Hostname: page-component-74656f9884-skzmx Total loading time: 0 Render date: 2026-11-06T11:23:36.973Z Has data issue: false hasContentIssue false
  • English
  • Français

Origin of Devonian mafic magmatism in the East Kunlun orogenic belt, northern Tibetan Plateau: implications for continental exhumation

Published online by Cambridge University Press:  09 January 2020

You-Jun Tang ,
Bin Liu Open the ORCID record for Bin Liu [Opens in a new window] ,
Mei-Jun Li ,
Yue Wu ,
Jian Huang ,
Shao-Qing Zhao  and
Yang Sun
You-Jun Tang
Affiliation:
Key Laboratory of Exploration Technologies for Oil and Gas Resources, Ministry of Education, Yangtze University, Wuhan430100, China College of Resources and Environment, Yangtze University, Wuhan430100, China
Bin Liu*
Affiliation:
Key Laboratory of Exploration Technologies for Oil and Gas Resources, Ministry of Education, Yangtze University, Wuhan430100, China State Key Laboratory of Petroleum Resources and Prospecting, China University of Petroleum, Beijing102249, China School of Geosciences, Yangtze University, Wuhan430100, China
Mei-Jun Li
Affiliation:
State Key Laboratory of Petroleum Resources and Prospecting, China University of Petroleum, Beijing102249, China
Yue Wu
Affiliation:
Key Laboratory of Exploration Technologies for Oil and Gas Resources, Ministry of Education, Yangtze University, Wuhan430100, China College of Resources and Environment, Yangtze University, Wuhan430100, China
Jian Huang
Affiliation:
State Key Laboratory of Geological Processes and Mineral Resources, China University of Geosciences, Wuhan430074, China
Shao-Qing Zhao
Affiliation:
Key Laboratory of Exploration Technologies for Oil and Gas Resources, Ministry of Education, Yangtze University, Wuhan430100, China School of Geosciences, Yangtze University, Wuhan430100, China
Yang Sun
Affiliation:
Key Laboratory of Exploration Technologies for Oil and Gas Resources, Ministry of Education, Yangtze University, Wuhan430100, China School of Geosciences, Yangtze University, Wuhan430100, China
*
Author for correspondence: Bin Liu, Email: binliu@yangtzeu.edu.cn;liubincug@foxmail.com
Rights & Permissions [Opens in a new window]

Abstract

This paper presents a comprehensive study of the zircon geochronology, geochemistry and Sr–Nd isotope geology of Devonian mafic rocks developed in the East Kunlun orogenic belt, northern Tibetan Plateau, and reveals their mantle sources, petrogenesis and geodynamic implications for continental exhumation. The zircon geochronology of typical samples indicates that these mafic rocks crystallized at 406∼408 Ma. They can be classified into two different groups based on petrographic observations and geochemical compositions. Group 1 rocks exhibit low TiO2 and FeOt contents and Nb/Y ratios and have enriched mid-ocean ridge basalt (E-MORB)-like compositions with slight negative Nb and Ta anomalies. However, Group 2 rocks have distinctly high TiO2 and FeOt contents and Nb/Y ratios, comparable to typical Fe–Ti-rich mafic rocks worldwide. All the samples exhibit weak enrichments in light rare earth elements, Nb and Ta relative to the primitive mantle. Based on geochemical and isotopic studies, Group 1 rocks are suggested to be derived from depleted asthenospheric mantle that was metasomatized by c. 3–5 % continental crustal components, while Group 2 rocks originated from partial melting of enriched lithospheric mantle. The high contents of Fe, Ti and Nb for Group 2 rocks could be attributed to a high degree of olivine crystallization under low fO2 conditions with delayed nucleation of Fe–Ti oxides. Combining those results with other geological data, we conclude that slab break-off was the key factor causing exhumation of eclogites and triggering flare-up of the Devonian magmatism, and that continental collision or continental subduction may have initiated at 431∼436 Ma.

Keywords

mafic magmatismcontinental exhumationmagmatic flare-upDevonianEastern Kunlun

Information

Type
Original Article
Information
Geological Magazine , Volume 157 , Issue 8 , August 2020 , pp. 1265 - 1280
Copyright
© Cambridge University Press 2020

1. Introduction

Slab subduction on the Earth is generally regarded as an important mechanism for transporting crustal materials into the mantle, triggering extensive crust–mantle interactions and causing chemical heterogeneity in the mantle (e.g. Hofmann, Reference Hofmann1997). Oceanic subduction, accompanied by melting of the subduction-modified mantle wedge and/or a young subducting oceanic crust and melt–peridotite interactions, produces various types of arc magmatic rocks (such as the typical calc-alkaline rock association, adakites and high-Mg andesites; Defant & Drummond, Reference Defant and Drummond1990; Tatsumi & Eggins, Reference Tatsumi and Eggins1995; Spandler & Pirard, Reference Spandler and Pirard2013). In the past 35 years, many studies have discovered that continental crust can also be subducted into the deep mantle, even to depths of >100 km (e.g. Smith, Reference Smith1984; Sobolev & Shatsky, Reference Sobolev and Shatsky1990) where the crust would experience high-pressure (HP) to ultra-high-pressure (UHP) metamorphism and then return to crustal levels by continental exhumation (Zhao et al. Reference Zhao, Dai and Zheng2013). Compared with oceanic subduction, continental subduction rarely develops syn-subduction arc magmatic rocks owing to the relatively low contents of aqueous fluids in the subducting continental crust (e.g. Zheng, Reference Zheng2012). However, syn-exhumation or post-collisional magmatic rocks have been discovered in many well-known continental collision-related orogenic belts, such as the Alpine–Himalaya, Pan-African, Dabie-Sulu and Eastern Kunlun orogens (Bonin, Reference Bonin2004; Chung et al. Reference Chung, Chu, Zhang, Xie, Lo, Lee, Lan, Li, Zhang and Wang2005; Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012; Zhao et al. Reference Zhao, Dai and Zheng2013). Such magmatic rocks can record critical information about the geodynamic processes of continental subduction and exhumation as well as the effects of deep crust–mantle interactions.

The East Kunlun orogenic belt (EKOB) is one of the very large magmatic belts in the northern part of the Tibetan Plateau (Harris et al. Reference Harris, Xu, Lewis and Jin1988; Mo et al. Reference Mo, Luo, Deng and Yu2007; Dong et al. Reference Dong, He, Sun, Liu, Zhou, Zhang, Yang, Cheng, Zhao and Li2018) and is mainly composed of early Palaeozoic and late Permian to Triassic granitic intrusions (Fig. 1; Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012; Xiong et al. Reference Xiong, Ma, Zhang, Liu and Jiang2014). In recent years, voluminous eclogites have been identified in the Precambrian basement rocks of the EKOB, which constitutes one of the largest HP metamorphic belts (∼500 km long) on the Earth (Fig. 1a; Bi et al. Reference Bi, Song, Dong, Yang, Qi and Allen2018; Song et al. Reference Song, Bi, Qi, Yang, Allen, Niu, Su and Li2018). The protoliths of the eclogites are dominated by Proterozoic continental basaltic rocks (∼934 Ma; Meng et al. Reference Meng, Zhang and Cui2013), as evidenced by their continental tholeiite basalt-like geochemical composition, significant crustal contamination and abundant old inherited zircons (Meng et al. Reference Meng, Zhang and Cui2013, Reference Meng, Cui, Jia, Ren and Feng2015). U–Pb dating of metamorphic zircons from the eclogites and their surrounding rocks (e.g. garnet-mica schists) has shown that the timing of eclogite-facies metamorphism is 425∼436 Ma (Meng et al. Reference Meng, Zhang and Cui2013; Zhang et al. Reference Zhang, Qian, Li, Wang, Zhang, You and Liu2017; Song et al. Reference Song, Bi, Qi, Yang, Allen, Niu, Su and Li2018) and the timing of exhumation of these HP rocks is c. 409∼411 Ma (Zhang et al. Reference Zhang, Qian, Li, Wang, Zhang, You and Liu2017; Song et al. Reference Song, Bi, Qi, Yang, Allen, Niu, Su and Li2018). The discovery of coesite inclusions in these eclogites provides robust evidence that the continental crust was subducted into the deep mantle to depths of 100–120 km and then exhumed to crustal levels (Bi et al. Reference Bi, Song, Dong, Yang, Qi and Allen2018; Song et al. Reference Song, Bi, Qi, Yang, Allen, Niu, Su and Li2018). The widespread Devonian magmatic rocks are generally considered to be products of Proto-Tethyan collision-related magmatism in the EKOB (e.g. Zhao et al. Reference Zhao, Ma, Wang, Bai, Li and Ji2008; Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012; Dong et al. Reference Dong, He, Sun, Liu, Zhou, Zhang, Yang, Cheng, Zhao and Li2018). They consist of abundant granitic intrusions and scarce mafic rocks, which mostly intrude the Precambrian basement rocks (Figs 1, 2). Moreover, their zircon U–Pb ages range from 411 Ma to 402 Ma (Zhang et al. Reference Zhang, Meng, Wan, Yang and Dong2003; Long et al. Reference Long, Jin, Ge and Yu2006; Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012; Wang et al. Reference Wang, Dong and Dong2018), which is identical within error to the timing of exhumation of the eclogites. Compared with the granitic intrusions, the genetic mechanism, magma source and geodynamics of these mafic rocks are still poorly understood. Thus, investigations of these Devonian magmatic rocks could provide excellent insights into the process of continental exhumation and its related deep crust–mantle interactions. This paper presents a comprehensive study of the zircon geochronology, geochemistry and Sr–Nd isotope geology of the Devonian mafic rocks in the EKOB. The data are used to ascertain the composition of the mantle sources and petrogenesis and then to reveal their geodynamic relationship with continental exhumation.

Fig. 1.

(a) Simplified geological map showing the distribution of magmatic rocks in the EKOB; and (b) the location of the EKOB in the Tibetan Plateau.

Fig. 2.

Simplified geological map of the study area.

2. Geological background and sample descriptions

The EKOB is situated in the northern part of the Tibetan Plateau and is bounded by the Qaidam basin and Bayan Har – Songpan-Garzê terrane (Fig. 1). The Central Kunlun fault zone (CKLF) divides the EKOB into two segments: the north terrane of the East Kunlun (NEKL) and the south terrane of the East Kunlun (SEKL). The NEKL is mainly composed of Precambrian metamorphic basement rocks (namely, the ‘Jinshuikou Group’) and Silurian to Triassic magmatic rocks. The Jinshuikou Group can be divided into two parts (Wang et al. Reference Wang, Wei, Jia, Zhang, Li, Zhu and Xiang2007; Chen et al. Reference Chen, Pei, Li, Liu, Li, Zhang, Chen, Liu, Ding and Guo2011). The lower part (namely, the ‘Baishahe Formation’) is dominated by marbles, gneisses, migmatites and amphibolites and is considered to have formed during Palaeoproterozoic times (c. 1.9∼2.0 Ga; Wang et al. Reference Wang, Wei, Jia, Zhang, Li, Zhu and Xiang2007), while the upper part (namely, the ‘Xiaomiao Formation’) consists of quartzites, marbles, gneisses and schists with formation ages of 1.6∼1.7 Ga (Chen et al. Reference Chen, Pei, Li, Liu, Li, Zhang, Chen, Liu, Ding and Guo2011). The SEKL is mainly composed of Proterozoic metamorphic rocks (namely, ‘the Kuhai Group and the Wanbaogou Group’), Triassic flysch and some granitic rocks. The CKLF comprises ophiolites, volcanic rocks, mafic intrusions and some metamorphic rocks. The recent identification of eclogites or coesite-bearing eclogites suggests that the CKLF also marks an early Palaeozoic HP metamorphic belt (Meng et al. Reference Meng, Zhang and Cui2013; Bi et al. Reference Bi, Song, Dong, Yang, Qi and Allen2018; Song et al. Reference Song, Bi, Qi, Yang, Allen, Niu, Su and Li2018). The formation ages of the ophiolites are 518∼522 Ma (e.g. Yang et al. Reference Yang, Robinson, Jiang and Xu1996; Lu et al. Reference Lu, Yu, Jin, Li and Zheng2002), suggesting the presence of a Proto-Tethyan Ocean in the EKOB. Previous studies on the Qingshuiquan granulites in the CKLF suggested that the subducted oceanic slab arrived at depths of 40∼50 km before ∼508 Ma (e.g. Li et al. Reference Li, Lu, Xiang, Zhou, Guo, Song, Zheng and Gu2006). The development of calc-alkaline magmatic intrusions with zircon U–Pb ages of 438∼515 Ma (Mo et al. Reference Mo, Luo, Deng and Yu2007; Zhang et al. Reference Zhang, Pei, Ding, Li, Feng, Sun, Li and Chen2010, Reference Zhang, Ma, Xiong, Liu, Li and Pan2014; Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012, Reference Liu, Ma, Jiang, Guo, Zhang and Xiong2013b; Xiong et al. Reference Xiong, Ma, Jiang, Liu, Zhang and Zhou2013; Zhou et al. Reference Zhou, Dong, Zhang, Yang, Sun and He2016; Dong et al. Reference Dong, He, Sun, Liu, Zhou, Zhang, Yang, Cheng, Zhao and Li2018) suggests prolonged subduction of the Proto-Tethyan Ocean during early Cambrian to early Silurian times. The widespread Devonian magmatic rocks are generally considered to have been formed in a collision-related environment (Zhao et al. Reference Zhao, Ma, Wang, Bai, Li and Ji2008; Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012; Dong et al. Reference Dong, He, Sun, Liu, Zhou, Zhang, Yang, Cheng, Zhao and Li2018). They mostly intrude the Precambrian basement rocks and consist of numerous granitic intrusions and minor mafic rocks (Figs 1, 2). The granitic intrusions account for a large proportion (∼60 %) of all Palaeozoic granites in the EKOB and display a narrow range of crystallization ages (402∼411 Ma; Fig. 1a, 2; Zhang et al. Reference Zhang, Meng, Wan, Yang and Dong2003; Long et al. Reference Long, Jin, Ge and Yu2006; Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012; Wang et al. Reference Wang, Dong and Dong2018). Many studies have addressed the idea that partial melting of Precambrian basement rocks and associated magma mixing had played essential roles in the generation of these granitic intrusions (Zhao et al. Reference Zhao, Ma, Wang, Bai, Li and Ji2008; Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012).

In this study, we have identified two different groups of mafic rocks in the NEKL (Fig. 2). Both of them intrude a typical Devonian granitic intrusion with zircon U–Pb ages of ∼407 Ma (Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012). The host granitic intrusion is mainly composed of monzonitic granite and granodiorite (Fig. 2), which are peraluminous granites in composition. The two groups of mafic rocks crop out differently in the field (Figs 2, 3), and there are no cross-cutting relationships between them. Group 1 rocks occur as dykes with a NE–SW-trending strike (Fig. 3a). We have investigated more than five dykes, and the dykes usually have thicknesses ranging from 1.0 m to 1.6 m. All the dykes show sharp contacts with the host granitic intrusion (Fig. 3a), and most of them have fine-grained textures (Fig. 3a, c). The Group 2 mafic rocks appear as a small pluton with a distribution area of ∼0.1 km2 (Figs 2, 3b). The mafic pluton also has a sharp contact with the enclosing granitic rocks, and xenoliths from the host granites are not found in the pluton. Most of the Group 2 rocks are medium to coarse grained, although some rocks near the contact zone exhibit relatively fine-grained textures.

Fig. 3.

Field photographs and photomicrographs for the two groups of Devonian mafic rocks. Cpx – clinopyroxene; Hb – hornblende; Opa – opaque minerals; Pl – plagioclase. Hammer for scale is ∼35 cm long.

Group 1 rocks are mainly composed of plagioclase (40∼45 %), clinopyroxene (15∼20 %) and hornblende (25∼30 %; Fig. 3c). Plagioclases have subhedral crystals and mostly display polysynthetic twinning. Clinopyroxenes are subhedral and partly altered along the margins and cleavage of the grains. Hornblendes are subhedral to anhedral and exhibit obvious pleochroism under plane-polarized light. Group 2 rocks consist of plagioclase (45∼50 %), clinopyroxene (5–10 %) and hornblende (30∼35 %) with some opaque minerals (5∼10 %; Fig. 3d). Plagioclases are subhedral, and they exhibit apparent polysynthetic twinning. Most of the clinopyroxenes have been partly or entirely replaced by hornblendes, uralites, chlorites and some opaque minerals. Opaque minerals show anhedral columnar crystals and occur in the interstices between the clinopyroxenes and plagioclases (Fig. 3d), suggesting they may have crystallized in the late stage of magma evolution.

3. Analytical methods

The samples for the experimental analysis of the chemical and isotopic compositions were selected from fresh outcrops, and detailed analytical methods are described as follows.

To obtain sufficient numbers of magmatic zircons, zircon grains were isolated from ∼22 kg of Group 1 mafic rocks (NM16-1) by using mechanical crushing, magnetic techniques and handpicking under microscopes. The appropriate analytical sites for the zircon grains were selected based on systematic petrographic observations and analyses of cathodoluminescence (CL) images. Analysis of U–Pb isotopes and trace-element compositions of the zircon grains were conducted at the State Key Laboratory of Geological Processes and Mineral Resources (GPMR), China University of Geosciences (Wuhan), using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS). A GeoLas 2005 instrument and an ICP-MS instrument (Agilent 7500a) were used for laser sampling and obtaining ion signal intensities, respectively. Harvard zircon 91500 was used as an external standard to normalize isotopic fractionation during analysis, while NIST SRM 610 glass was used as an external standard and Zr as an internal standard to calculate the trace-element concentrations of unknowns. The detailed operating conditions of these instruments, analytical procedures and calculation of off-line data were the same as those described by Liu et al. (Reference Liu, Hu, Gao, Günther, Xu, Gao and Chen2008a, Reference Liu, Gao, Hu, Gao, Zong and Wang2010a,b).

Analysis of whole-rock major elements was carried out at the Hubei Province Geological Experimental Testing Centre, China. The contents of major oxides were determined using X-ray fluorescence (XRF) spectrometry (Magix_pro2440), while the values of loss on ignition (LOI) were measured by the weight loss after heating at high temperature. The relative standard deviations (RSDs) for major elements were better than 1 %. Trace-element contents were measured using an ICP-MS instrument (Agilent 7500a) at the GPMR. HF+HNO3 in Teflon bombs was used to digest the powder samples. Detailed methods of sample digestion and pre-treatment as well as RSDs for trace elements are described by Liu et al. (Reference Liu, Zong, Kelemen and Gao2008b). The analytical results for the international reference materials and replicate samples are listed in online Supplementary Material Table S1.

Sr–Nd isotope ratios of the mafic rocks were measured at the GPMR. Ratios of 87Sr/86Sr were measured using a Finnigan Triton thermo-ion mass isotope spectrometer, while ratios of 143Nd/144Nd were determined using a multiple collector (MC)-ICP-MS instrument. The detailed sample preparation and experimental procedures were the same as those described by Gao et al. (Reference Gao, Rudnick, Yuan, Liu, Liu, Xu, Ling, Ayers, Wang and Wang2004). The mass fractionation correction for Sr–Nd isotope compositions was carried out using 88Sr/86Sr = 8.375209 and 146Nd/144Nd = 0.721900. Replicate analyses of the NBS 987 standard gave a 88Sr/86Sr ratio of 0.710274 ± 0.000009, showing good agreement with approved published data (0.710241 ± 0.000011; Fantle & DePaolo, Reference Fantle and DePaolo2006). Replicate analyses of the JNdi-1 standard yielded a 146Nd/144Nd ratio of 0.512118 ± 0.000009, also displaying good agreement with approved published data (0.512115 ± 0.000007; Tanaka et al. Reference Tanaka, Togashi, Kamioka, Amakawa, Kagami, Hamamoto, Yuhara, Orihashi, Yoneda, Shimizu, Kunimaru, Takahashi, Yanagi, Nakano, Fujimaki, Shinjo, Asahara, Tanimizu and Dragusanu2000). In addition, replicate analyses of two samples (YJ03-4 and YJ05-1) are shown in Table 3.

4. Results

4.a. Zircon U–Pb ages

The zircon U–Pb isotope analyses for the Group 2 mafic rocks were reported in Liu et al. (Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012) by the authors, and the results yielded a zircon U–Pb age of 406 ± 3 Ma (MSWD = 1.8). Therefore, zircon U–Pb isotope analysis of the Group 1 mafic rocks was performed in this study. The representative sample (NM16-1) was selected from the site at 36° 13.212′ N, 96° 29.003′ E. The results and CL images of the zircon grains are presented in Table 1 and Figure 4, and trace-element compositions of the zircon grains are listed in online Supplementary Material Table S2.

Table 1.

Zircon LA-ICP-MS U–Pb dating results of Devonian mafic rocks

Fig. 4.

(a) CL images and (b) LA-ICP-MS U–Pb concordant diagrams for zircons from the Devonian mafic rocks. Data for chondrite are from Taylor & McLennan (Reference Taylor and McLennan1985). CL images of typical magmatic zircons from the host granitic intrusion (after Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012) are also shown for comparison.

All the zircon grains are colourless and transparent, and most of them show subhedral to columnar shapes with lengths of ∼50–160 μm (Fig. 4a). LA-ICP-MS analyses of U–Pb isotope compositions were carried out on 24 grains. Eight of them give relatively old 206Pb–238U ages, including 2510 Ma and 431–440 Ma (Fig. 4b), which could be interpreted as the formation ages of inherited zircons. The other 16 grains have relatively high Th/U ratios and Sm contents, positive Ce anomalies and obvious enrichment of heavy rare earth elements (HREEs; Fig. 5a; online Supplementary Material Table S2), comparable to those of typical magmatic zircons (Hoskin & Schaltegger, Reference Hoskin and Schaltegger2003). These grains show broadly spaced oscillatory zoning or weak zoning without complex internal structures (e.g. inherited cores) in the CL images (Fig. 4a), which is consistent with the features of typical zircons crystallized from mafic melts (Corfu et al. Reference Corfu, Hanchar, Hoskin and Kinny2003; Wang et al. Reference Wang, Wu, Qin, Zhu, Liu, Liu, Gao, Wijbrans, Zhou, Gong and Yuan2013). Furthermore, zircon grains from the surrounding Devonian granitic intrusion described by Liu et al. (Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012) exhibit entirely different internal structures (such as narrow oscillatory zoning and distinct old inherited cores; Fig. 4a), and most of them have relatively lower Eu/Eu* ratios (Fig. 5b; online Supplementary Material Table S3). These features strongly suggest that the 16 grains may be crystallized from the Group 1 mafic melts rather than inherited from the host granitic intrusion. Sixteen analyses on the zircon grains yielded relatively young 206Pb–238U ages of 404∼411 Ma (Fig. 4b) with a weighted mean 206Pb–238U age of 408 ± 2 Ma (MSWD = 0.25, n = 16), representing the crystallization age of the Group 1 rocks. In addition, the mean 206Pb–238U age is similar to the crystallization age of the Group 2 mafic rocks (Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012), suggesting that both groups of mafic rocks from the EKOB formed contemporaneously.

Fig. 5.

(a) Chondrite-normalized REE patterns and (b) Eu/Eu* v. Sm plot for zircons from the Devonian mafic rocks. Trace-element compositions of typical magmatic zircons from the host granitic intrusion (after Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012) are also shown for comparison.

4.b. Whole-rock geochemistry

The whole-rock chemical compositions of the representative samples are listed in Table 2. All of the Group 1 rocks have moderate SiO2, low FeOt and TiO2 contents, and moderate MgO contents and Mg no. values (Table 2). They have relatively low Nb/Y ratios of 0.13∼0.33 and plot in the field of sub-alkaline basalt on the SiO2 versus Nb/Y diagram (Fig. 6a). The samples exhibit weak enrichment in light rare earth elements (LREEs) relative to HREEs, with LREE/HREE values of 1.70∼3.28 (Fig. 7a). Most of the samples show slight negative Eu anomalies with Eu/Eu* values of 0.89∼1.03. The primitive mantle-normalized trace-element patterns display moderate enrichments in Th and LREEs with slight depletions in Nb and Ta; these characteristics are similar to those of enriched mid-ocean ridge basalt (E-MORB) (Fig. 7b). Compared with Group 1 rocks, Group 2 rocks are distinguished by relatively higher contents of TiO2 and FeOt and low MgO and Mg nos. (Table 2). Most of the samples have relatively high Nb/Y ratios of 0.45∼0.80 and belong to the alkaline basalt series in composition (Fig. 6a). All of the Group 2 rocks show strong enrichments in Fe and Ti, comparable to those of typical Fe–Ti-rich mafic rocks worldwide (Fig. 6b; e.g. Brooks et al. Reference Brooks, Larsen and Nielsen1991; Zhou et al. Reference Zhou, Robinson, Lesher, Keays, Zhang and Malpas2005). All the samples display small enrichments in LREEs relative to HREEs (LREE/HREE = 2.65∼3.48) and weak positive Eu anomalies (Eu/Eu* = 1.17∼1.69) in the chondrite-normalized REE patterns (Fig. 7c). Compared with the primitive mantle, all of the Group 2 rocks exhibit apparent enrichments in Th, LREEs, Nb and Ta, with depletions in Zr, Hf and P (Fig. 7d).

Table 2.

Major- (in wt %) and trace-element (in ppm) compositions of Devonian mafic rocks

FeOt = 0.89981*Fe2O3 + FeO; Mg no. = 100*Mg2+/(Mg2+ + Fet2+); - = undetected.

Table 3.

Sr–Nd isotopic compositions of the selected samples from the Devonian mafic rocks

* Replicate analyses at the State Key Laboratory of Geological Processes and Mineral Resources (GPMR), China University of Geosciences (Wuhan).

Fig. 6.

SiO2–Nb/Y (Winchester & Floyd, Reference Winchester and Floyd1977) and (CaO + Al2O3)–(FeOt + TiO2)–MgO (Zhou et al. Reference Zhou, Robinson, Lesher, Keays, Zhang and Malpas2005) plots of Devonian mafic rocks. Data for Fe–Ti-rich mafic rocks are from Aden & Frizzo (Reference Aden and Frizzo1996), Brooks et al. (Reference Brooks, Larsen and Nielsen1991) and Zhou et al. (Reference Zhou, Robinson, Lesher, Keays, Zhang and Malpas2005).

Fig. 7.

(a, c) Chondrite-normalized REE distribution patterns and (b, d) primitive mantle-normalized trace-element spider diagrams for the Devonian mafic rocks. Data for chondrite are from Taylor & McLennan (Reference Taylor and McLennan1985); data for primitive mantle and E-MORB are from Sun & McDonough (Reference Sun, McDonough, Saunders and Norry1989); IAT (island arc tholeiitic basalt) compositions are from Niu & O’Hara (Reference Niu and O’Hara2003); data for mafic rocks derived from the enriched lithospheric mantle are from Liu et al. (Reference Liu, Ma, Jiang, Guo, Zhang and Xiong2013b).

4.c. Sr–Nd isotope compositions

The whole-rock Sr–Nd isotope compositions of the two groups of samples are listed in Table 3 and plotted in Figure 8. One sample (NM02-1) from the Silurian mafic rocks (∼438 Ma; Liu et al. Reference Liu, Ma, Jiang, Guo, Zhang and Xiong2013b) was also analysed for comparison in the discussion below. The initial ratios of Sr and Nd isotopes for the mafic rocks are calculated using a time of 400 Ma. Group 1 rocks have relatively higher and positive εNd(t) values (+1.2 ∼ +3.9) with lower initial 87Sr/86Sr ratios (I sr) of 0.7065∼0.7074, while Group 2 samples have apparent lower and negative εNd(t) values (−6.8 ∼ −3.4) and higher I sr values that range from 0.7112 to 0.7144. Sample NM02-1 has a moderate I sr value of 0.7072 and a low εNd(t) value of −5.6.

Fig. 8.

Plot of εNd(t) versus Isr and εNd(t) versus Mg no. for the Devonian mafic rocks. Data for the mafic rocks with the ages of 425∼431 Ma are from Zhang et al. (Reference Zhang, Wang, Qian and Li2018). Data for the host granitic intrusion are from Liu et al. (Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012). Data for the Precambrian basement rocks in the EKOB are from Ba et al. (Reference Ba, Chen, Wang, Wang, Zhang and Wang2012) and Yu et al. (Reference Yu, Jin, Ge and Long2005). Values of 3.015 ppm Nd, 92.52 ppm Sr, 87Sr/86Sr = 0.703829 and 147Nd/143Nd = 0.513192 are from Bian et al. (Reference Bian, Li, Pospelov, Yin, Li, Zhao, Chang, Luo, Gao, Astrakhantsev and Chamov2004). Values of 31.4 ppm Nd, 281.40 ppm Sr, 87Sr/86Sr = 0.729107 and 147Nd/143Nd = 0.511899 and values of 31.41 ppm Nd, 243.48 ppm Sr, 87Sr/86Sr = 0.74902 and 147Nd/143Nd = 0.511796 for the two different types of Precambrian basement rocks (Rock 1 and Rock 2) are from Ba et al. (Reference Ba, Chen, Wang, Wang, Zhang and Wang2012) and Yu et al. (Reference Yu, Jin, Ge and Long2005).

5. Discussion

5.a. Petrogenesis of the Devonian mafic rocks

5.a.1. Roles of fractional crystallization and crustal assimilation

Both groups of samples have low and variable MgO, Mg no., Cr and Ni (Table 2) compared to primitive basaltic melts (Tatsumi & Eggins, Reference Tatsumi and Eggins1995), indicating that the samples may have experienced some degree of fractional crystallization. Both groups of rocks have decreasing Cr and Ni with decreasing Mg no. (Fig. 9), suggesting the differentiation of olivine and clinopyroxene. The differentiation of Fe–Ti oxides and plagioclases did not play a vital role in the generation of the two groups of samples, as indicated by the constant or increasing Eu*/Eu ratios, FeOt and TiO2 with decreasing Mg no. (Fig. 9). The Group 2 mafic rocks exhibit strong enrichments in FeOt and TiO2 (Table 2), and opaque minerals (such as Fe–Ti oxides) occur as interstitial phases in the thin-sections (Fig. 3b), suggesting they record the late-stage differentiation of Fe–Ti oxides under low oxygen fugacity (fO2), as discussed below. Moreover, the Group 2 mafic rocks may have experienced some degree of apatite differentiation because of the positive correlation between P2O5 and Mg no. Crustal assimilation accompanied by fractional crystallization (AFC) is prevalent when magma passes through the continental crust and can markedly modify the chemical and isotopic composition of the primary magma (DePaolo, Reference DePaolo1981; Halama et al. Reference Halama, Marks, Brügmann, Siebel, Wenzel and Markl2004). For instance, crustal assimilation into mafic magma would result in obviously increasing contents of large ion lithophile elements, Na2O and K2O with decreasing contents of P2O5 and TiO2, Nb/La ratios and εNd(t) values (e.g. Zhao & Zhou, Reference Zhao and Zhou2009). Both groups of samples presented above exhibit constant or increasing TiO2 and Nb/La with decreasing Mg no. (Figs 9, 10a), which is different from the characteristics of crustal contamination. Group 1 rocks have no correlation between εNd(t) values and Mg no., while Group 2 rocks display a negative correlation between εNd(t) values and Mg no. (Fig. 8b). These features strongly support a negligible contribution from crustal contamination. Lu/Yb ratio cannot be modified by partial melting or fractional crystallization of mafic melts and has been used to evaluate the effects of crustal assimilation (Zhou et al. Reference Zhou, Zhao, Jiang, Gao, Wang and Yang2009). Both groups of mafic rocks have relatively low Lu/Yb ratios (0.14∼0.16 and 0.15∼0.17, respectively), which are comparable to those of mantle-derived melts (0.14–0.15; Sun & McDonough, Reference Sun, McDonough, Saunders and Norry1989), and thus argue against significant crustal contamination. In addition, the uniform distribution patterns of the trace elements (Fig. 7) further indicate that crustal contamination was insignificant for the generation of the two different groups of Devonian mafic rocks in the EKOB.

Fig. 9.

Plots of FeOt, TiO2, P2O5, Cr, Ni and Eu*/Eu versus Mg no. for the Devonian mafic rocks.

Fig. 10.

Plots of Nb/La versus Mg no. and Nb/La versus FeOt for the Devonian mafic rocks.

5.a.2. Mantle sources and magma generation

All the Group 1 samples exhibit enriched mid-ocean ridge basalt (E-MORB)-like trace-element patterns relative to the primitive mantle; such a signature is not typically associated with mafic rocks derived from the enriched lithospheric mantle of the EKOB (Fig. 7b; Liu et al. Reference Liu, Ma, Jiang, Guo, Zhang and Xiong2013b). All the Group 1 samples have relatively higher εNd(t) values (+1.2 ∼ +3.9) than mafic rocks derived from the enriched lithospheric mantle of the EKOB (Fig. 8a), indicating that the mantle source may have been dominated by depleted mantle materials. Most of the samples plot along the line of the MORB–ocean island basalt (OIB) array in Figure 11a, while some samples plot above the line of the MORB–OIB array. These features point to the participation of some crustal materials in their magma generation (Pearce, Reference Pearce2008). Two main pathways have been proposed: source mixing before partial melting and crustal contamination during magma evolution at crustal levels, which could incorporate crustal materials into basaltic melts (Zhao et al. Reference Zhao, Dai and Zheng2013). As mentioned above, the contribution of crustal contamination seems to have been insignificant during magma evolution. The relatively high Th/Yb ratios of some samples may be related to source mixing between the depleted mantle and crustal materials. The results of source mixing modelling suggest that the Group 1 rocks may have originated from a depleted mantle metasomatized by nearly 3–5 % continental crustal components (e.g. Precambrian basement rocks; Fig. 8a). The Group 2 rocks have distinct low and negative εNd(t) values of −6.8 ∼ −3.4 with high I sr values of 0.7112∼0.7144. They plot on the line of binary mixing between crustal materials and depleted mantle on the I sr–εNd(t) diagram (Fig. 8a). The results of source mixing modelling further suggest that up to 30 % crustal components would be required to mix with the mantle to balance the observed variations in the Sr–Nd isotope composition of the Group 2 rocks. However, for such a large volume of crustal components mixed with the depleted mantle, maintaining their relatively low SiO2 contents is impossible. Moreover, the Group 2 rocks mostly plot in the field of mafic rocks derived from the enriched lithospheric mantle of the EKOB (Fig. 8a), further suggesting an enriched lithospheric mantle.

Fig. 11.

Plots of Th/Yb versus Nb/Yb (Pearce, Reference Pearce2008) and 10000Zn/Fet versus 100Mn/Fe (Murray et al. Reference Murray, Ducea, Schoenbohm, DeCelles, Ducea, Carrapa and Kapp2015) for the Devonian mafic rocks.

Partial melting of non-peridotite lithologies (such as pyroxenites) in the deep mantle could be responsible for the generation of mafic melts (e.g. Pilet et al. Reference Pilet, Baker, Muntener and Stolper2011; Dai et al. Reference Dai, Zhao, Zheng and Zhang2012; Murray et al. Reference Murray, Ducea, Schoenbohm, DeCelles, Ducea, Carrapa and Kapp2015). Therefore, ascertaining the petrological characteristics of mantle sources for the Devonian mafic rocks is very necessary. Experimental studies have demonstrated that pyroxenite-derived mafic melts usually exhibit relatively higher Zn/Fe ratios and lower Mn/Fe ratios than peridotite-derived mafic melts (e.g. Sobolev et al. Reference Sobolev, Hofmann, Kuzmin, Yaxley, Arndt, Chung, Danyushevsky, Elliott, Frey, Garcia, Gurenko, Kamenetsky, Kerr, Krivolutskaya, Matvienkov, Nikogosian, Rocholl, Sigurdsson, Sushchevskaya and Teklay2007; Le Roux et al. Reference Le Roux, Dasgupta and Lee2011). As shown in Figure 11b, most of the rocks in the two groups display obvious low Zn/Fe ratios with high Mn/Fe ratios, comparable to those of typical mantle peridotite-derived melts and MORB. These features suggest that the mantle lithologies of the two groups of mafic rocks were mainly composed of peridotites rather than pyroxenites. Furthermore, both groups of mafic rocks have relatively low Ce/Y ratios (0.39∼0.73 and 0.63∼0.86, respectively), low (Tb/Yb)N ratios (1.04∼1.57 and 1.11∼1.35, respectively) and flat HREE patterns (Fig. 7), suggesting that the mantle source may have been dominated by spinel-bearing peridotites (McKenzie, & Bickle, Reference McKenzie and Bickle1988; Wang et al. Reference Wang, Plank, Walker and Smith2002).

Although the Group 2 mafic rocks have been suggested to have been derived from an enriched lithospheric mantle, the genetic mechanisms responsible for the distinct geochemical variations (such as obvious Fe–Ti enrichments and positive Nb anomalies; Figs 6b, 7b) are still unconstrained. Fe–Ti enrichments in mafic rocks have been attributed to various mechanisms, including liquid immiscibility (e.g. Zhou et al. Reference Zhou, Robinson, Lesher, Keays, Zhang and Malpas2005), melting of an Fe-rich eclogite-bearing peridotite (e.g. Gibson, Reference Gibson2002) and crystallization under low oxygen fugacities (fO2) (e.g. Perfit & Fornari, Reference Perfit and Fornari1983; Peng et al. Reference Peng, Wilde, Fan, Peng and Mao2013). Liquid immiscibility is widely used to explain the high Fe–Ti contents of typical layered mafic intrusions (e.g. the Bushveld complex and Skaergaard intrusion; Jakobsen et al. Reference Jakobsen, Veksler, Tegner and Brooks2011; Vantongeren & Mathez, Reference Vantongeren and Mathez2012). The Group 2 mafic rocks display rock associations and field outcrops that are different to the mafic layered intrusions mentioned above, excluding the model of liquid immiscibility. Melting of an Fe-rich eclogite-bearing peridotite produces melts with high contents of TiO2 and FeOt, and the melts usually show OIB- or MORB-type trace-element patterns with high Al2O3 contents (mostly >15 %) and εNd(t) values (>+6; Herzberg & Zhang, Reference Herzberg and Zhang1996; Gibson, Reference Gibson2002). However, this interpretation cannot be the case for the Group 2 mafic rocks because of their relatively low Al2O3 contents of 14.55∼16.28 and strongly negative εNd(t) values of −6.8 ∼ −3.4. Experimental studies and observations have suggested that high degrees of magma differentiation (such as olivine and clinopyroxene differentiation) can promote Fe–Ti enrichment in the residual melts and that low fO2 can also delay Ti-magnetite saturation and then produce melts with high Fe–Ti contents (e.g. Toplis & Carroll, Reference Toplis and Carroll1995; Rutherford et al. Reference Rutherford, Barovich, Hand and Foden2006). As mentioned above, all the Group 2 rocks have relatively low Mg no. values (47–50) and contents of Cr (3.8∼60.4 ppm) and Ni (1.8∼12.5 ppm), suggesting that they may have experienced a high degree of olivine and clinopyroxene fractionation. Increasing FeOt and TiO2 contents with decreasing Mg no. values (Fig. 9) further suggest a delay in crystallization of Fe–Ti oxides during magma evolution, which is also evidenced by the fact that these minerals occur as interstitial phases surrounding clinopyroxenes and plagioclases in the thin-sections (Fig. 3d). The empirical method provided by Smythe & Brenan (Reference Smythe and Brenan2016) was applied for estimating the fO2 conditions at the time of crystallization from Ce anomalies in zircon, and the results show relatively low fO2 conditions (−21 ∼ −20; online Supplementary Material Table S4). Therefore, a high degree of olivine and clinopyroxene crystallization under low fO2 conditions could have been the key mechanism for promoting the high Fe–Ti contents of the Group 2 mafic melts. Experimental studies have shown that high-field-strength elements (such as Nb and Ta) could be moderately compatible in Fe–Ti oxides (e.g. Nielsen & Beard, Reference Nielsen and Beard2000; Klemme et al. Reference Klemme, Günther, Hametner, Prowatke and Zack2006), and the differentiation of Fe–Ti oxides would reduce the contents of Nb and Ta in the evolved magma. The Group 2 rocks exhibit increasing Nb/La ratios with increasing FeOt (Fig. 10b), and a negative correlation between FeOt and Mg no., indicating that the delayed nucleation of Fe–Ti oxides would lead to the generation of Nb-rich melts during the late stage of mafic magma evolution.

5.b. Geodynamic implications

The voluminous eclogites in the EKOB are generally considered to have been formed by the HP metamorphism of Proterozoic continental basaltic rocks (∼934 Ma; e.g. Meng et al. Reference Meng, Zhang and Cui2013, Reference Meng, Cui, Jia, Ren and Feng2015), pointing to deep subduction of the continent crust in the EKOB during continental collision. Moreover, the recent discovery of coesite inclusions in the eclogites further indicates that the continental crust could be subducted into the deep mantle even to depths of 100–120 km (e.g. Bi et al. Reference Bi, Song, Dong, Yang, Qi and Allen2018; Song et al. Reference Song, Bi, Qi, Yang, Allen, Niu, Su and Li2018). How such voluminous eclogites, buried in the deep mantle, can be exhumed to the crustal level of the EKOB is very interesting. However, the mechanism of exhumation of these HP rocks remains poorly understood.

The formation ages of Devonian magmatic rocks in the EKOB (402∼411 Ma) are identical within error to the time of retrograde metamorphism or exhumation of the HP rocks (409∼411 Ma; Fig. 12a; Zhang et al. Reference Zhang, Qian, Li, Wang, Zhang, You and Liu2017; Song et al. Reference Song, Bi, Qi, Yang, Allen, Niu, Su and Li2018), strongly suggesting that the Devonian magmatism may have a close genetic relationship with continental exhumation. It has been documented that the Devonian granitic intrusions were generated by partial melting of Precambrian basement rocks and associated magma mixing in a collision-related environment (e.g. Zhao et al. Reference Zhao, Ma, Wang, Bai, Li and Ji2008; Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012). Underplating of mafic magmas may have played a vital role in the formation of granitic magmas, owing to their unusually high crystallization temperatures (up to 853 °C; Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012). Such a process can also be supported by the new results of this study for the mafic rocks intruding the granitic intrusions. The crystallization ages of the Devonian mafic rocks (406∼408 Ma) are similar to those of the Devonian granitic intrusions, suggesting the activity of mantle-derived magma during the generation of granitic magmas. Compared to the other mafic rocks with ages of 425∼438 Ma, the Group 1 mafic rocks have relatively high and positive εNd(t) values (+1.2 ∼ +3.9; Fig. 8a), further indicating the upwelling of a large volume of hot depleted mantle materials (such as asthenospheric mantle). Lithospheric delamination and slab break-off are generally regarded as two key factors for triggering the upwelling of asthenospheric mantle and resulting in large-scale crustal anatexis in many typical collisional orogenic belts (e.g. Davies & Von Blanckenburg, Reference Davies and Von Blanckenburg1995; Bonin, Reference Bonin2004; Dokuz, Reference Dokuz2011). Collision-related magmatism induced by lithospheric delamination usually exhibits a regional distribution, while the magmatism caused by slab break-off occurs in a narrow and linear zone (Mahéo et al. Reference Mahéo, Blichert-Toft, Pin, Guillot and Pêcher2009; Dokuz, Reference Dokuz2011). The Devonian magmatic rocks exhibit a narrow linear NW–SE distribution along the CKLF (Fig. 1b; Liu et al. Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012), suggestive of slab break-off rather than lithospheric delamination. Many studies have proposed that slab break-off would lead to crustal uplift over a short time period (Davies & Von Blanckenburg, Reference Davies and Von Blanckenburg1995; Xu et al. Reference Xu, Yang, Wu, Li, Zhang, Qi, Song and Qiu2006). The rapid uplift of continental crust also occurred in the EKOB during Devonian time, as shown by the deposition time of local molasse rocks (400∼423 Ma; Lu et al. Reference Lu, Wu, Hu, Barosh, Hao and Zhou2010). In addition, slab break-off usually induces calc-alkaline magmatism associated with small volumes of mafic rocks and leads to the relatively rapid exhumation of eclogites in many typical HP–UHP belts (such as the HP metamorphic belt in the Qiangtang terrane; e.g. Zhang et al. Reference Zhang, Tang, Wang and Zhang2011). The development of voluminous granitic intrusions and minor mafic rocks in the EKOB, associated with the contemporaneous exhumation of extensive HP rocks, further suggests that slab break-off may be a reasonable interpretation for the continental exhumation and the Devonian magmatic activity (Fig. 12).

Fig. 12.

(a) A timeline of Cambrian to Devonian magmatism and typical metamorphism of the EKOB; and (b) a slab break-off model for the continental exhumation and magmatic flare-up in the EKOB. Data sources for the zircon U–Pb ages are as follows: 1 – Zhang et al. (Reference Zhang, Pei, Ding, Li, Feng, Sun, Li and Chen2010); 2 – Cui et al. (Reference Cui, Meng and Wu2011); 3 – Gao et al. (Reference Gao, Xiao, Xie, Fan, Guo and Xi2010); 4 – Xiong et al. (Reference Xiong, Ma, Jiang, Liu, Zhang and Zhou2013); 5 – Zhou et al. (Reference Zhou, Dong, Zhang, Yang, Sun and He2016); 6 – Wang et al. (Reference Wang, Li, Chen, Wang, Li and Jin2016); 7 – Zhang et al. (Reference Zhang, Ma, Xiong, Liu, Li and Pan2014); 8 – Wei et al. (Reference Wei, Sun, Li, Yan, Zhang, Wang, Shen and Yang2018); 9 – Liu et al. (Reference Liu, Ma, Jiang, Guo, Zhang and Xiong2013b); 10 – Liu et al. (Reference Liu, Ma, Zhang, Xiong, Huang and Jiang2012); 11 – Long et al. (Reference Long, Jin, Ge and Yu2006); 12 – Zhang et al. (Reference Zhang, Meng, Wan, Yang and Dong2003); 13 – Zhao et al. (Reference Zhao, Ma, Wang, Bai, Li and Ji2008); 14 – Wang et al. (Reference Wang, Dong and Dong2018); 15 – Liu et al. (Reference Liu, Ma, Guo, Zhang, Xiong and Huang2013a); 16 – Lu et al. (Reference Lu, Yu, Jin, Li and Zheng2002); 17 – Yang et al. (Reference Yang, Robinson, Jiang and Xu1996); 18 – Zhang et al. (Reference Zhang, Wang, Qian and Li2018); 19 – Li et al. (Reference Li, Lu, Xiang, Zhou, Guo, Song, Zheng and Gu2006); 20 – Meng et al. (Reference Meng, Zhang and Cui2013); 21 – Song et al. (Reference Song, Bi, Qi, Yang, Allen, Niu, Su and Li2018); 22 – Zhang et al. (Reference Zhang, Qian, Li, Wang, Zhang, You and Liu2017).

In the classic scenario of slab break-off (e.g. Davies & Von Blanckenburg, Reference Davies and Von Blanckenburg1995; Garzanti et al. Reference Garzanti, Radeff and Malusà2018), the opposing buoyancy forces could induce the extensional deformation of the subducted slab and form early narrow rifts, and the asthenosphere would then rise into the narrow rifts. After the termination of subduction and the replacement of oceanic lithosphere by hot asthenosphere, a series of geological processes (such as magmatic activity, exhumation of eclogite-facies rocks and rapid uplift of the crust) would occur. Moreover, if so, the development of the mafic rocks at ∼425 Ma (Figs 1a, 12a) may mark the early narrow rifting of the subducted slab. Most of these rocks display obviously higher εNd(t) values than the subduction-related mafic rocks (∼438 Ma; Figs 1, 12a) identified in the EKOB (Liu et al. Reference Liu, Ma, Jiang, Guo, Zhang and Xiong2013b), possibly indicating the early upwelling of asthenospheric materials. Furthermore, the Devonian magmatism and rapid crustal uplift associated with the exhumation of extensive HP rocks should be the manifestation of the subsequent slab break-off (Fig. 12b).

Numerical modelling and observations from continental collision zones reveal that the slab strength controls the initiation of slab break-off and the speed of slab tearing (Van Hunen & Allen, Reference Van Hunen and Allen2011). For instance, the timing of slab break-off after continental collision for an old, strong subducting oceanic slab (nearly 20∼25 Ma after the onset of continental collision) is later than that for a young, weak subducting oceanic slab (nearly 10 Ma after the onset of continental collision; Van Hunen & Allen, Reference Van Hunen and Allen2011). Previous studies have documented that the Proto-Tethyan oceanic lithosphere in the EKOB formed during Cambrian time (518∼522 Ma; Fig. 12a; Yang et al. Reference Yang, Robinson, Jiang and Xu1996; Lu et al. Reference Lu, Yu, Jin, Li and Zheng2002) and then experienced prolonged subduction during early Cambrian to early Silurian times (438∼515 Ma; Fig. 12a; Gao et al. Reference Gao, Xiao, Xie, Fan, Guo and Xi2010; Zhang et al. Reference Zhang, Pei, Ding, Li, Feng, Sun, Li and Chen2010; Xiong et al. Reference Xiong, Ma, Jiang, Liu, Zhang and Zhou2013; Liu et al. Reference Liu, Ma, Jiang, Guo, Zhang and Xiong2013b), suggesting an old and strong subducting oceanic slab. If the time of the slab break-off is assumed to be 411 Ma, the continental collision and the beginning of continental subduction should have initiated during the interval from 431 to 436 Ma.

6. Conclusions

  1. (1) Zircon U–Pb ages of the Devonian mafic rocks are 406∼408 Ma and are comparable to the exhumation time of extensive HP rocks in the EKOB.

  2. (2) Group 1 mafic rocks were formed by partial melting of depleted asthenospheric mantle metasomatized by c. 3–5 % continental crustal components, while Group 2 rocks originated from partial melting of enriched lithospheric mantle.

  3. (3) The high degree of olivine crystallization under low fO2 conditions and the delayed nucleation of Fe–Ti oxides could have been responsible for the enrichment in Fe, Ti and Nb in the Group 2 rocks.

  4. (4) Slab break-off was the key factor causing the exhumation of extensive HP rocks and triggering Devonian magmatic activity, and continental collision or continental subduction may have initiated during the interval from 431 to 436 Ma.

Supplementary Material

To view supplementary material for this article, please visit https://doi.org/10.1017/S0016756819001353.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (Grant 41972148 and 41502050), the Foundation of the State Key Laboratory of Petroleum Resources and Prospecting, China University of Petroleum, Beijing (PRP/open-1908), the China Geological Survey (Grants DD20160022 and 12120115026901) and the Yangtze Youth Fund (Grant 2015cqn29). We thank Editor-in-Chief Dr Kathryn Goodenough and two anonymous reviewers for their valuable comments and constructive suggestions on the manuscript. We also thank Zhao-Chu Hu, Lian Zhou and Hai-Hong Chen for their help with the analytical work.

References

Aden, A and Frizzo, P (1996) Geochemistry and origin of low and high TiO2 mafic rocks in the Barkasan complex: a comparison with common Neoproterozoic gabbros of northern Somali crystalline basement. Journal of African Earth Sciences 22, 43–54.CrossRefGoogle Scholar
Ba, J, Chen, NS, Wang, QY, Wang, XY, Zhang, L and Wang, SQ (2012) Nd-Sr-Pb isotopic compositions of cordierite granite on southern margin of the Qaidam block, NW China, and constraints on its petrogenesis, tectonic affinity of source region and tectonic implications. Earth Science – Journal of China University of Geosciences 37, 80–92 (in Chinese with English abstract).Google Scholar
Bi, H, Song, SG, Dong, JL, Yang, LM, Qi, SS and Allen, MB (2018) First discovery of coesite in eclogite from East Kunlun, northwest China. Science Bulletin 63, 1536–8.CrossRefGoogle Scholar
Bian, QT, Li, DH, Pospelov, I, Yin, LM, Li, HS, Zhao, DS, Chang, CF, Luo, XQ, Gao, SL, Astrakhantsev, Q and Chamov, N (2004) Age, geochemistry, and tectonic setting of Buqingshan ophiolites, North Qinghai-Tibet Plateau, China. Journal of Asian Earth Sciences 23, 577–96.CrossRefGoogle Scholar
Bonin, B (2004) Do coeval mafic and felsic magmas in post-collisional to within-plate regimes necessarily imply two contrasting, mantle and crustal, sources? A review. Lithos 78, 1–24.CrossRefGoogle Scholar
Brooks, CK, Larsen, LM and Nielsen, TFD (1991) Importance of iron-rich tholeiitic magmas at divergent plate margins: a reappraisal. Geology 19, 269–72.2.3.CO;2>CrossRefGoogle Scholar
Chen, YX, Pei, XZ, Li, RB, Liu, ZQ, Li, ZC, Zhang, XF, Chen, GC, Liu, ZG, Ding, SP and Guo, JF (2011) Zircon U–Pb age of Xiaomiao Formation of Proterozoic in the eastern section of the East Kunlun Orogenic Belt. Geoscience 25, 510–21 (in Chinese with English abstract).Google Scholar
Chung, S, Chu, M, Zhang, Y, Xie, Y, Lo, C, Lee, T, Lan, C, Li, X, Zhang, Q and Wang, Y (2005) Tibetan tectonic evolution inferred from spatial and temporal variations in post-collisional magmatism. Earth-Science Reviews 68, 173–96.CrossRefGoogle Scholar
Corfu, F, Hanchar, JM, Hoskin, P and Kinny, P (2003) Atlas of zircon textures. Reviews in Mineralogy and Geochemistry 53, 469–500.CrossRefGoogle Scholar
Cui, MH, Meng, FC and Wu, XK (2011) Early Ordovician island arc of Qimantag Mountain, eastern Kunlun: evidences from geochemistry, Sm-Nd isotope and geochronology of intermediate-basic igneous rocks. Acta Petrologica Sinica 27, 3365–79 (in Chinese with English abstract).Google Scholar
Dai, L, Zhao, Z, Zheng, Y and Zhang, J (2012) The nature of orogenic lithospheric mantle: geochemical constraints from postcollisional mafic–ultramafic rocks in the Dabie orogen. Chemical Geology 334, 99–121.CrossRefGoogle Scholar
Davies, HJ and Von Blanckenburg, F (1995) Slab breakoff: a model of lithosphere detachment and its test in the magmatism and deformation of collisional orogens. Earth and Planetary Science Letters 129, 85–102.CrossRefGoogle Scholar
Defant, MJ and Drummond, MS (1990) Derivation of some modern arc magmas by melting of young subducted lithosphere. Nature 347, 662–5.CrossRefGoogle Scholar
DePaolo, DJ (1981) Trace element and isotopic effects of combined wallrock assimilation and fractional crystallization. Earth and Planetary Science Letters 53, 189–202.CrossRefGoogle Scholar
Dokuz, A (2011) A slab detachment and delamination model for the generation of Carboniferous high-potassium I-type magmatism in the Eastern Pontides, NE Turkey: the Köse composite pluton. Gondwana Research 19, 926–44.CrossRefGoogle Scholar
Dong, YP, He, DF, Sun, SY, Liu, XM, Zhou, XH, Zhang, FF, Yang, Z, Cheng, B, Zhao, GC and Li, JH (2018) Subduction and accretionary tectonics of the East Kunlun orogen, western segment of the Central China Orogenic System. Earth-Science Reviews 186, 231–61.CrossRefGoogle Scholar
Fantle, MS and DePaolo, DJ (2006) Sr isotopes and pore fluid chemistry in carbonate sediment of the Ontong Java Plateau: calcite recrystallization rates and evidence for a rapid rise in seawater Mg over the last 10 million years. Geochimica et Cosmochimica Acta 70, 3883–904.CrossRefGoogle Scholar
Gao, S, Rudnick, RL, Yuan, HL, Liu, XM, Liu, YS, Xu, WL, Ling, WL, Ayers, J, Wang, XC and Wang, QH (2004) Recycling lower continental crust in the North China craton. Nature 432, 892–7.CrossRefGoogle ScholarPubMed
Gao, XF, Xiao, PX, Xie, CR, Fan, LY, Guo, L and Xi, RG (2010) Zircon LA-ICP-MS U–Pb dating and geological significance of Bashierxi granite in the eastern Kunlun area, China. Geological Bulletin of China 29, 1001–8 (in Chinese with English abstract).Google Scholar
Garzanti, E, Radeff, G and Malusà, MG (2018) Slab breakoff: a critical appraisal of a geological theory as applied in space and time. Earth-Science Reviews 177, 303–19.CrossRefGoogle Scholar
Gibson, SA (2002) Major element heterogeneity in Archean to recent mantle plume starting-heads. Earth and Planetary Science Letters 195, 59–74.CrossRefGoogle Scholar
Halama, R, Marks, M, Brügmann, G, Siebel, W, Wenzel, T and Markl, G (2004) Crustal contamination of mafic magmas: evidence from a petrological, geochemical and Sr–Nd–Os–O isotopic study of the Proterozoic Isortoq dike swarm, South Greenland. Lithos 74, 199–232.CrossRefGoogle Scholar
Harris, NBW, Xu, RH, Lewis, CL and Jin, C (1988) Plutonic rocks of the 1985 Tibet geotraverse, Lhasa to Golmud. Philosophical Transactions of the Royal Society of London A: Mathematical, Physical and Engineering Sciences 327, 145–68.Google Scholar
Herzberg, C and Zhang, J (1996) Melting experiments on anhydrous peridotite KLB-1: compositions of magmas in the upper mantle and transition zone. Journal of Geophysical Research: Solid Earth 101, 8271–95.CrossRefGoogle Scholar
Hofmann, AW (1997) Mantle geochemistry: the message from oceanic volcanism. Nature 385, 219–29.CrossRefGoogle Scholar
Hoskin, PWO and Schaltegger, U (2003) The composition of zircon and igneous and metamorphic petrogenesis. Reviews in Mineralogy and Geochemistry 53, 27–62.CrossRefGoogle Scholar
Jakobsen, JK, Veksler, IV, Tegner, C and Brooks, CK (2011) Crystallization of the Skaergaard intrusion from an emulsion of immiscible iron- and silica-rich liquids: evidence from melt inclusions in plagioclase. Journal of Petrology 52, 345–73.CrossRefGoogle Scholar
Klemme, S, Günther, D, Hametner, K, Prowatke, S and Zack, T (2006) The partitioning of trace elements between ilmenite, ulvospinel, armalcolite and silicate melts with implications for the early differentiation of the moon. Chemical Geology 234, 251–63.CrossRefGoogle Scholar
Le Roux, V, Dasgupta, R and Lee, C-A (2011) Mineralogical heterogeneities in the Earth’s mantle: constraints from Mn, Co, Ni and Zn partitioning during partial melting. Earth and Planetary Science Letters 307, 395–408.CrossRefGoogle Scholar
Li, HK, Lu, SN, Xiang, ZQ, Zhou, HY, Guo, H, Song, B, Zheng, JK and Gu, Y (2006) SHRIMP U–Pb zircon age of the granulite from the Qingshuiquan area, Central Eastern Kunlun Suture Zone. Earth Science Frontiers 13, 311–21 (in Chinese with English abstract).Google Scholar
Liu, Y, Gao, S, Hu, Z, Gao, C, Zong, K and Wang, D (2010a) Continental and oceanic crust recycling-induced melt-peridotite interactions in the trans-North China orogen: U–Pb dating, Hf isotopes and trace elements in zircons from mantle xenoliths. Journal of Petrology 51, 537–71.CrossRefGoogle Scholar
Liu, Y, Hu, Z, Gao, S, Günther, D, Xu, J, Gao, C and Chen, H (2008a) 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.CrossRefGoogle Scholar
Liu, Y, Hu, Z, Zong, K, Gao, C, Gao, S, Xu, J and Chen, H (2010b) Reappraisement and refinement of zircon U–Pb isotope and trace element analyses by LA-ICP-MS. Chinese Science Bulletin 55, 1535–46.CrossRefGoogle Scholar
Liu, B, Ma, CQ, Guo, P, Zhang, JY, Xiong, FH and Huang, J (2013a) Discovery of the Middle Devonian A-type granite from the Eastern Kunlun orogen and its tectonic implications. Earth Science – Journal of China University of Geosciences 38, 947–62 (in Chinese with English abstract).Google Scholar
Liu, B, Ma, CQ, Jiang, HA, Guo, P, Zhang, JY and Xiong, FH (2013b) Early Paleozoic tectonic transition from ocean subduction to collisional orogeny in the Eastern Kunlun region: evidence from Huxiaoqin mafic rocks. Acta Petrologica Sinica 29, 2093–106 (in Chinese with English abstract).Google Scholar
Liu, B, Ma, CQ, Zhang, JY, Xiong, FH, Huang, J and Jiang, HA (2012) Petrogenesis of Early Devonian intrusive rocks in the east part of Eastern Kunlun Orogen and implications for Early Palaeozoic orogenic processes. Acta Petrologica Sinica 28, 1785–807 (in Chinese with English abstract).Google Scholar
Liu, Y, Zong, K, Kelemen, PB and Gao, S (2008b) Geochemistry and magmatic history of eclogites and ultramafic rocks from the Chinese continental scientific drill hole: subduction and ultrahigh-pressure metamorphism of lower crustal cumulates. Chemical Geology 247, 133–53.CrossRefGoogle Scholar
Long, XP, Jin, W, Ge, WC and Yu, N (2006) Zircon U–Pb geochronology and geological implications of the granitoids in Jinshuikou, East Kunlun, NW China. Geochimica 35, 333–45 (in Chinese with English abstract).Google Scholar
Lu, L, Wu, ZH, Hu, DG, Barosh, PJ, Hao, S and Zhou, CJ (2010) Zircon U–Pb age for rhyolite of the Maoniushan Formation and its tectonic significance in the East Kunlun Mountains. Acta Petrologica Sinica 26, 1150–8 (in Chinese with English abstract).Google Scholar
Lu, SN, Yu, HF, Jin, W, Li, HK and Zheng, JK (2002) Microcontinents on the eastern margin of Tarim paleocontinent. Acta Petrologica et Mineralogica 21, 317–26 (in Chinese with English abstract).Google Scholar
Mahéo, G, Blichert-Toft, J, Pin, C, Guillot, S and Pêcher, A (2009) Partial melting of mantle and crustal sources beneath South Karakorum, Pakistan: implications for the Miocene geodynamic evolution of the India–Asia convergence zone. Journal of Petrology 50, 427–49.CrossRefGoogle Scholar
McKenzie, DP and Bickle, MJ (1988) The volume and composition of melt generated by extension of the lithosphere. Journal of Petrology 29, 625–79.CrossRefGoogle Scholar
Meng, FC, Cui, MH, Jia, LP, Ren, YF and Feng, HB (2015) Paleozoic continental collision of the East Kunlun orogen: evidence from protoliths of the eclogites. Acta Petrologica Sinica 31, 3581–94 (in Chinese with English abstract).Google Scholar
Meng, FC, Zhang, JX and Cui, MH (2013) Discovery of Early Paleozoic eclogite from the East Kunlun, Western China and its tectonic significance. Gondwana Research 23, 825–36.CrossRefGoogle Scholar
Mo, XX, Luo, ZH, Deng, JF and Yu, XH (2007) Granitoids and crustal growth in the east-Kunlun orogenic belt. Geological Journal of China Universities 3, 403–14 (in Chinese with English abstract).Google Scholar
Murray, KE, Ducea, MN and Schoenbohm, L (2015) Foundering-driven lithospheric melting: the source of central Andean mafic lavas on the Puna Plateau (22°S–27°S). In Geodynamics of a Cordilleran Orogenic System: The Central Andes of Argentina and Northern Chile (eds DeCelles, PG, Ducea, MN, Carrapa, B and Kapp, PA), pp. 139–66. Geological Society of America Memoirs no. 212.CrossRefGoogle Scholar
Nielsen, RL and Beard, JS (2000) Magnetite-melt HFSE partitioning. Chemical Geology 164, 21–34.CrossRefGoogle Scholar
Niu, Y and O’Hara, MJ (2003) Origin of ocean island basalts: a new perspective from petrology, geochemistry, and mineral physics considerations. Journal of Geophysical Research 108, 1–19.CrossRefGoogle Scholar
Pearce, JA (2008) Geochemical fingerprinting of oceanic basalts with applications to ophiolite classification and the search for Archean oceanic crust. Lithos 100, 14–48.CrossRefGoogle Scholar
Peng, T, Wilde, SA, Fan, W, Peng, B and Mao, Y (2013) Mesoproterozoic high Fe–Ti mafic magmatism in western Shandong, North China Craton: petrogenesis and implications for the final breakup of the Columbia supercontinent. Precambrian Research 235, 190–207.CrossRefGoogle Scholar
Perfit, MR and Fornari, DJ (1983) Geochemical studies of abyssal lavas recovered by DSRV Alvin from Eastern Galapagos Rift, Inca Transform, and Ecuador Rift: 2. Phase chemistry and crystallization history. Journal of Geophysical Research: Solid Earth 88, 10530–50.CrossRefGoogle Scholar
Pilet, S, Baker, MB, Muntener, O and Stolper, EM (2011) Monte Carlo simulations of meta-somatic enrichment in the lithosphere and implications for the source of alkaline basalts. Journal of Petrology 52, 1415–42.CrossRefGoogle Scholar
Rutherford, L, Barovich, K, Hand, M and Foden, J (2006) Continental ca 1.7–1.69 Ga Fe-rich metatholeiites in the Curnamona Province, Australia: a record of melting of a heterogeneous, subduction-modified lithospheric mantle. Journal of the Geological Society of Australia 53, 501–19.Google Scholar
Smith, DC (1984) Coesite in clinopyroxene in the Caledonides and its implications for geodynamics. Nature 310, 641–4.CrossRefGoogle Scholar
Smythe, DJ and Brenan, JM (2016) Magmatic oxygen fugacity estimated using zircon-melt partitioning of cerium. Earth and Planetary Science Letters 453, 260–6.CrossRefGoogle Scholar
Sobolev, AV, Hofmann, AW, Kuzmin, DV, Yaxley, GM, Arndt, NT, Chung, S-L, Danyushevsky, LV, Elliott, T, Frey, FA, Garcia, MO, Gurenko, AA, Kamenetsky, VS, Kerr, AC, Krivolutskaya, NA, Matvienkov, VV, Nikogosian, IK, Rocholl, A, Sigurdsson, IA, Sushchevskaya, NM and Teklay, M (2007) The amount of recycled crust in sources of mantle-derived melts. Science 316, 412–7.CrossRefGoogle ScholarPubMed
Sobolev, NV and Shatsky, VS (1990) Diamond inclusions in garnets from metamorphic rocks; a new environment of diamond formation. Nature 343, 742–6.CrossRefGoogle Scholar
Song, SG, Bi, HZ, Qi, SS, Yang, LM, Allen, MB, Niu, YL, Su, L and Li, WF (2018) HP–UHP metamorphic belt in the East Kunlun orogen: final closure of the proto-Tethys ocean and formation of the pan-North-China continent. Journal of Petrology 59, 2043–60.CrossRefGoogle Scholar
Spandler, C and Pirard, C (2013) Element recycling from subducting slabs to arc crust: a review. Lithos 170–171, 208–23.CrossRefGoogle Scholar
Sun, SS and McDonough, WF (1989) Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. In Magmatism in the Ocean Basins (eds Saunders, AD and Norry, MJ), pp. 313–45. Geological Society of London, Special Publication no. 42.CrossRefGoogle Scholar
Tanaka, T, Togashi, S, Kamioka, H, Amakawa, H, Kagami, H, Hamamoto, T, Yuhara, M, Orihashi, Y, Yoneda, S, Shimizu, H, Kunimaru, T, Takahashi, K, Yanagi, T, Nakano, T, Fujimaki, H, Shinjo, R, Asahara, Y, Tanimizu, M and Dragusanu, C (2000) JNdi-1: a neodymium isotopic reference in consistency with LaJolla neodymium. Chemical Geology 168, 279–81.CrossRefGoogle Scholar
Tatsumi, Y and Eggins, S (1995) Subduction Zone Magmatism. Cambridge: Blackwell Publishing House Science, 211 pp.Google Scholar
Taylor, SR and McLennan, SM (1985) The Continental Crust: Its Composition and Evolution. Oxford: Blackwell Scientific Publications, 328 pp.Google Scholar
Toplis, MJ and Carroll, MR (1995) An experimental study of the influence of oxygen fugacity on Fe-Ti oxide stability, phase relations, and mineral–melt equilibria in ferro-basaltic systems. Journal of Petrology 36, 1137–70.CrossRefGoogle Scholar
Van Hunen, J and Allen, MB (2011) Continental collision and slab break-off: a comparison of 3-D numerical models with observations. Earth and Planetary Science Letters 302, 27–37.CrossRefGoogle Scholar
Vantongeren, JA and Mathez, EA (2012) Large-scale liquid immiscibility at the top of the Bushveld Complex. Geology 40, 3.CrossRefGoogle Scholar
Wang, X, Dong, GC and Dong, LQ (2018) Genesis of Early Devonian granitoid in east segment of East Kunlun: evidences from Longmoka Pluton. Journal of Earth Sciences and Environment 40, 414–27.Google Scholar
Wang, T, Li, B, Chen, J, Wang, JS., Li, WF and Jin, TT (2016) Characteristics of chronology and geochemistry of the early Silurian monzogranite in the Wulonggou area, East Kunlun and its geological significance. Journal of Mineralogical and Petrological Sciences 36, 62–70 (in Chinese with English abstract).Google Scholar
Wang, K, Plank, T, Walker, JD and Smith, EI (2002) A mantle melting profile across the Basin and Range, SW USA. Journal of Geophysical Research: Solid Earth 107, ECV 5-1–ECV 5-21 doi: 10.1029/2001JB000209.CrossRefGoogle Scholar
Wang, GC, Wei, QR, Jia, XC, Zhang, KX, Li, DW, Zhu, YH and Xiang, SY (2007) Some ideas of Precambrian geology in the East Kunlun, China. Geological Bulletin of China 26, 929–37 (in Chinese with English abstract).Google Scholar
Wang, H, Wu, YB, Qin, ZW, Zhu, LQ, Liu, Q, Liu, XC, Gao, S, Wijbrans, JR, Zhou, L, Gong, HJ and Yuan, HL (2013) Age and geochemistry of Silurian gabbroic rocks in the Tongbai orogen, central China: implications for the geodynamic evolution of the North Qinling arc–back-arc system. Lithos 179, 1–15.CrossRefGoogle Scholar
Wei, X, Sun, FY, Li, L, Yan, JM, Zhang, YT, Wang, YC, Shen, TS and Yang, YJ (2018) The Wulonggou metaluminous A2-type granites in the Eastern Kunlun Orogenic Belt, NW China: rejuvenation of subduction-related felsic crust and implications for post-collision extension. Lithos 312–313, 108–27.Google Scholar
Winchester, JA and Floyd, PA (1977) Geochemical discrimination of different magma series and their differentiation products using immobile elements. Chemical Geology 20, 325–43.CrossRefGoogle Scholar
Xiong, FH, Ma, CQ, Jiang, HA, Liu, B, Zhang, JY and Zhou, Q (2013) Petrogenetic and tectonic significance of Permian calc-alkaline lamprophyres, East Kunlun orogenic belt, Northern Qinghai-Tibet Plateau. International Geology Review 55, 1817–34. doi: 10.1080/00206814.2013.804683.CrossRefGoogle Scholar
Xiong, FH, Ma, CQ, Zhang, JY, Liu, B and Jiang, HA (2014) Reworking of old continental lithosphere: an important crustal evolution mechanism in orogenic belts, as evidenced by Triassic I-type granitoids in the East Kunlun orogen, northern Tibetan Plateau. Journal of the Geological Society, London 171, 847–63.CrossRefGoogle Scholar
Xu, Z, Yang, J, Wu, C, Li, H, Zhang, J, Qi, X, Song, S and Qiu, H (2006) Timing and mechanism of formation and exhumation of the northern Qaidam ultrahigh-pressure metamorphic belt. Journal of Asian Earth Sciences 28, 160–73.Google Scholar
Yang, JS, Robinson, PT, Jiang, CF and Xu, ZQ (1996) Ophiolites of the Kunlun mountains, China and their tectonic implications. Tectonophysics 258, 215–31.CrossRefGoogle Scholar
Yu, N, Jin, W, Ge, WC and Long, XP (2005) Geochemical study on peraluminous granite from Jinshuikou in East Kunlun. Global Geology 24, 123–8 (in Chinese with English abstract).Google Scholar
Zhang, JY, Ma, CQ, Xiong, FH, Liu, B, Li, JW and Pan, YM (2014) Early Paleozoic high-Mg diorite-granodiorite in the eastern Kunlun Orogen, western China: response to continental collision and slab break-off. Lithos 210–211, 129–46.CrossRefGoogle Scholar
Zhang, JX, Meng, FC, Wan, YS, Yang, JS and Dong, GA (2003) Early Paleozoic tectono-thermal event of the Jingshuikou Group on the southern margin of Qaidam: zircon U–Pb SHRIMP age evidence. Geological Bulletin of China 22, 397–404 (in Chinese with English abstract).Google Scholar
Zhang, YF, Pei, XZ, Ding, SP, Li, RB, Feng, JY, Sun, Y, Li, ZC and Chen, YX (2010) LA-ICP-MS zircon U–Pb age of quartz diorite at the Kekesha area of Dulan County, eastern section of the East Kunlun orogenic belt, China and its significance. Geological Bulletin of China 29, 79–85.Google Scholar
Zhang, ZW, Qian, B, Li, WY, Wang, YL, Zhang, JW, You, MX and Liu, YG (2017) The discovery of Early Paleozoic eclogite from the Xiarihamu magmatic Ni-Cu sulfide deposit in eastern Kunlun orogenic belt: zircon U–Pb chronologic evidence. Geology in China 44, 816–7 (in Chinese with English abstract).Google Scholar
Zhang, K, Tang, X, Wang, Y and Zhang, Y (2011) Geochronology, geochemistry, and Nd isotopes of early Mesozoic bimodal volcanism in northern Tibet, western China: constraints on the exhumation of the central Qiangtang metamorphic belt. Lithos 121, 167–75.CrossRefGoogle Scholar
Zhang, ZW, Wang, CY, Qian, B and Li, WY (2018) The geochemistry characteristics of Silurian gabbro in East Kunlun Orogenic Belt and its mineralization relationship with magmatic Ni-Ci sulphide deposit. Acta Petrologica Sinica 34, 2262–74 (in Chinese with English abstract).Google Scholar
Zhao, ZF, Dai, LQ and Zheng, YF (2013) Postcollisional mafic igneous rocks record crust–mantle interaction during continental deep subduction. Scientific Reports 3, 3413.CrossRefGoogle ScholarPubMed
Zhao, ZM, Ma, HD, Wang, BZ, Bai, YS, Li, RS and Ji, WH (2008) The evidence of intrusive rocks about collision-orogeny during Early Devonian in Eastern Kunlun Area. Geological Review 54, 47–56 (in Chinese with English abstract).Google Scholar
Zhao, J and Zhou, M (2009) Secular evolution of the Neoproterozoic lithospheric mantle underneath the northern margin of the Yangtze block, South China. Lithos 107, 152–68.CrossRefGoogle Scholar
Zheng, YF (2012) Metamorphic chemical geodynamics in continental subduction zones. Chemical Geology 328, 5–48.CrossRefGoogle Scholar
Zhou, B, Dong, YP, Zhang, FF, Yang, Z, Sun, SS and He, DF (2016) Geochemistry and zircon U–Pb geochronology of granitoids in the East Kunlun Orogenic Belt, northern Tibetan Plateau: origin and tectonic implications. Journal of Asian Earth Sciences 130, 265–81.CrossRefGoogle Scholar
Zhou, MF, Robinson, PT, Lesher, CM, Keays, RR, Zhang, C and Malpas, J (2005) Geochemistry, petrogenesis and metallogenesis of the Panzhihua gabbroic layered intrusion and associated Fe–Ti–V oxide deposits, Sichuan Province, SW China. Journal of Petrology 46, 2253–80.CrossRefGoogle Scholar
Zhou, MF, Zhao, JH, Jiang, CY, Gao, JF, Wang, W and Yang, SH (2009) OIB-like, heterogeneous mantle sources of Permian basaltic magmatism in the western Tarim Basin, NW China: implications for a possible Permian large igneous province. Lithos 113, 583–94.CrossRefGoogle Scholar
Figure 0

Fig. 1. (a) Simplified geological map showing the distribution of magmatic rocks in the EKOB; and (b) the location of the EKOB in the Tibetan Plateau.

Figure 1

Fig. 2. Simplified geological map of the study area.

Figure 2

Fig. 3. Field photographs and photomicrographs for the two groups of Devonian mafic rocks. Cpx – clinopyroxene; Hb – hornblende; Opa – opaque minerals; Pl – plagioclase. Hammer for scale is ∼35 cm long.

Figure 3

Table 1. Zircon LA-ICP-MS U–Pb dating results of Devonian mafic rocks

Figure 4

Fig. 4. (a) CL images and (b) LA-ICP-MS U–Pb concordant diagrams for zircons from the Devonian mafic rocks. Data for chondrite are from Taylor & McLennan (1985). CL images of typical magmatic zircons from the host granitic intrusion (after Liu et al.2012) are also shown for comparison.

Figure 5

Fig. 5. (a) Chondrite-normalized REE patterns and (b) Eu/Eu* v. Sm plot for zircons from the Devonian mafic rocks. Trace-element compositions of typical magmatic zircons from the host granitic intrusion (after Liu et al.2012) are also shown for comparison.

Figure 6

Table 2. Major- (in wt %) and trace-element (in ppm) compositions of Devonian mafic rocks

Figure 7

Table 3. Sr–Nd isotopic compositions of the selected samples from the Devonian mafic rocks

Figure 8

Fig. 6. SiO2–Nb/Y (Winchester & Floyd, 1977) and (CaO + Al2O3)–(FeOt + TiO2)–MgO (Zhou et al. 2005) plots of Devonian mafic rocks. Data for Fe–Ti-rich mafic rocks are from Aden & Frizzo (1996), Brooks et al. (1991) and Zhou et al. (2005).

Figure 9

Fig. 7. (a, c) Chondrite-normalized REE distribution patterns and (b, d) primitive mantle-normalized trace-element spider diagrams for the Devonian mafic rocks. Data for chondrite are from Taylor & McLennan (1985); data for primitive mantle and E-MORB are from Sun & McDonough (1989); IAT (island arc tholeiitic basalt) compositions are from Niu & O’Hara (2003); data for mafic rocks derived from the enriched lithospheric mantle are from Liu et al. (2013b).

Figure 10

Fig. 8. Plot of εNd(t) versus Isr and εNd(t) versus Mg no. for the Devonian mafic rocks. Data for the mafic rocks with the ages of 425∼431 Ma are from Zhang et al. (2018). Data for the host granitic intrusion are from Liu et al. (2012). Data for the Precambrian basement rocks in the EKOB are from Ba et al. (2012) and Yu et al. (2005). Values of 3.015 ppm Nd, 92.52 ppm Sr, 87Sr/86Sr = 0.703829 and 147Nd/143Nd = 0.513192 are from Bian et al. (2004). Values of 31.4 ppm Nd, 281.40 ppm Sr, 87Sr/86Sr = 0.729107 and 147Nd/143Nd = 0.511899 and values of 31.41 ppm Nd, 243.48 ppm Sr, 87Sr/86Sr = 0.74902 and 147Nd/143Nd = 0.511796 for the two different types of Precambrian basement rocks (Rock 1 and Rock 2) are from Ba et al. (2012) and Yu et al. (2005).

Figure 11

Fig. 9. Plots of FeOt, TiO2, P2O5, Cr, Ni and Eu*/Eu versus Mg no. for the Devonian mafic rocks.

Figure 12

Fig. 10. Plots of Nb/La versus Mg no. and Nb/La versus FeOt for the Devonian mafic rocks.

Figure 13

Fig. 11. Plots of Th/Yb versus Nb/Yb (Pearce, 2008) and 10000Zn/Fet versus 100Mn/Fe (Murray et al.2015) for the Devonian mafic rocks.

Figure 14

Fig. 12. (a) A timeline of Cambrian to Devonian magmatism and typical metamorphism of the EKOB; and (b) a slab break-off model for the continental exhumation and magmatic flare-up in the EKOB. Data sources for the zircon U–Pb ages are as follows: 1 – Zhang et al. (2010); 2 – Cui et al. (2011); 3 – Gao et al. (2010); 4 – Xiong et al. (2013); 5 – Zhou et al. (2016); 6 – Wang et al. (2016); 7 – Zhang et al. (2014); 8 – Wei et al. (2018); 9 – Liu et al. (2013b); 10 – Liu et al. (2012); 11 – Long et al. (2006); 12 – Zhang et al. (2003); 13 – Zhao et al. (2008); 14 – Wang et al. (2018); 15 – Liu et al. (2013a); 16 – Lu et al. (2002); 17 – Yang et al. (1996); 18 – Zhang et al. (2018); 19 – Li et al. (2006); 20 – Meng et al. (2013); 21 – Song et al. (2018); 22 – Zhang et al. (2017).

Supplementary material: File

Tang et al. supplementary materials

Tang et al. supplementary materials

Download Tang et al. supplementary materials(File)
File 251.4 KB