0·7114 0·5117 Upper crust (U) 150 40 0·7160 0·5117 Partition coefficients (D) used in AFC models Host magmas Opx Cpx Grt Pl Amp Bio Qz Kfs DSr Andesitic magma 0·04 0·06 0·015 2·82 0·30 0·012 DNd 0·028 0·18 0·530 0·149 1·2 0·290 DSr Dacitic magma 4·4 0·30 0·012 0·020 3·76 DNd 0·149 4·26 2·230 0·014 0·024 Mixing model is after DePaolo & Wasserburg (1979). AFC, assimilation and fractional crystallization (Depaolo, 1981); r, ratio of the rate of assimilation to the rate of fractional crytsallisation. Cpx, Clinopyroxene; Opx, Orthopyroxene; Grt, Garnet; Pl, Plagioclase; Amp, Amphibole; Bio, Biotite; Qz, Quartz; Kfs, K-feldspar. Balsaltic magmas are from Luo et al. (2012) and Li et al. (2015). Crustal melt and Lower crust assimilation is represented by the leucogranites in this study. Upper crust assimilation is from Liu et al. (2008a). Mineral assemblage in AFC models has referred to the experimental results by Alonso-Perez et al. (2009) and petrography observation in this study. Partition coefficients from Rollinson, (1993) and GERM Partition Coefficient (Kd) Database (http://earthref.org/KDD/). Fig. 14 View largeDownload slide Variation of (a) εNd(t) v. ISr and(b) ISrv. Sr highlighting the effect of magma mixing and assimilation and fractional crystallization (AFC) processes. B-Mix-1 and -2 represent mixing between mantle-derived basaltic magmas and crustal melts. For mixing lines, each circle represents 10% addition of the mafic endmember. A-AFC-Lower crust and -Upper crust are models in which andesitic magmas are assumed to have continuously evolved in a two-stage AFC process taking place in the lower crust and upper crust, respectively. F is the remaining melt fraction and each circle represents 10% fractionation. The grey dotted line with arrow represents the actual trend defined by the samples. Fig. 14 View largeDownload slide Variation of (a) εNd(t) v. ISr and(b) ISrv. Sr highlighting the effect of magma mixing and assimilation and fractional crystallization (AFC) processes. B-Mix-1 and -2 represent mixing between mantle-derived basaltic magmas and crustal melts. For mixing lines, each circle represents 10% addition of the mafic endmember. A-AFC-Lower crust and -Upper crust are models in which andesitic magmas are assumed to have continuously evolved in a two-stage AFC process taking place in the lower crust and upper crust, respectively. F is the remaining melt fraction and each circle represents 10% fractionation. The grey dotted line with arrow represents the actual trend defined by the samples. A magma mixing scenario is also compatible with the whole-rock element compositions (Fig. 13). A Permian basalt (sample 114–1) from the West Qinling Orogen has been used in the mixing models as the potential mafic end-member (Kou et al., 2007). Sample 0987, from this study, was taken as the felsic end-member. The modeling results reveal that mixing of equal proportions of the basaltic and felsic end-members can account for the high Mg# and Cr contents of most andesitic rocks (Fig. 13). If this hybrid andesitic magma entrained another 10% orthopyroxene and 5% clinoproxene, the resultant mixed magma would have a higher Mg# and Cr content, as observed in some MSD andesites (Fig. 13). In addition, some MSD andesites also display relatively low MgO (Mg#), Cr and Ni contents, but high Al2O3 and Sr abundances (Fig. 10), which could be interpreted as a result of entrainment of plagioclase in evolved dacitic magmas. This interpretation is consistent with the petrographic observation that the high-Mg andesites contain more pyroxenes, whereas the low-Mg andesites have more plagioclase (Fig. 2). Therefore, we suggest that the MSD high-Mg andesites were actually mixtures of andesitic and dacitic magmas with each carrying different crystal cargoes. Petrogenesis of the peraluminous garnet-bearing porphyries and crystal-poor rhyolites Peraluminous magmas are usually assumed to be produced by partial melting of metasedimentary rocks (Chappell & White, 1974) but it has also been suggested that they can be produced by: (1) fractional crystallization of amphibole and clinopyroxene from metaluminous magmas (Cawthorn & Brown, 1972; Zen, 1988; Alonso-Perez et al., 2009); (2) fractional crystallization of metaluminous magmas accompanied by crustal assimilation (AFC) (Barley, 1987; Barbarin, 1996; Harangi et al., 2001); and (3) mixing of mafic and crustal felsic magmas (Keay et al., 1997). The Saierqingou and Dewulu volcanic rocks and Fandelongwa porphyries collectively define a set of continuous geochemical evolution trends, indicating that they are likely to have shared a similar petrogenesis. The various compositional and zoning patterns of garnets can be used as indicators of magmatic evolution and prevailing conditions. Origin of garnets in the peraluminous magmas Compositions of garnet in igneous rocks may reflect melt composition, pressure, temperature, and oxygen fugacity (fO2) in the course of magma evolution. They can also provide insights into their ultimate origin, e.g. metamorphic, restitic, peritectic or magmatic (Day et al., 1992; Harangi et al., 2001; Stevens et al., 2007; Acosta-Vigil et al., 2010; Bach et al., 2012). Harangi et al. (2001) have demonstrated that garnets crystallized from mantle-derived (M-type) or I-type magmas at high-pressure are enriched in CaO (>5 wt %) and depleted in MnO (<2 wt %), whereas those formed at relatively low-pressure have high CaO (>4 wt %) and high MnO (>3 wt %) contents. Garnets formed in peraluminous magmas and, or, metapelites have low CaO (<4 wt %) and variable MnO contents (Fig. 15). In this study, garnet textures, compositions and inclusions are jointly used to distinguish their origins. Fig. 15 View largeDownload slide CaO v. MnO contents for garnet described in this study. Fields of garnet from different types of rocks are from Harangi et al. (2001). Compositions for garnet formed in experimental andesitic magmas at 8 kbar and 12 kbar are from Alonso-Perez et al. (2009). Fig. 15 View largeDownload slide CaO v. MnO contents for garnet described in this study. Fields of garnet from different types of rocks are from Harangi et al. (2001). Compositions for garnet formed in experimental andesitic magmas at 8 kbar and 12 kbar are from Alonso-Perez et al. (2009). Type-1 garnets from the peraluminous porphyries and dacites enclose minor zircon, apatite and glass inclusions (Figs 3a–c and 4a–b) and they are, therefore, considered to be of magmatic origin. Type-1 garnets or their cores have high CaO and low MnO contents, which is inconsistent with crystallization from a peraluminous magma (Fig. 15). They plot in the field of high-pressure garnets that may crystallize from M/I-type magmas and also have compositions similar to the garnets from the Maixiu and Dewulu andesites (Wang & Ding, 1990) (Fig. 15). These features are interpreted to indicate that they are antecrysts that were crystallized from andesitic magmas at relatively high pressure. Garnet 0988#1 with a plagioclase corona (Fig. 3a) and 0981#2 (Fig. 4b) with nearly consistent CaO and increasing MnO from core to rim (Fig. 7c) indicate rapid magma ascent and decompression processes (Harangi et al., 2001). In garnet 0988#2, the core and rim contain different inclusions (Fig. 3b) and have different chemical compositions (Fig. 7b), indicating that the melt composition changed during growth. The rim with high MnO and low CaO contents may be due to interaction with a peraluminous magma (Fig. 15). Type-2 garnets, mainly found in the Saierqingou dacites, have clear section surfaces and minor zircon and apatite inclusions (Fig. 3e–h), indicative of a magmatic origin. Some resorbed garnets are enclosed by plagioclase (Fig. 3f–h), indicating that they crystallized earlier than plagioclase. Garnet 09131#0 with a cordierite corona (Fig. 3d) implies a decompression event that probably corresponds to magma ascent. Type-2 garnets with CaO and MnO contents that plot in the field of garnets crystallized from S-type magmas (Fig. 15), are likely orthocrysts crystallized from peraluminous magmas. Type-3 garnets occur as clusters in metapelitic xenoliths and are, therefore, interpreted as xenocrysts (Fig. 4g). Their cores have compositions that overlap with those of garnets formed from S-type magmas or metapelites (Fig. 15). Type-3 garnets contain trichoid sillimanite inclusions and they are concentrated in the studied biotite-rich metapelitic enclave, which distinguishes them from Type-2 garnets (Figs 3 and 4). All these features indicate that they are residual phases formed by peritectic reaction (e.g. Harangi et al., 2001; Acosta-Vigil et al., 2010; Lackey et al., 2012). The garnet-rich metapelitic xenolith is surrounded by a corona of plagioclase glomerocrysts (Figs 2g and 4g), which was probably produced by interaction with the host magma. Some Type-3 garnets have high CaO and MnO rims, similar to the rim of 0981#2 (Fig. 15), which is interpreted to represent overgrowth by reaction with andesitic magma at a relatively low pressure. Type-4 garnets include xenocrysts introduced from the wall-rocks. Garnets 09131#5 (Fig. 3i), 09163#2 (Fig. 4d) and the core of 09163#1 (Fig. 4c) have low CaO and MnO contents, similar to the garnets that form in S-type magmas or metapelites (Fig. 15). Compared with the Type-3 garnets, they contain distinct mineral inclusions and have different compositional zoning patterns (Figs 3 and 4), indicating that they have different origins and evolutionary histories. Garnet 09131#5 with abundant fluid inclusions (Fig. 3i) is likely to have formed under fluid-present conditions. The rim of garnet 09163#1 (Fig. 4c) has a composition similar to those of Type-1 garnets, indicative of interaction with andesitic magma at high pressure (Fig. 15). The primary composition of 09164#4 (Fig. 4e and f) has been completely replaced during interaction with andesitic magma at relatively low pressure (Fig. 15). Assimilation and fractional crystallization of andesitic magmas The contents of MgO (Mg#), Cr and Ni decrease with increasing SiO2 content in these volcanic and subvolcanic rocks (Fig. 10), which is consistent with fractionation of pyroxenes, amphibole and biotite. At SiO2 < 65 wt %, Yb (HREE) decreases and Dy/Yb ratios increase as SiO2 content increases (Fig. 10), which implies that cryptic garnet fractionation occurred at depth. Al2O3 and Sr abundances firstly increase and then decrease, and Eu*/Eu is constant (Fig. 10) at SiO2 < 65 wt %, which might indicate that plagioclase was not a dominant phase fractionated at an early stage. Zirconium, Hf, Th, LREE and HREE contents increase significantly in the more felsic volcanic and sub-volcanic rocks (Fig. 10), indicating that these elements behaved incompatibly in the late stages of crystallization. Previous studies have proposed that metaluminous magmas can evolve to peraluminous compositions by closed-system fractional crystallization of amphibole and clinopyroxene (Cawthorn & Brown, 1972; Zen, 1988; Alonso-Perez et al., 2009). However, the coexistence of various types of garnet (Figs 3 and 4) and plagioclase (Fig. 2e and f) clearly indicates an open-system evolution for their host magmas. In addition, for the Saierqingou and Dewulu volcanic rocks and Fandelongwa porphyries, with increasing SiO2 abundance, the ISr values increase, and εNd(t) and zircon εHf(t) values decrease (Fig. 16), indicating that AFC processes played an important role in the evolution of these rocks. These observations support the argument that peraluminous garnet-bearing felsic magmas could have resulted from AFC processes involving metalumiunous andesitic magmas (Barley, 1987; Harangi et al., 2001). Fig. 16 View largeDownload slide Plots of (a) ISrv. SiO2, (b) εNd(t) v. SiO2 and (c) zircon εHf(t) v. SiO2 for the igneous rocks described in this study. SiO2 for pyroclastic sample 0984 is an average of SiO2 values (∼60 wt %) for the andesites described in this study. Grey circles with error bars represent average zircon εHf(t) values and 2-sigma uncertainty. Fig. 16 View largeDownload slide Plots of (a) ISrv. SiO2, (b) εNd(t) v. SiO2 and (c) zircon εHf(t) v. SiO2 for the igneous rocks described in this study. SiO2 for pyroclastic sample 0984 is an average of SiO2 values (∼60 wt %) for the andesites described in this study. Grey circles with error bars represent average zircon εHf(t) values and 2-sigma uncertainty. Petrographic and geochemical data indicate that two-stage AFC processes at different depths could have been involved in the petrogenesis of these peraluminous igneous rocks. To model these processes, we have assumed that Saierqingou andesitic sample 09129 could represent the most primitive magma from which other magmas were derived. The presence of garnet indicates that the AFC process initially occurred in the deep crust. The mineral assemblage, observed in the MSD andesites, consists of orthopyroxene, clinopyroxene, garnet, plagioclase, amphibole and biotite, which is consistent with experimental results for crystallization of andesitic liquids at high pressure (12 kbar) (Alonso-Perez et al., 2009). The assimilated material could be represented by the Sangke leucogranite. At a second stage, some of these magmas, like the Fandelongwa porphyry, may have been stored in the shallow crust. The mineral assemblage in these rocks is plagioclase, K-feldspar, quartz, amphibole and biotite. Sedimentary wall-rocks could represent a shallow level crustal contaminant (Liu et al., 2008b); this would be consistent with the presence of sedimentary xenoliths. The detailed parameters used in the models are given in Table 4 and the results of AFC modeling (Depaolo, 1981) are illustrated in Fig. 14. The modeling results for lower crust contamination (curves A-AFC-Lower crust in Fig. 14) show that a less evolved dacitic magma (equivalent to sample 0991) could be generated by ∼50% fractional crystallization of andesite by removal of an assemblage consisting of 40 vol.% orthopyroxene, 35 vol. % clinopyroxene, 5 vol. % garnet, 10 vol. % biotite, 5 vol. % plagioclase, and 5 vol. % amphibole, a high r value (∼0·5), a relatively high DNd (0·70) and a low DSr (0·20). The modeling results for upper crust contamination (curves A-AFC-Upper crust in Fig. 14) show that derivation of the highly evolved rhyolitic magma (represented by sample 0987) requires a further 20–40% crystallization of 40 vol. % plagioclase, 30 vol. % K-feldspar, 20 vol. % quartz, 5 vol. % amphibole and 5 vol. % biotite, a high r value of ∼0·5, a relatively low DNd (0·39) and a high DSr (2·91). In conclusion, we suggest that the peraluminous garnet-bearing dacites and porphyries and crystal-poor rhyolites were produced by fractional crystallization of andesitic magmas, accompanied by assimilation of crustal materials and entrainment of a range of crystal assemblages at different crustal levels. Petrogenesis of the leucogranites The Sangke leucogranites have geochemical compositions similar to those of typical North Himalaya leucogranites that are interpreted to have been generated by melting of metasedimentary rocks (Zhang et al., 2004; King et al., 2011; Guo & Wilson, 2012; Gao & Zeng, 2014), except that the former exhibit greater HREE depletion (Figs 10 and 11). Because in granitic systems, Rb, Sr and Ba are mainly hosted by micas and feldspars, the ratios and concentrations of there elements can be used to constrain the source composition and petrogenetic processes (Inger & Harris, 1993). Anatectic melts generated by dehydration melting of metapelites usually have higher Rb/Ba and Rb/Sr ratios than melts derived from meta-greywackes. The Sangke leucogranites have relatively high Rb/Sr (2·1–3·3) and Rb/Ba (1·0–1·8) ratios, implying that metapelites may have been involved in their source. On an Rb/Sr v. Ba diagram (Fig. 17a), the data for the Sangke leucogranites display a negative correlation trend, which is a characteristic of some North Himalaya leucogranites that are suggested to be produced by dehydration melting of muscovite under fluid-absent condition (Types I and II) (Fig. 17a) (Zhang et al., 2004; King et al., 2011; Guo & Wilson, 2012; Huang et al., 2013). However, the leucogranites that are interpreted to be derived from fluid-fluxed muscovite melting (Type III) (Fig. 17a) (Aoya et al., 2005; Gao & Zeng, 2014) and those that are argued to be generated by fluid-fluxed biotite melting (Type IV) (Zhang et al., 2004; King et al., 2011) commonly show constant Rb/Sr ratios and variable Ba contents (Fig. 17a). Thus, the Sangke leucogranites are most likely to have originated from muscovite dehydration melting of metapelites under fluid-absent conditions. Fig. 17 View largeDownload slide Plots of (a) Rb/Sr v. Ba (Inger & Harris, 1993) and (b) (Dy/Yb)Nv. YbN for the Sangke (SK) leucogranites. Mu(VP), vapour saturated muscovite melting; Mu(VA), vapour absent muscovite melting; Bi(VA), vapour absent biotite melting. Data for North Himalaya leucogranites (Types I-IV) are the same as in Figure 12. Yellow circles with grey line represent the amount garnet in the residue calculated by the method of Guo & Wilson (2012). Fig. 17 View largeDownload slide Plots of (a) Rb/Sr v. Ba (Inger & Harris, 1993) and (b) (Dy/Yb)Nv. YbN for the Sangke (SK) leucogranites. Mu(VP), vapour saturated muscovite melting; Mu(VA), vapour absent muscovite melting; Bi(VA), vapour absent biotite melting. Data for North Himalaya leucogranites (Types I-IV) are the same as in Figure 12. Yellow circles with grey line represent the amount garnet in the residue calculated by the method of Guo & Wilson (2012). Particularly, the Sangke leucogranites display strongly fractionated REE patterns with very high (Dy/Yb)N ratios (14–22) and extreme HREE depletion (YbN = 0·14–0·22) (Figs 11f and 17b), relative to the North Himalaya leucogranites. Most of the North Himalaya leucogranites (Types I and III) have a limited range of YbN (3–10) and (Dy/Yb)N (1–2) (Fig. 17b), indicating minor or no residual garnet in their source. The Type IV Himalaya leucogranites have high YbN (10–50) and low (Dy/Yb)N (0·65–1·3) (Fig. 17b), indicating entrainment of some garnet (Zhang et al., 2004; King et al., 2011). The Type II Himalaya leucogranites have low YbN (1·3–5·8) and high (Dy/Yb)N (2–4) (Fig. 17b); trace element modeling of non-modal batch melting indicates that the amount of residual garnet in their source is 7–10% (Guo & Wilson, 2012; Huang et al., 2013). The Sangke leucogranites have lower YbN and higher (Dy/Yb)N values than those of the Type II Himalaya leucogranites (Fig. 17b), which we interpret to indicate that at least 10% residual garnet existed in their source. The visibly negative Eu anomalies of the Sangke leucogranites (Fig. 11f) were controlled by either plagioclase fractionation during magma differentiation or abundant residual plagioclase in the source. The homogeneous composition of the Sangke leucogranites indicates that this feature was mostly inherited from the source. The Paleozoic sedimentary rocks from the West Qinling Orogen have been metamorphosed at low-grade and they do not contain garnet; thus, they are unsuitable as a magma source for the Sangke leucogranites. The source rock may be represented by the garnet-rich metapelitic xenoliths (Fig. 2g). The Sangke leucogranites have high ISr and negative εNd(t) and zircon εHf(t) values (Fig. 12), coupled with Nd–Hf isotope model ages of ∼1·8–2·1 Ga (Table 3 and Supplementary Data Table S7). These results indicate that their magma source rocks are not exposed and may be old Proterozoic meta-supracrustal rocks. The influence of recycling of crystal cargoes on the whole-rock composition The mechanical incorporation of crystals of multiple origins (xenocrysts, antecrysts and autocrysts) may have had a pronounced influence on the final whole-rock compositions. Melt inclusion, mineral composition, whole-rock geochemistry and experimental studies indicate that many andesites worldwide are not equivalent to magmatic liquids, but are mixtures of felsic melts, crystals and lithic fragments from various crustal and mantle sources (Reubi & Blundy, 2009; Bach et al., 2012; Price et al., 2012; Parat et al., 2014), and even some apparently high-MgO primitive basalts may have been affected by this mechanism (Smith et al., 2010; Larrea et al., 2013; Parat et al., 2014; Cassidy et al., 2015). In addition, recycling of residual and peritectic phases (Chappell et al., 1987, Stevens et al., 2007, Dorais et al., 2009) has been proposed as a process affecting chemical variations in peraluminous magmas. Our observations presented in this study support the view that the interpretation of whole-rock compositional variation should be very closely linked with petrography and mineral chemistry (e.g. Reubi & Blundy, 2009; Smith et al., 2010; Price et al., 2012). Implications for the interpretation of the petrogenesis of high-Mg andesites Kamei et al. (2004) have suggested that Cenozoic high-Mg andesites (HMA) can be divided into four types: sanukitic HMA, adakitic HMA, bajaitic HMA, and boninitic HMA. Archean high-Mg sanukitoids have geochemical compositions similar to adakitic and bajaitic HMA, which are characterized by strong LILE and LREE enrichment, HREE depletion, and high Sr, Sr/Y and (La/Yb)N (Smithies & Champion, 2000). The boninitic HMA have extremely high Mg, Cr and Ni contents, and very low TiO2, HFSE and REE contents (Kamei et al., 2004). The sanukitic HMAs also display LILE and LREE enrichment, but they have low Sr contents and low Sr/Y and (La/Yb)N ratios (Tatsumi, 2006). The compositions of the MSD high-Mg andesites, with low Sr contents and low Sr/Y and (La/Yb)N ratios, are broadly similar to those of the Setouchi sanukitic HMA (Fig. 11a and b), which were proposed to be generated by interaction between slab sediment-derived melts and the overriding mantle wedge (Tatsumi, 2006). However, there are remarkable differences in petrography between the MSD andesites and the Setouchi sanukitic HMA. The MSD andesites are strongly porphyritic (35–50 vol. %) with plagioclase as the primary phenocryst (Fig. 2b and c). In contrast, the Setouchi sanukitic HMA contain <10 vol. % phenocrysts and the phenocryst assemblage is dominated by olivine without plagioclase (Tatsumi, 2006). These features rule out the possibility that the MSD high-Mg andesites are direct equivalents of typical primary, high-Mg andesitic melts generated at mantle depths. The MSD high-Mg andesites are more comparable to garnet-bearing andesites from the Northland arc, New Zealand (Fig. 13), which have been interpreted to represent mixtures of diverse liquids and complex crystal cargoes (Bach et al., 2012). These observations indicate that intracrustal processes (magma mixing coupled with crystal recycling) may be a widespread mechanism for the generation of high-Mg andesitic magmas in different tectonic settings, such as active continental margins (Streck et al., 2007; Chiaradia et al., 2014), oceanic arcs (Smith et al., 2010), and intracontinental environments (Qian & Hermann, 2010; Chen et al., 2013; Zhang et al., 2013). Thus, it is critical to identify whether the high-Mg andesitic samples described here can represent primary magmas derived directly from a mantle source. Implications for chemical diversity in peraluminous magmas Except for leucogranites, the compositions of many peraluminous granites do not match those of experimental melts (Montel & Vielzeuf, 1997; Patiño Douce, 1999). Several processes have been proposed to explain this geochemical contrast, including restite unmixing, fractional crystallization, source heterogeneity, magma mixing and mingling, assimilation of crustal rocks by metaluminous magmas and entrainment of peritectic phases (Cawthorn & Brown, 1972; Chappell & White, 1974; Chappell et al., 1987; Barbarin, 1996; Stevens et al., 2007; Villaros et al., 2009). One of the conclusions from our study is that AFC processes can be an effective mechanism for generating peraluminous magmas. We have demonstrated that garnets in the studied rocks have diverse origins (orthocrysts, antecrysts, xenocrysts or peritectic phase). The garnet-bearing peraluminous volcanic and subvolcanic rocks examined in this study have K2O/Na2O < 1, in contrast to the potassic (K2O/Na2O > 1) peraluminous melts generated by melting of metasediments (Montel & Vielzeuf, 1997; Patiño Douce & Harris, 1998;  Patiño Douce, 1999) (Fig. 10e). Moreover, the Saierqingou felsic volcanic rocks are characterized by increasing Th, LREE, HREE and Zr and constant P2O5 contents with increasing SiO2 abundance (Fig. 10). These trends are typical of highly fractionated I-type granites, rather than highly fractionated S-type granite suites (Chappell, 1999). Therefore, one should be cautious when defining igneous rocks as S-type on the basis of A/CNK >1·1 and the presence of garnet. Compared with the North Himalayan leucogranites, the Sangke leucogranites are characterized by strong depletion of HREE, suggesting abundant residual garnet in their source. The North Himalayan leucogranites display wide variations in their trace element and REE patterns (Fig. 11), which have been attributed to source heterogeneity and variable melting conditions (Zhang et al., 2004; King et al., 2011; Guo & Wilson, 2012; Gao & Zeng, 2014). Particularly, the Type II North Himalayan leucogranites show HREE enrichment, which has been interpreted to be related to the entrainment of peritectic garnets from fluid-fluxed melting of biotite (King et al., 2011). Overall, the entrainment of peritectic assemblages may be universal, but it is not a major factor in controlling the chemical variation of peraluminous magmas. Crystals with different origins could be incorporated into host magmas to modify their chemical compositions. Reconstruction of the magma plumbing system Petrographic features, mineral zoning and geochemical data have been used to show that complex intracrustal processes were involved in the genesis of the high-Mg andesites, peraluminous garnet-bearing dacites and porphyries, crystal-poor rhyolites and leucogranites from the West Qinling Orogen. One of the most critical steps for reconstructing the whole magmatic plumbing system from which all the magmas represented by these rocks were derived is to constrain the pressure and temperature conditions for different components along the different evolutionary pathways. Temperature and pressure estimations Experimental (Green, 1982; Montel & Vielzeuf, 1997; Stevens et al., 1997; Alonso-Perez et al., 2009) and petrological studies (Day et al., 1992; Harangi et al., 2001; Villaros et al., 2009; Bach et al., 2012) have demonstrated that garnets with low MnO and CaO (< 4 wt %) generally crystallize from peraluminous magmas at low pressure (4 to 7 kbar) and temperature (700–800°C), whereas those crystallizing in metaluminous magmas have high CaO (> 4 wt %) and form at high pressures (7 to 12 kbar) and higher temperatures (700–1000°C). Therefore, some Type-1 garnets and garnets reported by Wang & Ding (1990) with high CaO (Fig. 15) may have formed at 7–12 kbar and 700–1000°C, while Type-2 and Type-3 garnets related to peraluminous magmas with lower CaO and MnO contents (Fig. 15) may have formed at pressures above 5–7 kbar. Type-4 garnets with variable compositions indicate a wider range of pressure. Alonso-Perez et al. (2009) show that MnO content increases and CaO content is constant in garnets when the crystallization pressure of andesitic magmas decreases from 12 to 8 kbar (Fig. 15). Thus, some Type-1 garnets rims with high MnO contents might have crystallized from andesitic magmas at a relatively low pressure. Ti-in-zircon thermometry (TTiZ) can also be applied to estimate the crystallization temperatures of magmas (Supplementary Data Table S6) (Ferry & Watson, 2007). Except for andesitic sample 09133, quartz is present in all the studied samples, thus, the activity of SiO2 (aSiO2) is considered to be close to one. The absence of rutile in the samples indicates aTiO2 < 1, and the presence of titanite and ilmenite indicates that aTiO2 is likely to have been > 0·5 (Hayden & Watson, 2007). In applying Ti-in-zircon thermometry, we have assumed an intermediate aTiO2 = 0·7 for all samples (e.g. Claiborne et al., 2010). The Saierqingou pyroclastic rock (0984) and andesite (09133) gives TTiZ of 770 to 933°C (mean = 857 ± 32°C, n = 14) and 700 to 802 °C (mean = 759 ± 19°C, n = 12), respectively (Fig. 18a). The Dewulu dacite (0991) gives TTiZ = 634–821°C (mean = 724 ± 43°C, n = 11) (Fig. 18a). The Fandelongwa porphyry (0981) yields TTiZ = 641–767°C (mean = 708 ± 35°C, n = 9) (Fig. 18a). The Sangke leucogranite (09117) has TTiZ = 674–833°C (mean = 783 ± 25°C, n = 13), which straddles the temperature for the onset of dehydration melting of muscovite (Fig. 18). Fig. 18 View largeDownload slide Pressure and temperature estimates for the formation and evolution of the igneous rocks described in this study. (a) SiO2v. Ti-in-Zircon temperature (TTiz). Grey circles with error bars represent weighted average TTiz temperatures and 2-sigma uncertainty. (b) Phase diagram for andesite + 5 wt % H2O over a range of pressures and temperatures (Green, 1982). Muscovite dehydration melting solidus is from Patiño Douce (1999). Mineral assemblages in bold were observed in this study. The dashed black line with arrowhead (L1) is the potential evolutionary path for the garnet-bearing andesites. The grey line with arrowhead (L2) displays P–T ranges for the generation of the Sangke leucogranite. The grey circles are P–T values obtained for the garnet from the volcanic rocks of the Northland Arc, New Zealand (Bach et al., 2012). L, liquid; Px, pyroxene; Cpx, clinopyroxene; Grt, garnet; Am, amplhibole; Pl, plagioclase; Qz, quartz. Fig. 18 View largeDownload slide Pressure and temperature estimates for the formation and evolution of the igneous rocks described in this study. (a) SiO2v. Ti-in-Zircon temperature (TTiz). Grey circles with error bars represent weighted average TTiz temperatures and 2-sigma uncertainty. (b) Phase diagram for andesite + 5 wt % H2O over a range of pressures and temperatures (Green, 1982). Muscovite dehydration melting solidus is from Patiño Douce (1999). Mineral assemblages in bold were observed in this study. The dashed black line with arrowhead (L1) is the potential evolutionary path for the garnet-bearing andesites. The grey line with arrowhead (L2) displays P–T ranges for the generation of the Sangke leucogranite. The grey circles are P–T values obtained for the garnet from the volcanic rocks of the Northland Arc, New Zealand (Bach et al., 2012). L, liquid; Px, pyroxene; Cpx, clinopyroxene; Grt, garnet; Am, amplhibole; Pl, plagioclase; Qz, quartz. These results agree well with earlier temperature and pressure estimations obtained by Li et al. (2013). Using clinopyroxene-melt thermobarometers (Putirka, 2008b), they calculated temperatures and pressures distributed in two clusters: one group (n = 7) gave an average pressure of  7 ± 1 kbar and temperature of  975 ± 20 °C; the other group (n = 22) gave an average pressure of 11 ± 1 kbar and temperature of 1050 ± 25°C (Fig. 18b). The calculated temperatures are higher than the TTiZ values for andesitic sample (09133) (Fig. 18). This difference is probably due to the fact that the clinopyroxenes could be antecrysts or the earliest crystallization phases from more mafic magmas, while zircon may have not been present at that time. As suggested above, garnet-bearing Saierqingou dacites and Fandelongwa porphyries could have been produced by fractional crystallization of andesites and assimilation of crustal materials. Thus, a phase equilibrium diagram for andesitic magma with 5 wt % H2O can be used to constrain the P–T conditions (Green, 1982) (Fig. 18b). For the earlier stage of deep crustal evolution, the assemblage garnet–clinopyroxene crystallized at 12–14 kbar and 980–1050°C. The assemblage plagioclase + pyroxenes + garnet ± amphibole in the andesitic magmas dominantly crystallized at 8–14 kbar and 860–960°C. Quartz is rare in the andesitic magmas but abundant in the garnet-bearing dacitic magmas, indicating that the dacitic magmas mainly crystallized at pressures > 8 kbar over relatively a low and wide range of temperatures (650–860°C) (Fig. 18b). In summary, the high-Mg andesites and garnet-bearing igneous rocks examined in this study may have crystallized at high pressures (7–12 kbar) and over a wide range of temperatures (650–1000°C) (Fig. 18). The Sangke leucogranites may have been generated at pressure above 5–7 kbar and temperature of 674–830°C (Fig. 18). An integrated petrogenetic model There are adjacent and coeval high Sr/Y igneous rocks in the West Qinling Orogen that have been recently described by Huang et al. (2014) and Luo et al. (2015). They suggested that the high Sr/Y magmas represented by these rocks could have been generated at pressures of 10–12 kbar, equivalent to a crustal thickness of 30–40 km. This implies that there was a thickened crust (at least > 40 km) beneath the West Qinling Orogen during the Middle Triassic. Ti-in-zircon thermometry, garnet compositions, clinopyroxene-melt thermobarometry, and experimental studies have provided approximate first-order estimations of pressure and temperature conditions (Fig. 18). These results indicate that multiple magma reservoirs co-existed at various crustal levels. On the basis of the above arguments presented above, we propose an integrated petrogenetic model (Fig. 19) to explain the complex intra-crustal processes responsible for forming the diverse lithological and mineral characteristics of the igneous rocks of the West Qinling Orogen. Fig. 19 View largeDownload slide A schematic model for the intra-crustal plumbing system in which the magmas represented by the rocks described in this study were generated. (a) The successive under-plating or intra-plating of mantle-derived basaltic magmas formed MASH or deep crustal hot zones in which fractional crystallization and primitive mafic magma recharge occurred, inducing crustal anatexis. Mixing between evolved basaltic magma and crustal melts generated hybrid andesitic magmas. These ascended rapidly and erupted at the surface and incorporated diverse crystal cargoes along the way. Crystal-rich high-Mg andesites were generated by these processes. (b) The ascent of some hybrid andesitic magmas stalled at different crustal levels where they evolved, through assimilation and fractional crystallization, into dacitic and rhyolitic magmas. Various types of garnet were carried up to the upper crust by the rapid ascent of these dacitic magmas. Extreme HREE-depleted leucogranites formed by anatectic melting of metasediments with abundant garnet being retained in the residue. Fig. 19 View largeDownload slide A schematic model for the intra-crustal plumbing system in which the magmas represented by the rocks described in this study were generated. (a) The successive under-plating or intra-plating of mantle-derived basaltic magmas formed MASH or deep crustal hot zones in which fractional crystallization and primitive mafic magma recharge occurred, inducing crustal anatexis. Mixing between evolved basaltic magma and crustal melts generated hybrid andesitic magmas. These ascended rapidly and erupted at the surface and incorporated diverse crystal cargoes along the way. Crystal-rich high-Mg andesites were generated by these processes. (b) The ascent of some hybrid andesitic magmas stalled at different crustal levels where they evolved, through assimilation and fractional crystallization, into dacitic and rhyolitic magmas. Various types of garnet were carried up to the upper crust by the rapid ascent of these dacitic magmas. Extreme HREE-depleted leucogranites formed by anatectic melting of metasediments with abundant garnet being retained in the residue. (1) Partial melting of enriched lithospheric mantle generated basaltic magmas (Guo et al., 2012; Li et al., 2014) and these melts successively underplated the crust to form MASH (Hildreth & Moorbath, 1988) or deep crustal hot zones (Annen et al., 2006). Within these zones, mantle-derived magmas were repeatedly modified by fractional crystallization involving high Mg# clinopyroxene and orthopyroxene and high An plagioclase (Li et al., 2013), and ultramafic and mafic cumulates were formed (Fig. 19a). Clinopyroxene with reverse zoning and orthopyroxene with complex zoning (Fig. 5a and b) indicate that magma recharge occurred in these deep crustal zones. (2) The basaltic magmas provided enough heat and volatiles to induce crustal anatexis. Mixing between evolved basaltic magmas and crustal melts generated hybrid andesitic magmas (Fig. 19a). The hybrid andesitic magmas incorporated abundant early-stage antecrysts (high Mg# pyroxenes, high An plagioclase and cumulates). This was followed by crystallization of low Mg# pyroxenes as thin overgrowth rims, low An matrix plagioclase (Li et al., 2013) and cryptic Type-1garnet on a short timescale. Magmas derived at this stage were rapidly erupted to form some of the MSD high-Mg andesites (Fig. 19a). Evolved dacitic magmas preferentially incorporated abundant plagioclase and minor pyroxenes, resulting in the formation of some of the MSD high-Al andesites (Fig. 19a). (3) Some hybrid andesitic magmas did not erupt and were stored at relatively high levels in the middle and lower crust (Fig. 19b). These magmas continuously evolved from metaluminous to strongly peraluminous through fractional crystallization of plagioclase, pyroxenes, garnet, quartz, amphibole and biotite and assimilation of anatectic melts derived from dehydration melting of metasedimentary rocks. Diverse types of garnet were generated during this process (Fig. 19b). Type-1 garnets (high CaO and low MnO) are antecrysts that crystallized from andesitic magmas at high pressure. Some Type-1 garnet rims with high CaO and MnO may have grown within andesitic magmas at a relatively low pressure, and some garnets rims with low CaO and MnO may have formed by interaction with anatectic melts. Type-3 garnets are peritectic phases that were generated by dehydration melting of metapelitic xenoliths. They also interacted with andesitic magmas to form reaction rims with compositions similar to those of low-pressure Type-1 garnets. Type-2 garnets are autocrysts that crystallized from evolved peraluminous dacitic magmas. Type-4 garnets are xenocrysts that were carried to surface by discrete magma batches from various depths. (4) Evolved peraluminous dacitic magmas that had entrained different types of garnet and plagioclase and a small amount of metasedimentary xenoliths, were erupted to form the Saierqingou dacites and, or, intruded into the upper crust (<3 km) to form the Fandelongwa porphyries (Fig. 19b). Some dacitic magmas were further modified by fractional crystallization and assimilation of sedimentary rocks in upper crustal magma reservoirs. The Fandelongwa porphyries may have evolved in a similar way, but it is more likely that they represent crystal mushes (Fig. 19b). From the upper parts of these crystal mushes, crystal-poor melts erupted to form the Saierqingou crystal-poor rhyolites (Fig. 19b). In addition, anatectic melts were derived from dehydration melting of metapelite, leaving abundant residual garnet (Type-3) in the source. These anatectic melts intruded into the upper crust (<3 km) to form the Sangke leucogranites (Fig. 19b). Multiple magma reservoirs could have coexisted and been distributed throughout the entire crust, feeding magmas represented by the high-Mg andesites, peraluminous garnet-bearing igneous rocks and leucogranites higher into shallower crustal levels and ultimately to the surface (Fig. 19). These magmas evolved through polybaric processes operating at different crustal depths. Based on our new data and observations in combination with previous work, we are able to suggest that the crust-mantle interaction and crustal differentiation mainly occurred in MASH (Hildreth & Moorbath, 1988) or deep crustal hot zones (Annen et al., 2006) in the middle-to-lower crust, where diverse intermediate and felsic magmas were generated by complex intra-crustal processes. Tectonic control for the eruption of the highly porphyritic and garnet-bearing magmas Garnet-bearing magmas are uncommon and garnet formed at high pressure can only be preserved in the shallow crust under particular conditions (e.g. along deep-seated faults, in extensional regimes or with high volatile contents) where rapid magma ascent can occur (Day et al., 1992; Harangi et al., 2001; Bach et al., 2012). However, high crystal contents will change the rheological behavior of magmas, increasing the viscosity and density, which is unfavorable for rapid magma ascent. This raises an important question about how these crystal-rich and garnet-bearing magmas erupted on the surface. It has been suggested that garnet-bearing volcanic rocks are mainly formed in extensional regimes, either related to subduction (Fitton, 1977; Bach et al., 2012) or in post-collisional settings (Barley, 1987; Gilbert & Rogers, 1989; Harangi et al., 2001). Comparatively, there are two conflicting views about the generation of the Middle Triassic magmatism in the West Qinling orogen: (1) a subduction setting (Guo et al., 2012; Li et al., 2013, 2014; Huang et al., 2014) and (2) a post-collisional setting (Zhang et al., 2006; Zhang et al., 2008; Luo et al., 2012, 2015). A petrogenic model whereby slab sediment-derived melts interacted with the overriding mantle wedge has been proposed for the generation of the Maixiu high-Mg andesites (Li et al., 2013), which would imply that the A’nimaque Ocean was still subducting in the Middle Triassic. However, the Late Gequ molasse formation, which has a basal conglomerate, lies unconformably over the A’nimaque ophiolitic mélange, leading Chen et al. (2010) to advocate that the A’nimaque Ocean had been closed by the end of the Middle Permian. Combining findings from our study with information available for the regional geology, we suggest that the Middle Triassic magmatism of the West Qinling occurred in an early-stage, post-collisional, extensional tectonic regime, which resulted from the break-off of the subducted A’nimaque Oceanic slab (Zhang et al., 2006; Luo et al., 2012, 2015). For the eruption of crystal-rich magmas, some researchers have suggested that several prerequisites must be satisfied, including: (1) the presence of preexisting early cumulates or crystal mushes; (2) a rapid ascent velocity of the host magmas, which is greater than the settling rate of the entrained crystals; and (3) a short crustal residence time for the host magmas without extensive modification in shallow magma reservoirs (Smith et al., 2010; Lange et al., 2013; Cassidy et al., 2015). The petrographic observations, mineral compositions and zoning indicate that the magmatic system in which it is proposed that these Middle Triassic igneous rocks were generated meets all the above conditions. Middle Triassic igneous rocks in West Qinling are mainly distributed along strike-slip faults and they were emplaced parallel to major tectonic lineaments (Fig. 1c), which implies that the fault system has promoted rapid ascent of crystal-rich magmas in an extensional tectonic regime (e.g. Lange et al., 2013; Cassidy et al., 2015). CONCLUSIONS During the Middle Triassic, the break-off of the subducted A’nimaque oceanic slab caused asthenospheric upwelling which induced partial melting of overlying enriched lithospheric mantle to generate hydrous basalts. These basaltic magmas repeatedly under-plated or intra-plated the mid-lower crust, forming MASH or deep crustal hot zones, in which diverse intermediate and felsic magmas were generated by complex intracrustal processes, including fractional crystallization, recharge, crustal anatexis, magma mixing and mingling, assimilation and random entrainment of various crystal cargoes and crustal xenoliths. (1) The MSD high-Mg andesites in the three volcanic suites are crystal-rich and are characterized by high MgO (Mg#), Cr and Ni contents. Nevertheless, they cannot be equivalent to primitive high-Mg andesitic liquids. They are blends of felsic melts and complex crystal assemblages with diverse origins. (2) The peraluminous garnet-bearing dacites and porphyries and crystal-poor rhyolites were produced by fractional crystallization of andesitic magmas, assimilation of crustal materials and entrainment of crystals of different origins at different crustal levels. The leucogranites were produced by muscovite dehydration melting of metapelites under fluid-absent conditions. They show extreme depletion of HREE and pronounced negative Eu anomalies, indicating that garnet and plagioclase were abundant as residual phases in their sources. (3) Four types of garnet have been identified in this study: Type-1 are antecrysts that crystallized from andesitic magmas; Type-2 are orthocrysts from peraluminous magmas; Type-3 are peritectic phases formed by melting reactions; Type-4 are xenocrysts entrained from metapelitic wall-rocks. Consistent with the whole-rock geochemical and Sr–Nd–Hf isotopic data, the various types of garnets record the petrogenetic links and the complex interactions between magmas represented by these igneous rocks. (4) The incorporation of diverse crystal populations into magmas would inevitably effect the final bulk-rock compositions of the igneous rocks. The interpretation of whole-rock composition variation should, therefore, be very closely linked with petrography and mineral chemistry. (5) Multiple dispersed magma reservoirs could have co-existed throughout the entire crust. The high-Mg andesites, peraluminous garnet-bearing magmatic rocks and leucogranites evolved and interacted in these separated magma storage regions on their way to the surface. The rapid ascent and eruption of highly porphyritic and garnet-bearing magmas may have been closely associated with a post-collisional extensional tectonic regime and promoted by fault systems. Acknowledgements We are grateful to Tod Waight, Jennifer Garrison, Szabolcs Harangi and editor Richard Price for their thoughtful and insightful reviews, which greatly helped us to improve the paper. We also sincerely thank Yongsheng Liu and Zhaochu Hu for zircon U–Pb dating and Hf isotopic analyses, Detao He and Chunfei Chen for the preparation of thin sections, and Hua Huang, Chao Zhang and Yuan-Bao Wu for their helpful discussion. Funding This research was supported by Natural Science Foundation of China (No. 41403026), China Postdoctoral Science Foundation Grant (No. 2014M552107), China Geological Survey (No. 12120113100900) and MOST Special Fund from State Key Laboratory of Geological Processes and Mineral Resources (MSFGPMR201601–2). SUPPLEMENTARY DATA Supplementary data are available at Journal of Petrology online. REFERENCES Acosta-Vigil A. , Buick I. , Hermann J. , Cesare B. , Rubatto D. , London D. , Morgan G. B. ( 2010 ). Mechanisms of crustal anatexis: a geochemical study of partially melted metapelitic enclaves and host dacite, SE Spain . Journal of Petrology 51 , 785 – 821 . Google Scholar CrossRef Search ADS Alonso-Perez R. , Müntener O. , Ulmer P. ( 2009 ). Igneous garnet and amphibole fractionation in the roots of island arcs: experimental constraints on andesitic liquids . Contributions to Mineralogy and Petrology 157 , 541 – 558 . Google Scholar CrossRef Search ADS Annen C. , Blundy J. D. , Sparks R. S. J. ( 2006 ). The genesis of intermediate and silicic magmas in deep crustal hot zones . Journal of Petrology 47 , 505 – 539 . Google Scholar CrossRef Search ADS Annen C. , Blundy J. D. , Leuthold J. , Sparks R. S. J. ( 2015 ). Construction and evolution of igneous bodies: towards an integrated perspective of crustal magmatism . Lithos 230 , 206 – 221 . Google Scholar CrossRef Search ADS Aoya M. , Wallis S. R. , Terada K. , Lee J. , Kawakami T. , Wang Y. , Heizler M. ( 2005 ). North-south extension in the Tibetan crust triggered by granite emplacement . Geology 33 , 853 – 856 . Google Scholar CrossRef Search ADS Bach P. , Smith I. E. M. , Malpas J. G. ( 2012 ). The origin of garnets in andesitic rocks from the Northland Arc, New Zealand, and their implication for sub-arc processes . Journal of Petrology 53 , 1169 – 1195 . Google Scholar CrossRef Search ADS Barbarin B. ( 1996 ). Genesis of the two main types of peraluminous granitoids . Geology 24 , 295 – 298 . Google Scholar CrossRef Search ADS Barley M. E. ( 1987 ). Origin and evolution of midcretaceous, garnet-bearing, intermediate and silicic volcanics from Canterbury, New-Zealand . Journal of Volcanology and Geothermal Research 32 , 247 – 267 . Google Scholar CrossRef Search ADS Blichert-Toft J. , Chauvel C. , Albarède F. ( 1997 ). Separation of Hf and Lu for high-precision isotope analysis of rock samples by magnetic sector-multiple collector ICP-MS . Contributions to Mineralogy and Petrology 127 , 248 – 260 . Google Scholar CrossRef Search ADS Cao X. F. , Lü X. B. , Yao S. Z. , Mei W. , Zou X. Y. , Chen C. , Liu S. T. , Zhang P. , Su Y. Y. , Zhang B. ( 2011 ). LA–ICP–MS U–Pb zircon geochronology, geochemistry and kinetics of the Wenquan ore-bearing granites from West Qinling, China . Ore Geology Reviews 43 , 120 – 131 . Google Scholar CrossRef Search ADS Cashman K. V. , Sparks R. S. J. , Blundy J. D. ( 2017 ). Vertically extensive and unstable magmatic systems: a unified view of igneous processes . Science 355 , eaag3055 . Google Scholar CrossRef Search ADS PubMed Cassidy M. , Edmonds M. , Watt S. F. L. , Palmer M. R. , Gernon T. M. ( 2015 ). Origin of basalts by hybridization in andesite-dominated arcs . Journal of Petrology 56 , 325 – 346 . Google Scholar CrossRef Search ADS Cawthorn R. G. , Brown P. A. ( 1972 ). A model for the formation and crystallization of corundum-normative calc-alkaline magmas through amphibole fractionation . Journal of Geology 26 , 523 – 533 . Cesare B. ( 2000 ). Incongruent melting of biotite to spinel in a quartz-free restite at El Joyazo (SE Spain): textures and reaction characterization . Contributions to Mineralogy and Petrology 139 , 273 – 284 . Google Scholar CrossRef Search ADS Chappell B. W. ( 1999 ). Aluminium saturation in I- and S-type granites and the characterization of fractionated haplogranites . Lithos 46 , 535 – 551 . Google Scholar CrossRef Search ADS Chappell B. W. , White A. J. R. ( 1974 ). Two contrasting granite types . Pacific Geology 8 , 173 – 174 . Chappell B. W. , White A. J. R. , Wyborn D. ( 1987 ). The importance of residual source material (restite) in granite petrogenesis . Journal of Petrology 28 , 1111 – 1138 . Google Scholar CrossRef Search ADS Chen B. , Jahn B.-M. , Suzuki K. ( 2013 ). Petrological and Nd-Sr-Os isotopic constraints on the origin of high-Mg adakitic rocks from the North China Craton: tectonic implications . Geology 41 , 91 – 94 . Google Scholar CrossRef Search ADS Chen S. J. , Li R. S. , Ji W. H. , Zhao Z. M. , Liu R. L. , Jia B. H. , Zhang Z. F. , Wang G. G. ( 2010 ). The Permian lithofacies paleogeographic characteristics and basin-mountain conversion in the Kunlun orogenic belt . Geology in China 37 , 374 – 393 . Chiaradia M. , Muntener O. , Beate B. ( 2014 ). Quaternary sanukitoid-like andesites generated by intracrustal processes (Chacana caldera complex, Ecuador): implications for Archean sanukitoids . Journal of Petrology 55 , 769 – 802 . Google Scholar CrossRef Search ADS Claiborne L. L. , Miller C. F. , Wooden J. L. ( 2010 ). Trace element composition of igneous zircon: a thermal and compositional record of the accumulation and evolution of a large silicic batholith, Spirit Mountain, Nevada . Contributions to Mineralogy and Petrology 160 , 511 – 531 . Google Scholar CrossRef Search ADS Clemens J. D. , Wall V. J. ( 1984 ). Origin and evolution of a peraluminous silicic ignimbrite suite—the Violet Town volcanics . Contributions to Mineralogy and Petrology 88 , 354 – 371 . Google Scholar CrossRef Search ADS Dahren B. , Troll V. R. , Andersson U. B. , Chadwick J. P. , Gardner M. F. , Jaxybulatov K. , Koulakov I. ( 2012 ). Magma plumbing beneath Anak Krakatau volcano, Indonesia: evidence for multiple magma storage regions . Contributions to Mineralogy and Petrology 163 , 631 – 651 . Google Scholar CrossRef Search ADS Davidson J. P. , Morgan D. J. , Charlier B. L. A. , Harlou R. , Hora J. M. ( 2007 ). Microsampling and isotopic analysis of igneous rocks: implications for the study of magmatic systems . Annual Review of Earth and Planetary Sciences 35 , 273 – 311 . Google Scholar CrossRef Search ADS Day R. A. , Green T. H. , Smith I. E. M. ( 1992 ). The origin and significance of garnet phenocrysts and garnet-bearing xenoliths in miocene calc-alkaline volcanics from Northland, New-Zealand . Journal of Petrology 33 , 125 – 161 . Google Scholar CrossRef Search ADS Depaolo D. J. ( 1981 ). Trace element and isotopic effects of combined wallrock assimilation and fractional crystallization . Earth and Planetary Science Letters 53 , 189 – 202 . Google Scholar CrossRef Search ADS DePaolo D. J. , Wasserburg G. J. ( 1979 ). Petrogenetic mixing models and Nd-Sr isotopic patterns . Geochimica et Cosmochimica Acta 43 , 615 – 627 . Google Scholar CrossRef Search ADS Dorais M. J. , Pett T. K. , Tubrett M. ( 2009 ). Garnetites of the Cardigan Pluton, New Hampshire: evidence for peritectic garnet entrainment and implications for source rock compositions . Journal of Petrology 50 , 1993 – 2016 . Google Scholar CrossRef Search ADS Erdmann S. , Jamieson R. A. , MacDonald M. A. ( 2009 ). Evaluating the origin of garnet, cordierite, and biotite in granitic rocks: a case study from the South Mountain Batholith, Nova Scotia . Journal of Petrology 50 , 1477 – 1503 . Google Scholar CrossRef Search ADS Feng Y. M. , Cao X. Z. , Zhang E. P. ( 2002 ). Structure, Orogenic Processes and Geodynamic of the the Western Qinling Orogen . Xi’an : Xi’an Map Press . Ferry J. M. , Watson E. B. ( 2007 ). New thermodynamic models and revised calibrations for the Ti-in-zircon and Zr-in-rutile thermometers . Contributions to Mineralogy and Petrology 154 , 429 – 437 . Google Scholar CrossRef Search ADS Fitton J. G. ( 1977 ). The genetic significance of almandine-pyrope phenocrysts in the calc-alkaline Borrowdale Volcanic Group, Northern England . Contributions to Mineralogy and Petrology 36 , 231 – 248 . Google Scholar CrossRef Search ADS Gao L. E. , Zeng L. S. ( 2014 ). Fluxed melting of metapelite and the formation of Miocene high-CaO two-mica granites in the Malashan gneiss dome, southern Tibet . Geochimica et Cosmochimica Acta 130 , 136 – 155 . Google Scholar CrossRef Search ADS Gao S. , Rudnick R. L. , Yuan H. L. , Liu X. M. , Liu Y. S. , Xu W. L. , Ling W. L. , Ayers J. , Wang X. C. , Wang Q. H. ( 2004 ). Recycling lower continental crust in the North China craton . Nature 432 , 892 – 897 . Google Scholar CrossRef Search ADS PubMed Gilbert J. S. , Rogers N. W. ( 1989 ). The significance of garnet in the Permo-Carboniferous volcanic rocks of the Pyrenees . Journal of the Geological Society 146 , 477 – 490 . Google Scholar CrossRef Search ADS Green T. ( 1982 ). Anatexis of mafic crust and high pressure crystallization of andesite. In: Thorpe R. S. (ed.) Andesites Orogenic Andesites and Realted Rocks . Chichester : John Wiley & Sons Ltd , pp. 465 – 487 . Grove T. L. , Elkins-Tanton L. T. , Parman S. W. , Chatterjee N. , Muntener O. , Gaetani G. A. ( 2003 ). Fractional crystallization and mantle-melting controls on calc-alkaline differentiation trends . Contributions to Mineralogy and Petrology 145 , 515 – 533 . Google Scholar CrossRef Search ADS Guo A. L. , Zhang G. W. , Sun Y. G. , Cheng S. Y. , Qiang J. ( 2007 ). Sr-Nd-Pb isotopic geochemistry of late-Paleozoic mafic volcanic rocks in the surrounding areas of the Gongge basin, Qinghai province and geological implications . Acta Petrologica Sinica 4 , 747 – 754 . Guo X. Q. , Yan Z. , Wang Z. Q. , Wang T. , Hou K. J. , Fu C. L. , Li J. L. ( 2012 ). Middle Triassic arc magmatism along the northeastern margin of the Tibet: U–Pb and Lu–Hf zircon characterization of the Gangcha complex in the West Qinling terrane, central China . Journal of the Geological Society 169 , 327 – 336 . Google Scholar CrossRef Search ADS Guo Z. F. , Wilson M. ( 2012 ). The Himalayan leucogranites: constraints on the nature of their crustal source region and geodynamic setting . Gondwana Research 22 , 360 – 376 . Google Scholar CrossRef Search ADS Harangi S. , Downes H. , Kosa L. , Szabo C. , Thirlwall M. F. , Mason P. R. D. , Mattey D. ( 2001 ). Almandine garnet in calc-alkaline volcanic rocks of the northern Pannonian Basin (eastern-central Europe): geochemistry, petrogenesis and geodynamic implications . Journal of Petrology 42 , 1813 – 1843 . Google Scholar CrossRef Search ADS Hayden L. A. , Watson E. B. ( 2007 ). Rutile saturation in hydrous siliceous melts and its bearing on Ti-thermometry of quartz and zircon . Earth and Planetary Science Letters 258 , 561 – 568 . Google Scholar CrossRef Search ADS Hildreth W. , Moorbath S. ( 1988 ). Crustal contributions to arc magmatism in the Andes of Central Chile . Contributions to Mineralogy and Petrology 98 , 455 – 489 . Google Scholar CrossRef Search ADS Hirose K. ( 1997 ). Melting experiments on Iherzolite KLB-1 under hydrous conditions and generation of high-magnesian andesitic melts . Geology 25 , 42 – 44 . Google Scholar CrossRef Search ADS Huang C. M. , Zhao Z. D. , Zhu D. C. , Liu D. , Huang Y. , Dung M. C. , Hu Z. C. , Zheng J. P. ( 2013 ). Geochemistry, zircon U–Pb chronology and Hf isotope of Luozha leucogranite, southern Tibet: implication for petrogenesis . Acta Petrologica Sinica 29 , 3689 – 3702 . Huang X. F. , Mo X. X. , Yu X. H. , Li X. W. , Yang M. C. , Luo M. F. , He W. Y. , Yu J. C. ( 2014 ). Origin and geodynamic settings of the Indosinian high Sr/Y granitoids in the West Qinling: an example from the Shehaliji pluton in Tongren area . Acta Petrologica Sinica 30 , 3255 – 3270 . Inger S. , Harris N. ( 1993 ). Geochemical constraints on leucogranite magmatism in the Langtang Valley, Nepal Himalaya . Journal of Petrology 34 , 345 – 368 . Google Scholar CrossRef Search ADS Jerram D. A. , Davidson J. P. ( 2007 ). Frontiers in textural and microgeochemical analysis . Elements 3 , 235 – 238 . Google Scholar CrossRef Search ADS Jerram D. A. , Martin V. M. ( 2008 ). Understanding crystal populations and their significance through the magma plumbing system . Geological Society of London Special Publications 304 , 133 – 148 . Google Scholar CrossRef Search ADS Jin W. J. , Zhang Q. , He D. F. , Jia X. Q. ( 2005 ). SHRIMP dating of adakites in western Qinling and their implications . Acta Petrologica Sinica 3 , 950 – 966 . Johannes W. , Holtz F. O. ( 1996 ). Petrogenesis and Experimental Petrology of Granitic Rocks . Berlin : Springer . Google Scholar CrossRef Search ADS Kamei A. , Owada M. , Nagao T. , Shiraki K. ( 2004 ). High-Mg diorites derived from sanukitic HMA magmas, Kyushu Island, southwest Japan arc: evidence from clinopyroxene and whole rock compositions . Lithos 75 , 359 – 371 . Google Scholar CrossRef Search ADS Kawabata H. , Takafuji N. ( 2005 ). Origin of garnet crystals in calc-alkaline volcanic rocks from the Setouchi volcanic belt, Japan . Mineralogical Magazine 69 , 951 – 971 . Google Scholar CrossRef Search ADS Keay S. , Collins W. J. , McCulloch M. T. ( 1997 ). A three-component Sr-Nd isotopic mixing model for granitoid genesis, Lachlan fold belt, eastern Australia . Geology 25 , 307 – 310 . Google Scholar CrossRef Search ADS Kemp A. I. S. , Hawkesworth C. J. , Foster G. L. , Paterson B. A. , Woodhead J. D. , Hergt J. M. , Gray C. M. , Whitehouse M. J. ( 2007 ). Magmatic and crustal differentiation history of granitic rocks from Hf-O isotopes in zircon . Science 315 , 980 – 983 . Google Scholar CrossRef Search ADS PubMed King J. , Harris N. , Argles T. , Parrish R. , Zhang H. ( 2011 ). Contribution of crustal anatexis to the tectonic evolution of Indian crust beneath southern Tibet . Geological Society of America Bulletin 123 , 218 – 239 . Google Scholar CrossRef Search ADS Kou X. H. , Zhu Y. H. , Zhang K. X. , Shi B. , Luo G. M. ( 2007 ). Discovery and geochemistry of upper Permian volcanic rocks in Tongren Area, Qinghai Province and their tectonic significance . Earth Science—Journal of China University of Geosciences 32 , 45 – 58 . Lackey J. S. , Romero G. A. , Bouvier A. S. , Valley J. W. ( 2012 ). Dynamic growth of garnet in granitic magmas . Geology 40 , 171 – 174 . Google Scholar CrossRef Search ADS Lange A. E. , Nielsen R. L. , Tepley F. J. , Kent A. J. R. ( 2013 ). The petrogenesis of plagioclase-phyric basalts at mid-ocean ridges . Geochemistry Geophysics Geosystems 14 , 3282 – 3296 . Google Scholar CrossRef Search ADS Larrea P. , Franca Z. , Lago M. , Widom E. , Gale C. , Ubide T. ( 2013 ). Magmatic processes and the role of antecrysts in the genesis of Corvo Island (Azores Archipelago, Portugal) . Journal of Petrology 54 , 769 – 793 . Google Scholar CrossRef Search ADS Le Bas M. J. L. , Le Maitre R. W. L. , Streckeisen A. , Zanettin B. ( 1986 ). A chemical classification of volcanic rocks based on the total alkali-silica diagram . Journal of Petrology 27 , 745 – 750 . Google Scholar CrossRef Search ADS Li X. W. , Mo X. X. , Bader T. , Scheltens M. , Yu X. H. , Dong G. C. , Huang X. F. ( 2014 ). Petrology, geochemistry and geochronology of the magmatic suite from the Jianzha Complex, central China: petrogenesis and geodynamic implications . Journal of Asian Earth Sciences 95 , 164 – 181 . Google Scholar CrossRef Search ADS Li X. W. , Mo X. X. , Huang X. F. , Dong G. C. , Yu X. H. , Luo M. F. , Liu Y. B. ( 2015 ). U–Pb zircon geochronology, geochemical and Sr-Nd-Hf isotopic compositions of the Early Indosinian Tongren Pluton in West Qinling: petrogenesis and geodynamic implications . Journal of Asian Earth Sciences 97 , 38 – 50 . Google Scholar CrossRef Search ADS Li X. W. , Mo X. X. , Yu X. H. , Ding Y. , Huang X. F. , Wei P. , He W. Y. ( 2013 ). 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 , 158 – 174 . Google Scholar CrossRef Search ADS Liu H. J. , Chen Y. J. , Mao S. D. , Zhao C. H. , Yang R. S. ( 2008a ). Element and Sr-Nb-Pb isotope geochemistry of ganite-porphyry dykes in the Yangshan gold belt, western Qinling orogen . Acta Petrologica Sinica 24 , 1101 – 1111 . Liu Y. S. , Zong K. Q. , Kelemen P. B. , Gao S. ( 2008c ). Geochemistry and magmatic history of eclogues and ultramafic rocks from the Chinese continental scientific drill hole: subduction and ultrahigh-pressure metamorphism of lower crustal cumulates . Chemical Geology 247 , 133 – 153 . Google Scholar CrossRef Search ADS Liu Y. S. , Hu Z. , Gao S. , Günther D. , Xu J. , Gao C. , Chen H. ( 2008b ). In situ analysis of major and trace elements of anhydrous minerals by LA-ICP-MS without applying an internal standard . Chemical Geology 257 , 34 – 43 . Google Scholar CrossRef Search ADS Liu Y. S. , Hu Z. C. , Zong K. Q. , Gao C. G. , Gao S. , Xu J. , Chen H. H. ( 2010 ). Reappraisement and refinement of zircon U–Pb isotope and trace element analyses by LA-ICP-MS . Chinese Science Bulletin 55 , 1535 – 1546 . Google Scholar CrossRef Search ADS Luo B. J. , Zhang H. F. , Lu X. B. ( 2012 ). U–Pb zircon dating, geochemical and Sr-Nd-Hf isotopic compositions of Early Indosinian intrusive rocks in West Qinling, central China: petrogenesis and tectonic implications . Contributions to Mineralogy and Petrology 164 , 551 – 569 . Google Scholar CrossRef Search ADS Luo B. J. , Zhang H. F. , Xu W. C. , Guo L. , Pan F. B. , Yang H. ( 2015 ). The Middle Triassic Meiwu Batholith, West Qinling, Central China: implications for the evolution of compositional diversity in a composite Batholith . Journal of Petrology 56 , 1139 – 1172 . Google Scholar CrossRef Search ADS Maniar P. D. , Piccoli P. M. ( 1989 ). Tectonic discrimination of granitoids . Geological Society of America Bulletin 101 , 635 – 643 . Google Scholar CrossRef Search ADS Marsh B. D. ( 2015 ). 6.07—magmatism, magma, and magma chambers . Treatise on Geophysics 6 , 273 – 323 . Google Scholar CrossRef Search ADS Miller J. S. , Matzel J. E. P. , Miller C. F. , Burgess S. D. , Miller R. B. ( 2007 ). Zircon growth and recycling during the assembly of large, composite arc plutons . Journal of Volcanology and Geothermal Research 167 , 282 – 299 . Google Scholar CrossRef Search ADS Montel J. M. , Vielzeuf D. ( 1997 ). Partial melting of metagreywackes, Part II. Compositions of minerals and melts . Contributions to Mineralogy and Petrology 128 , 176 – 196 . Google Scholar CrossRef Search ADS Müntener O. , Kelemen P. B. , Grove T. L. ( 2001 ). The role of H2O during crystallization of primitive arc magmas under uppermost mantle conditions and genesis of igneous pyroxenites: an experimental study . Contributions to Mineralogy & Petrology 141 , 643 – 658 . Google Scholar CrossRef Search ADS Parat F. , Streck M. J. , Holtz F. , Almeev R. ( 2014 ). Experimental study into the petrogenesis of crystal-rich basaltic to andesitic magmas at Arenal volcano . Contributions to Mineralogy and Petrology 168 , 1 – 18 . Google Scholar CrossRef Search ADS Patiño Douce A. E. ( 1999 ). What do experiments tell us about the relative contributions of crust and mantle to the origin of granitic magmas ? Geological Society of London, Special Publications 168 , 55 – 75 . Google Scholar CrossRef Search ADS Patiño Douce A. E. , Harris N. ( 1998 ). Experimental constraints on Himalayan anatexis . Journal of Petrology 39 , 689 – 710 . Google Scholar CrossRef Search ADS Peccerillo A. , Taylor S. R. ( 1976 ). Geochemistry of Eocene calc-alkaline volcanic rocks from the Kastamonu area, northern Turkey . Contributions to Mineralogy and Petrology 58 , 63 – 81 . Google Scholar CrossRef Search ADS Price R. C. , Gamble J. A. , Smith I. E. M. , Maas R. , Waight T. , Stewart R. B. , Woodhead J. ( 2012 ). The anatomy of an andesite volcano: a time-stratigraphic study of andesite petrogenesis and crustal evolution at Ruapehu volcano, New Zealand . Journal of Petrology 53 , 2139 – 2189 . Google Scholar CrossRef Search ADS Putirka K. D. ( 2008a ). Introduction to minerals, inclusions and volcanic processes . Reviews in Mineralogy & Geochemistry 69 , 1 – 8 . Google Scholar CrossRef Search ADS Putirka K. D. ( 2008b ). Thermometers and barometers for volcanic systems . Minerals, Inclusions and Volcanic Processes 69 , 61 – 120 . Qian Q. , Hermann J. ( 2010 ). Formation of high-Mg diorites through assimilation of peridotite by monzodiorite magma at crustal depths . Journal of Petrology 51 , 1381 – 1416 . Google Scholar CrossRef Search ADS Qin J. F. , Lai S. C. , Grapes R. , Diwu C. R. , Ju Y. J. , Li Y. F. ( 2009 ). Geochemical evidence for origin of magma mixing for the Triassic monzonitic granite and its enclaves at Mishuling in the Qinling orogen (central China) . Lithos 112 , 259 – 276 . Google Scholar CrossRef Search ADS Rapp R. P. , Watson E. B. ( 1995 ). Dehydration melting of metabasalt at 8–32 kbar: implications for continental growth and crust-mantle recycling . Journal of Petrology 36 , 891 – 931 . Google Scholar CrossRef Search ADS Reubi O. , Blundy J. ( 2009 ). A dearth of intermediate melts at subduction zone volcanoes and the petrogenesis of arc andesites . Nature 461 , 1269 – 1273 . Google Scholar CrossRef Search ADS PubMed Rollinson H. R. ( 1993 ). Using Geochemical Data: Evaluation, Presentation, Interpretation. Routledge : Longman Scientific & Technical Essex . Shirey S. B. , Hanson G. N. ( 1984 ). Mantle-derived Archaean monozodiorites and trachyandesites . Nature 310 , 222 – 224 . Google Scholar CrossRef Search ADS Sisson T. W. , Grove T. L. ( 1993 ). Experimental investigations of the role of H2o in calc-alkaline differentiation and subduction zone magmatism . Contributions to Mineralogy and Petrology 113 , 143 – 166 . Google Scholar CrossRef Search ADS Sisson T. W. , Ratajeski K. , Hankins W. B. , Glazner A. F. ( 2005 ). Voluminous granitic magmas from common basaltic sources . Contributions to Mineralogy and Petrology 148 , 635 – 661 . Google Scholar CrossRef Search ADS Smith I. E. M. , Stewart R. B. , Price R. C. , Worthington T. J. ( 2010 ). Are arc-type rocks the products of magma crystallisation? Observations from a simple oceanic arc volcano: Raoul Island, Kermadec Arc, SW Pacific . Journal of Volcanology and Geothermal Research 190 , 219 – 234 . Google Scholar CrossRef Search ADS Smithies R. H. , Champion D. C. ( 2000 ). The Archaean high-Mg diorite suite: links to tonalite–trondhjemite–granodiorite magmatism and implications for early Archaean crustal growth . Journal of Petrology 41 , 1653 – 1671 . Google Scholar CrossRef Search ADS Stevens G. , Clemens J. D. , Droop G. T. R. ( 1997 ). Melt production during granulite-facies anatexis: experimental data from ''primitive'' metasedimentary protoliths . Contributions to Mineralogy and Petrology 128 , 352 – 370 . Google Scholar CrossRef Search ADS Stevens G. , Villaros A. , Moyen J. F. ( 2007 ). Selective peritectic garnet entrainment as the origin of geochemical diversity in S-type granites . Geology 35 , 9 – 12 . Google Scholar CrossRef Search ADS Streck M. J. , Leeman W. P. , Chesley J. ( 2007 ). High-magnesian andesite from Mount Shasta: a product of magma mixing and contamination, not a primitive mantle melt . Geology 35 , 351 – 354 . Google Scholar CrossRef Search ADS Sun S. S. , McDonough W. F. ( 1989 ). Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes . Geological Society, London, Special Publications 42 , 313 – 345 . Google Scholar CrossRef Search ADS Tatsumi Y. ( 2006 ). High-Mg andesites in the Setouchi volcanic belt, Southwestern Japan: analogy to archean magmatism and continental crust formation? Annual Review of Earth and Planetary Sciences 34 , 467 – 499 . Google Scholar CrossRef Search ADS Villaros A. , Stevens G. , Buick I. S. ( 2009 ). Tracking S-type granite from source to emplacement: clues from garnet in the Cape Granite Suite . Lithos 112 , 217 – 235 . Google Scholar CrossRef Search ADS Wang S. C. , Ding Y. ( 1990 ). The discovery of garnet in calc-alkaline volcanic rocks of Western Qinling and its geological signifiance . Acta Petrologica et Mineralogica 9 , 13 – 36 . Winchester J. A. , Floyd P. A. ( 1977 ). Geochemical discrimination of different magma series and their differentiation products using immobile elements . Chemical Geology 20 , 325 – 343 . Google Scholar CrossRef Search ADS Xu J. F. , Castillo P. R. , Li X. H. , Yu X. Y. , Zhang B. R. , Han Y. W. ( 2002 ). MORB-type rocks from the Paleo-Tethyan Mian-Lueyang northern ophiolite in the Qinling Mountains, central China: implications for the source of the low 206Pb/204Pb and high 143Nd/144Nd mantle component in the Indian Ocean . Earth and Planetary Science Letters 198 , 323 – 337 . Google Scholar CrossRef Search ADS Xu Q. P. ( 1994 ). Maixiu group of the Western Qinling area . Journal of Stratigraphy 18 , 282 – 288 . Zeck H. P. ( 1992 ). Restite-melt and mafic felsic magma mixing and mingling in an S-type dacite, Cerro-Del-Hoyazo, Southeastern Spain . Transactions of the Royal Society of Edinburgh-Earth Sciences 83 , 139 – 144 . Google Scholar CrossRef Search ADS Zen E. A. ( 1988 ). Phase-relations of peraluminous granitic-rocks and their petrogenetic implications . Annual Review of Earth and Planetary Sciences 16 , 21 – 51 . Google Scholar CrossRef Search ADS Zhang C. , Ma C. , Holtz F. , Koepke J. , Wolff P. E. , Berndt J. ( 2013 ). Mineralogical and geochemical constraints on contribution of magma mixing and fractional crystallization to high-Mg adakite-like diorites in eastern Dabie orogen, East China . Lithos 172-173 , 118 – 138 . Google Scholar CrossRef Search ADS Zhang C. L. , Wang T. , Wang X. X. ( 2008 ). Origin and tectonic setting of the early mesozoic granitoids in Qinling orogenic belt . Geological Journal of China Universities 3 , 304 – 316 . Zhang H. F. , Chen Y. L. , Xu W. C. , Liu R. , Yuan H. L. , Liu X. M. ( 2006 ). Granitoids around Gonghe basin in Qinghai province: petrogenesis and tectonic implications . Acta Petrologica Sinica 22 , 2910 – 2922 . Zhang H. F. , Harris N. , Parrish R. , Kelley S. , Zhang L. , Rogers N. , Argles T. , King J. ( 2004 ). Causes and consequences of protracted melting of the mid-crust exposed in the North Himalayan antiform . Earth and Planetary Science Letters 228 , 195 – 212 . Google Scholar CrossRef Search ADS Zhang H. F. , Jin L. L. , Zhang L. , Harris N. , Zhou L. , Hu S. H. , Zhang B. R. ( 2007 ). Geochemical and Pb-Sr-Nd isotopic compositions of granitoids from western Qinling belt: constraints on basement nature and tectonic affinity . Science in China Series D-Earth Sciences 50 , 184 – 196 . Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Petrology Oxford University Press

The Magmatic Plumbing System for Mesozoic High-Mg Andesites, Garnet-bearing Dacites and Porphyries, Rhyolites and Leucogranites from West Qinling, Central China

Journal of Petrology , Volume Advance Article (3) – Apr 12, 2018

Loading next page...
 
/lp/ou_press/the-magmatic-plumbing-system-for-mesozoic-high-mg-andesites-garnet-r0l0lns8Tl
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0022-3530
eISSN
1460-2415
D.O.I.
10.1093/petrology/egy035
Publisher site
See Article on Publisher Site

Abstract

ABSTRACT An integrated study of the petrography, mineral composition, zircon geochronology, whole-rock geochemistry and Sr–Nd–Hf isotopes was carried out for an unusual suite of igneous rocks, including high-Mg andesites, garnet-bearing dacites and porphyries, rhyolites and leucogranites, from West Qinling, central China. These data, particularly observations from garnets, are used to demonstrate the petrogenetic links among the associated magmatic components which eventually formed the observed lithologies, evaluate the influence of recycling crystal populations and reconstruct the whole magmatic plumbing system. The crystallization ages of these igneous rocks are ∼239–244 Ma. The high-Mg andesites are phenocryst-rich and characterized by high Mg# (>40; Mg#=100*mol. MgO/(MgO + FeO)), Cr and Ni abundances, low Sr/Y and (La/Yb)N ratios, and relatively high ISr and negative εNd(t) and εHf(t) values. The petrography, mineral chemistry and geochemical data indicate that the high-Mg andesites were generated by mixing between mantle-derived magmas and crustal melts, with subsequent entrainment of xenocrysts (e.g. high Mg# pyroxenes, high An plagioclase and some glomerocrysts) from various sources within the crust. The chemical compositions of the garnet-bearing dacites and porphyries and crystal-poor rhyolites define a common differentiation trend. They become more strongly peraluminous and have more evolved Sr–Nd–Hf isotopic compositions with increasing SiO2 content. Petrological and geochemical data indicate that these peraluminous magmas were likely produced by fractional crystallization of andesitic magma, accompanied by assimilation of crustal materials and, or, entrainment of various phenocryst/xenocryst assemblages. Four types of garnets have been identified, including antecrysts, orthocrysts, peritectic phases and xenocrysts, and the variations in mineral composition and inclusion assemblage indicate a complicated history of magma mixing and mineral-melt interaction/re-equilibrium. The leucogranites are strongly depleted in HREE ((La/Yb)N > 300) and show remarkable negative Eu anomalies (Eu/Eu* = 0·42–0·52). These geochemical features are indicative of the presence of both residual garnet and plagioclase in the magma source resulting from muscovite dehydration melting of metapelitic rocks. Together, all these observations consistently reflect magma evolution in several dispersed but interconnected magma reservoirs which formed a complicated trans-crustal magmatic plumbing system. Local magma compositions have been influenced by multiple processes, including crystallization and accumulation, recharging, anatexis, magma mixing and mingling, assimilation, remobilization of crystal mushes and random entrainment of various phenocryst assemblages and crustal xenoliths. Therefore, detailed petrographic information and mineral composition data are needed for interpreting the whole-rock geochemistry properly. The rapid ascent and eruption of crystal-rich and garnet-bearing magmas have been closely associated with an extensional regime in a post-collisional tectonic setting and facilitated by active fault systems. INTRODUCTION The origin of intermediate to silicic magmatic rocks is fundamental for understanding the evolution and differentiation of the continental crust (Annen et al., 2006; Kemp et al., 2007). Geophysical, geochemical, petrological, and volcanological studies have demonstrated that magmatic systems are trans-crustal, dispersed throughout the entire crust, and dominated by crystal mushes (Dahren et al., 2012; Annen et al., 2015; Marsh, 2015; Cashman et al., 2017). Rapidly ascending magmas can contain abundant mush fragments, which can be present as single crystals, crystal clusters (glomerocrysts), cumulate nodules, or restite (Cashman et al., 2017). These crystal populations may have contrasting origins and can be classified as orthocrysts, antecrysts, xenocrysts or peritectic phases (Davidson et al., 2007; Stevens et al., 2007; Jerram & Martin, 2008; Lackey et al., 2012). Orthocrysts are directly crystallized from their host magmas, whereas antecrysts are not in equilibrium with the host magmas but nevertheless related to the magmatic system (Miller et al., 2007; Jerram & Martin, 2008; Bach et al., 2012). Xenocrysts are derived from assimilated xenolithic rocks and peritectic phases form by dehydration melting of metamorphic rocks or xenoliths (Harangi et al., 2001; Stevens et al., 2007; Erdmann et al., 2009). These crystal populations reflect the diverse magma compositions with which they were in equilibrium (Davidson et al., 2007; Jerram & Martin, 2008), and they may provide important insights into the complex magmatic processes and pressure and temperature conditions in which the host magmas have evolved (Jerram & Davidson, 2007; Putirka, 2008a, b). Moreover, the incorporation of complex crystal populations influences final bulk-rock compositions (e.g. Stevens et al., 2007; Smith et al., 2010; Price et al., 2012; Larrea et al., 2013). Thus, evaluating the origin of complex crystal populations in igneous rocks can offer more robust information than can be obtained from whole-rock compositions, revealing the petrogenetic processes and prevailing physicochemical conditions by which the host magmas were generated and evolved, and facilitating the reconstruction of the whole magmatic plumbing system. In this contribution we focus on an unusual suite of garnet-bearing igneous rocks from the West Qinling Orogen, central China (Fig. 1a and b). Garnet is a common mineral in peraluminous dacitic to rhyolitic rocks (e.g. Clemens & Wall, 1984; Gilbert & Rogers, 1989; Zeck, 1992; Cesare, 2000; Acosta-Vigil et al., 2010), but rare in metaluminous andesites worldwide (e.g. Fitton, 1977; Day et al., 1992; Harangi et al., 2001; Kawabata & Takafuji, 2005; Bach et al., 2012). Previous studies have shown that garnets in igneous rocks can occur as orthocrysts, antecrysts, xenocrysts and, or, peritectic phases and usually crystallize at ∼5–12 kbar. Compositional data for garnets may record petrogenetic processes from source to emplacement or eruption (Harangi et al., 2001; Stevens et al., 2007; Acosta-Vigil et al., 2010; Bach et al., 2012; Lackey et al., 2012). The garnet-bearing andesitic rocks in the Hezuo and Maixiu areas from the West Qinling Orogen were first described by Wang & Ding (1990) (Fig. 1c). Recently, Li et al. (2013) suggested that some garnet-free andesites from the Maixiu area are high-Mg andesites. Our observations suggest that garnet-bearing dacites, granodioritic porphyries and pyroclastic rocks are spatially associated with high Mg-andesites, rhyolites and leucogranites (Fig. 1c). We present a comprehensive study of these igneous rocks based on combined petrography and mineral compositions, laser ablation inductively coupled mass spectrometric (LA-ICP-MS) U–Pb zircon dating, whole-rock geochemical and Sr–Nd–Hf isotopic compositions. These rock assemblages provide a rare opportunity to examine the petrogenetic links among these associated igneous rocks, and they provide an opportunity to examine the effects of recycled crystal populations on whole-rock compositions and to reconstruct the magmatic plumbing system in which the various magmas have evolved. Fig. 1 View largeDownload slide (a) Simplified geological map, showing the major tectonic units of China. (b) The distribution of Early Mesozoic magmatic rocks in the Qinling orogen (Feng et al., 2002). (c) Simplified geological map of the studied area. WQ, West Qinling; EQ, East Qinling; DB, Dabie belt; SL, Sulu belt; QD, Qaidam; QL, Qilian belt; KL (EKL), Kunlun belt; NQL, North Qinling; SQL, South Qinling. Pluton names: MB, Miba; MSL; Mishuling; WQ, Wenquan; XH, Xiahe; XK, Xiekeng; SPX, Shuanpengxi; TR, Tongren; JZ, Jianzha; HMH, Heimahe; WQ, Wenquan. Zircon U–Pb ages are indicated in Fig. 1 b (Luo et al., 2015 and references therein). Ages of the Maixiu andesite and Duowa granodiorite in Fig. 1c are from Li et al. (2013, 2015). Fig. 1 View largeDownload slide (a) Simplified geological map, showing the major tectonic units of China. (b) The distribution of Early Mesozoic magmatic rocks in the Qinling orogen (Feng et al., 2002). (c) Simplified geological map of the studied area. WQ, West Qinling; EQ, East Qinling; DB, Dabie belt; SL, Sulu belt; QD, Qaidam; QL, Qilian belt; KL (EKL), Kunlun belt; NQL, North Qinling; SQL, South Qinling. Pluton names: MB, Miba; MSL; Mishuling; WQ, Wenquan; XH, Xiahe; XK, Xiekeng; SPX, Shuanpengxi; TR, Tongren; JZ, Jianzha; HMH, Heimahe; WQ, Wenquan. Zircon U–Pb ages are indicated in Fig. 1 b (Luo et al., 2015 and references therein). Ages of the Maixiu andesite and Duowa granodiorite in Fig. 1c are from Li et al. (2013, 2015). GEOLOGICAL BACKGROUND The West Qinling Orogen is separated from the East Kunlun and Qaidam terranes by the Wenquan-Wahongshang fault to the west, bounded by the Qilian Orogen along the Qinghai Lake-Baoji fault to the north, and separated from the Songpan-Ganze block to the south by the A’nimaque-Mianlue suture zone (Fig. 1). The suture zone contains abundant ophiolite fragments, considered to represent fragments of a Late Palaeozoic Palaeo-Tethys oceanic subducted slab (Xu et al., 2002; Guo et al., 2007). The West Qinling Orogen is primarily covered by Devonian to Cretaceous sedimentary rocks and the Precambrian basement is rarely exposed (Feng et al., 2002; Liu et al., 2008a). Early Mesozoic igneous rocks are widespread in the Qinling Orogen (Fig. 1b). Abundant geochronological and geochemical studies have been performed on the granitoids in East Qinling and its neighbouring areas (Fig. 1b) (Luo et al., 2015, and references therein). Recent studies have revealed that the early Mesozoic igneous rocks in the West Qinling Orogen can be divided into two stages: (1) Middle Triassic (∼246–234 Ma), mainly distributed in the central and west parts of the West Qinling Orogen (Jin et al., 2005; Zhang et al., 2006; Guo et al., 2012; Luo et al., 2012, 2015; Li et al., 2013, 2014); (2) Late Triassic (∼228–205 Ma), widespread throughout the West Qinling Orogen (Fig. 1b) (Zhang et al., 2006, 2007; Qin et al., 2009; Cao et al., 2011). There are two hypotheses for the generation of the Middle Triassic magmatism in West Qinling: (a) an active continental margin setting related to subduction of the A’nimaque oceanic slab (Jin et al., 2005; Guo et al., 2012; Li et al., 2013, 2014); (b) an early-stage post-collisional setting induced by delamination of thickened lithosphere (Zhang et al., 2008) or break-off of the subducted A’nimaque oceanic slab (Zhang et al., 2006; Luo et al., 2012, 2015) after the collision between the West Qinling Orogen and the Songpan-Ganze block. FIELD GEOLOGY AND PETROGRAPHY The Mesozoic intermediate to felsic volcanic rocks are discontinuously distributed in a NW–SE direction over an area 140 km long and 10–16 km wide in the central part of the West Qinling Qrogen, extending from Dewulu, through Saierqingou to Maixiu (Fig. 1c). The garnet-bearing andesitic rocks from Maixiu and Dewulu (Fig. 1c) have been described by Wang & Ding (1990). In this study, the Saierqingou garnet-bearing volcanic rocks, the Fandelongwa garnet-bearing granodioritic porphyries and the Sangke leucogranites are newly recognized rock suites in the Xiahe area (Fig. 1c). Volcanic rocks The Maixiu volcanic rocks The Maixiu volcanic rocks belong to the Maixiu group, which unconformably overlies Lower to Middle Triassic strata and has a total thickness of 2·2–3·2 km and an exposed area of ∼272 km2 (Xu, 1994) (Fig. 1c), The Maixiu group can be divided into three sub-cycles (Xu, 1994). The lower unit (∼2·58 m to 1·12 km thick), located within the southern Maixiu Basin, consists of andesites, tuffaceous and pyroclastic rocks, intercalated with sandstone, slate and/or coal seams at the base overlain by andesites and dacites intercalated with sandstone, slate and pyroclastic rocks (Xu, 1994). The middle unit (> 0·78 km thick), mainly located in the centre of the Maixiu Basin, comprises dacites, dacitic lava breccias and brecciated lavas (Xu, 1994). The upper unit (1·04–1·41 km thick) is widely distributed in the central and northern parts of the Maixiu Basin. The lower part of the upper unit is composed of amphibole-bearing andesites and andesitic pyroclastic rocks, whereas the upper part consists of pyroclastic rocks and clastic sedimentary rocks (Xu, 1994). The andesites described in this study were collected from the lower unit of the Maixiu group. The Maixiu andesites are dark grey with abundant phenocrysts (35–45 vol. %, up to ∼50 vol. %). Petrographically, two distinct groups can be distinguished. One group contains abundant plagioclase (35–40 vol. %), minor clinopyroxene (1–3 vol. %), orthopyroxene (2–5 vol. %) and rare amphibole and biotite, whereas the other group contains less abundant plagioclase (25–35 vol. %) and more abundant clinopyroxene (5–7 vol. %) and orthopyroxene (5–10 vol. %) (Fig. 2a and b). Orthopyroxene is typically 0·2–0·8 mm in size and occurs as subhedral phenocrysts and glomeroporphyritic aggregates (Fig. 2a). Clinopyroxene (0·3–0·7 mm) and plagioclase (0·4–2 mm) are present as euhedral to subhedral phenocrysts. Some plagioclase grains display sieve-textured cores and pristine rims, and they commonly occur as glomeroporphyritic aggregates (Fig. 2b). The groundmass has a hyalopilitic and/or microcrystalline texture with plagioclase as the dominant microphenocryst (Fig. 2a and b). Some samples exhibit minor sericitization, epidotization and chloritization. Fig. 2 View largeDownload slide Representative photomicrographs. (a) and (b) Maixiu (MX) andesite (0929), orthopyroxene and plagioclase glomerocrysts, and plagioclase with sieve-textured core overgrown by a thin rim. (c) Dewulu (DWL) andesite (0993), plagioclase grains show different shapes, compositions and chemical zoning. (d) Saierqingou (SEQG) andesite (09129), orthopyroxene glomerocryst. (e) Saierqingou garnet-bearing dacite (0988), plagioclase grains have different compositions and patchy-zoning. (f) Fangdelongwa (FDLW) garnet-bearing porphyry (0981), plagioclase and quartz grains are zoned. (g) metapelitic enclave in the Fangdelongwa garnet-bearing porphyry (09164). (h) Sangke (SK) leucogranite (09117). (c), (e) and (f) are cathodoluminescence images. Opx, orthopyroxene; Pl, plagioclase; Bio, biotite; Q, quartz; Kf, k-feldspar; Mus, muscovite; Grt, Garnet. Fig. 2 View largeDownload slide Representative photomicrographs. (a) and (b) Maixiu (MX) andesite (0929), orthopyroxene and plagioclase glomerocrysts, and plagioclase with sieve-textured core overgrown by a thin rim. (c) Dewulu (DWL) andesite (0993), plagioclase grains show different shapes, compositions and chemical zoning. (d) Saierqingou (SEQG) andesite (09129), orthopyroxene glomerocryst. (e) Saierqingou garnet-bearing dacite (0988), plagioclase grains have different compositions and patchy-zoning. (f) Fangdelongwa (FDLW) garnet-bearing porphyry (0981), plagioclase and quartz grains are zoned. (g) metapelitic enclave in the Fangdelongwa garnet-bearing porphyry (09164). (h) Sangke (SK) leucogranite (09117). (c), (e) and (f) are cathodoluminescence images. Opx, orthopyroxene; Pl, plagioclase; Bio, biotite; Q, quartz; Kf, k-feldspar; Mus, muscovite; Grt, Garnet. The Dewulu volcanic rocks The Dewulu volcanic rocks, which cover an exposed area of ∼56 km2 and unconformably overlie Lower Permian strata, consist mainly of andesites and dacites with minor pyroclastic rocks and rhyolitic tuffs (Fig. 1c). They are intruded by a porphyritic quartz diorite stock and locally overlain by Jurassic strata. The Dewulu andesites and dacites were sampled in this study. The grey-black andesites are porphyritic, containing abundant plagioclase phenocrysts (35–40 vol. %), and minor pyroxene and amphibole (2–5 vol. %) (Fig. 2c). Plagioclase phenocrysts (0·3–2 mm in length) occur as euhedral long platy crystals and/or irregular porphyroclastic grains, and exhibit variable compositions and complex zoning (e.g. simple, oscillatory, and/or patchy zoning) (Fig. 2c). Small grains of pyroxenes and amphibole (0·2–0·4 mm) have been replaced by chlorite. The matrix has a hyalopilitic texture and consists of microlites of plagioclase, quartz, biotite and opaque minerals (Fig. 2c). The grey dacites display textures similar to those of the andesites, but have different phenocryst assemblages. Their phenocrysts comprise plagioclase (30–35 vol. %), quartz (10–15 vol. %) and minor biotite (2–3 vol. %), with a microlitic matrix consisting of K-feldspar, plagioclase, biotite and magnetite. The Saierqingou volcanic rocks The Saierqingou volcanic rocks, which unconformably overlie Lower Triassic strata, have a total thickness of ∼2·6 km over an area of ∼182 km2 (Fig. 1c). The volcanic succession can be divided into two units from bottom to top. The lower unit is composed of: (1) grey-black porphyritic andesites (∼0·05 km); (2) grey-black, medium-thick bedded siliceous rocks with thin layers of grey-green volcanic breccia (∼0·01 km); (3) yellow-grey rhyolitic tuff and grey-green dacites with minor andesitic breccia (∼0·03 km); and (4) grey-green dacitic breccia and purple red rhyolite. The upper unit consists mainly of purple red rhyolite and porphyries (∼2·05 km), with pyroclastic rocks at the top (∼0·35 km). The porphyritic andesites, garnet-bearing dacites and pyroclastic rocks, and rhyolite were sampled in this study. The Saierqingou grey-black andesites have a porphyritic texture with ∼40 vol. % phenocrysts. Similar to the Maixiu andesites, two groups of andesites can be identified. One group is characterized by abundant plagioclase (30–35 vol. %) and orthopyroxene (5–10 vol. %) (Fig. 2d). The other group is composed mainly of plagioclase (40–45 vol. %) with minor pyroxene, amphibole and biotite (2–3 vol. %). Orthopyroxene grains are generally < 0·5 mm in diameter and some glomeroporphyritic aggregates are also observed (Fig. 2d). Plagioclase (0·3–3 mm in length) mainly forms subhedral to euhedral and lath-shaped crystals, and shows albite twinning or oscillatory zoning (Fig. 2d). The matrix comprises plagioclase and biotite with minor zircon, apatite, and magnetite (Fig. 2d). The grey-green porphyritic garnet-bearing dacites contain ∼40 vol. % of phenocrysts (plagioclase 20–25 vol. %, quartz 10–15 vol. %, biotite 2–5 vol. %, garnet ∼1%) (Fig. 2e). The matrix is composed of plagioclase, K-feldspar and quartz, and minor zircon, apatite, and magnetite. Plagioclase grains (0·2–2 mm in length) show diverse types and some grains have patchy zoning (Fig. 2e). Quartz and biotite are anhedral and have irregular shapes with a diameter of 0·1–0·3 mm. The garnet-bearing pyroclastic rocks are grey-black and grey-green. Their mineral assemblage is comparable to those of the andesites and dacites. The purple red rhyodacites and rhyolites display aphanitic textures and consist mainly of quartz, plagioclase, albite, K-feldspar and biotite, and minor zircon, apatite and magnetite. Hypabyssal intrusive rocks The Fandelongwa garnet-bearing granodioritic porphyry The Fandelongwa garnet-bearing granodioritic porphyry intrudes Lower Triassic strata as a small stock (∼5 km2), which is exposed to the west of Xiahe town (Fig. 1c). The phenocrysts are plagioclase (15–20 vol. %), albite (5–10%), quartz (5–10 vol. %), biotite (7–10 vol. %) and minor garnet (∼1 vol. %) (Fig. 2f), with accessory magnetite, titanite, apatite and zircon. Both plagioclase and quartz have larger sizes (0·5–4 mm) than those in the garnet-bearing dacites. Some plagioclase and quartz grains have been partially resorbed into round or elliptical shapes and have clear core–rim textures (Fig. 2f). Some plagioclase grains also occur in glomeroporphyritic aggregates. Biotite (0·1–0·2 mm) is subhedral to anhedral. Enclaves of wall-rocks and fine-grained diorite and biotite-rich metapelite (Fig. 2g) are locally preserved. They are generally ovoid or ellipsoidal and commonly 2–20 centimetres in size. The Sangke two-mica porphyritic leucogranite The Sangke two-mica, porphyritic leucogranite stock (∼3 km2) crops out to the south of Xiahe town (Fig. 1c). This was emplaced into the Lower Triassic Longwuhe Formation and is covered to the west by Quaternary sediments. K-feldspar (10 vol. %) phenocrysts are generally 0.–2·0 mm in size. The matrix is very fine-grained (0·1–0·2 mm) and composed of quartz (30–35 vol. %), K-feldspar (20–30 vol. %), oligoclase (15–25 vol. %), muscovite (2–3 vol. %), and biotite (3–4 vol. %), with minor zircon, apatite and magnetite (Fig. 2h). Petrography and classification of garnets Different types of garnet are present in the Saierqingou dacites and pyroclastic rocks (Fig. 3), the Fandelongwa porphyries (Fig. 4), as well as in the rare metapelitic xenoliths (Fig. 4) in the Fandelongwa porphyries. According to crystal form, mineral inclusions and zoning pattern, garnets can be classified into the following four types. Fig. 3 View largeDownload slide Garnet in the Saierqingou volcanic rocks. (a) and (b) Type-1 garnets in dacite. (a) resorbed, rounded garnet 0988#1 is surrounded by a plagioclase corona; (b) garnet 0988#2 has a dark core and a pale rim; (c) Type-1 garnet (0984#1) from pyroclastic rock. (d)–(h) Type-2 garnets from dacite (09131). Grains #0 and #1 contain mineral inclusions, and #0 is also surrounded by a cordierite corona (d)–(e). Grains of #2–4 are rounded and cracked and are enclosed by plagioclase (f)–(h). (i) Type-4 garnet (09131#5) is an irregular fragment. Yellow circles represent spots analyzed by EMPA. Zrn, zircon; Ap, apatite; Ru, rutile; Crd, cordierite; MI, melt inclusions; FI, fluid inclusions. Fig. 3 View largeDownload slide Garnet in the Saierqingou volcanic rocks. (a) and (b) Type-1 garnets in dacite. (a) resorbed, rounded garnet 0988#1 is surrounded by a plagioclase corona; (b) garnet 0988#2 has a dark core and a pale rim; (c) Type-1 garnet (0984#1) from pyroclastic rock. (d)–(h) Type-2 garnets from dacite (09131). Grains #0 and #1 contain mineral inclusions, and #0 is also surrounded by a cordierite corona (d)–(e). Grains of #2–4 are rounded and cracked and are enclosed by plagioclase (f)–(h). (i) Type-4 garnet (09131#5) is an irregular fragment. Yellow circles represent spots analyzed by EMPA. Zrn, zircon; Ap, apatite; Ru, rutile; Crd, cordierite; MI, melt inclusions; FI, fluid inclusions. Fig. 4 View largeDownload slide View largeDownload slide Garnet in the Fangdelongwa granodioritic porphyry. Type-1 garnets: (a) resorbed garnet 0981#1; (b) garnet 0981#2 has a dark red core and a light-coloured rim. Type-4 garnets: (c) and (d) garnets 09163#1 and 09163#2 are surrounded by reaction coronae of albite and biotite, and 09163#1 has a dusty core and a clear rim; (e) and (f) garnet 09164#4 has been extensively replaced by biotite and chlorite. (g)–(k) Type-3 garnets occur as clusters in a metapelitic xenolith. (g) metapelitic xenolith in the porphyry is surrounded by glomerocrysts of plagioclase. (h)–(k) garnets (09164#1–3) have dusty cores and clear rims. Yellow circles represent spots analyzed by EMPA and red circles represent spots analyzed by LA-ICP-MS. Pl, plagioclase; Bio, biotite; Chl,chlorite; Sil, sillimanite; Ab,albite. Fig. 4 View largeDownload slide View largeDownload slide Garnet in the Fangdelongwa granodioritic porphyry. Type-1 garnets: (a) resorbed garnet 0981#1; (b) garnet 0981#2 has a dark red core and a light-coloured rim. Type-4 garnets: (c) and (d) garnets 09163#1 and 09163#2 are surrounded by reaction coronae of albite and biotite, and 09163#1 has a dusty core and a clear rim; (e) and (f) garnet 09164#4 has been extensively replaced by biotite and chlorite. (g)–(k) Type-3 garnets occur as clusters in a metapelitic xenolith. (g) metapelitic xenolith in the porphyry is surrounded by glomerocrysts of plagioclase. (h)–(k) garnets (09164#1–3) have dusty cores and clear rims. Yellow circles represent spots analyzed by EMPA and red circles represent spots analyzed by LA-ICP-MS. Pl, plagioclase; Bio, biotite; Chl,chlorite; Sil, sillimanite; Ab,albite. Type-1 garnets, which are subhedral to euhedral and usually 0·3–5 mm in diameter, occur as disseminated crystals in the Saierqingou volcanic rocks (Fig. 3) and the Fandelongwa porphyries (Fig. 4a and b). They are characterized by abundant inclusions (e.g. ilmenite, zircon and apatite), zoning of colour and inclusions and some reaction and resorption textures. A rounded and resorbed garnet 0988#1 is surrounded by a reaction corona consisting of plagioclase (Fig. 3a). Garnet 0988#2 has a dark red core containing inclusions of zircon and apatite and a light red rim with glass inclusions (Fig. 3b). Garnets 0984#1 and 0981#1 have also been resorbed and have dentated margins (Figs 3c and 4a). Garnet 0981#2 is intergrown with plagioclase and biotite and has a dark-red core and a light-colored rim (Fig. 4b). Type-2 garnets, mainly found in the Saierqingou dacites, are anhedral to subhedral, and mostly 0·1–1 mm in diameter and are lighter in colour than Type-1 garnets (Fig. 3d–h). Their cross sections are pristine and inclusion poor with only the occasional zircon, apatite or rutile inclusion (Fig. 3d and e). Most of the resorbed Type-2 garnets are spherical or roughly spherical, and some grains occur as inclusions in plagioclase (Fig. 3f–h). Garnet 09131#0 is surrounded by a reaction corona of cordierite (Fig. 3d). Type-3 garnets are anhedral to subhedral, 0·3 to 3 mm in diameter and occur as clusters