Abstract

Early Jurassic granitoids are widespread in the Lesser Xing’an–Zhangguangcai Ranges, providing excellent targets to understand the late Paleozoic to early Mesozoic tectonic framework and evolution of Northeast China, especially the Jiamusi block and its related structural belts. In this paper, we present new geochronological, geochemical, and isotopic data from the granitoids in the Lesser Xing’an–Zhangguangcai Ranges to constrain the early Mesozoic tectonic evolution of the Mudanjiang Ocean between the Jiamusi and Songnen blocks. Our results show that the granitic intrusions in the Lesser Xing’an–Zhangguangcai Ranges are mainly composed of syenogranite, monzogranite, granodiorite, and tonalite, which have crystallization ages from 196 to 181 Ma. Their geochemical features indicate that these Jurassic intrusions are all high-K calc-alkaline I-type granites with metaluminous to weakly peraluminous compositions. These granitoids are characterized by enrichments in large ion lithophile elements (e.g., Ba, Th, U) and light rare earth elements and depletions in high field strength elements (e.g., Nb and Ta) and heavy rare earth elements, which are typical for continental arc–type granites. The sources of these granitoids were likely derived from juvenile Mesoproterozoic to Neoproterozoic crustal materials (e.g., metabasaltic rocks). Integrated with data from regional coeval magmatism, metamorphism, metallogeny, and structure, our new data suggest that the granitoids in the Lesser Xing’an–Zhangguangcai Ranges were probably formed in an active continental margin setting, which fits well in our previous model of Early Jurassic westward subduction of the Mudanjiang Ocean between the Jiamusi and Songnen blocks.

INTRODUCTION

Granitoid rock, one of the principal components of continental crust, plays an important role in exploring the formation, evolution, and reworking of continental crust (Hu et al., 2016; Wu et al., 2011). It has been evidenced that granitoid rocks formed at distinct evolutionary stages of an orogenic belt have different geological and geochemical characteristics (Barbarin, 1999; Pearce et al., 1984). Thus, the study of granitoids can help to explore the related tectonic environments and better understand crustal growth and tectonic evolution on Earth (Maniar and Piccoli, 1989; Wu et al., 2011).

Northeast China (NE China) is characterized by exposure of large volumes of granitic intrusions with emplacement ages from the Paleozoic to Mesozoic (Wu et al., 2011). Despite intensive geochronological and geochemical examinations of these granitoids, there exists strong debate on the tectonic affinity of these granitoids, especially for those in the Lesser Xing’an–Zhangguangcai Ranges (LXZR) (Ge et al., 2017, 2018; Liu et al., 2017a; M.J. Xu et al., 2013a; Wu et al., 2011; Zhao et al., 2018; Zhu et al., 2017). The well-developed late Paleozoic to early Mesozoic granitoids in the LXZR are some of the most prominent products of the regional tectonic reorganization and amalgamation between the Jiamusi and Songnen blocks (Ge et al., 2017, 2018; Liu et al., 2017a; Wu et al., 2011). However, the formation mechanism of these granitoids is still controversial. On one hand, some previous views suggested that these granitoids could have formed as a result of delamination following the orogenic collapse of the Central Asian orogenic belt (Meng et al., 2011; W.L. Xu et al., 2013b). On the other hand, more recent studies, with evidence from the Heilongjiang Complex, reveal that an ancient ocean, namely, the Mudanjiang Ocean, existed between the Jiamusi and Songnen blocks in the Early Permian, and its subduction occurred during the Late Triassic–Early Jurassic (Ge et al., 2016; Wu et al., 2011; W.L. Xu et al., 2013b). Such a large ocean with a life span of more than 140 m.y. implies that the Mudanjiang Ocean was probably a branch of the Paleo–Pacific Ocean (Dong et al., 2017; Ge et al., 2017, 2018; Wu et al., 2011; Zhao et al., 2018; Zhu et al., 2017). Consequently, increasing numbers of studies propose that the granitoids in the LXZR were genetically related to the subduction of the Mudanjiang Ocean (Dong et al., 2017; Ge et al., 2017, 2018; Wu et al., 2011; Zhao et al., 2018; Zhu et al., 2017).

The controversial tectonic affinity of the LXZR granitoids further leads to ambiguity in the early Mesozoic tectonic model of the contacting Jiamusi block at a regional scale. Several tectonic models have been proposed for the Jiamusi block, including (1) postcollision extension of the southeastern Central Asian orogenic belt (Guo et al., 2015; Meng et al., 2011; Xu et al., 2009), (2) a back-arc extensional setting resulting from bipolar subduction of the paleo–Pacific plate beneath the Eurasian continent in the east and the Mongol-Okhotsk Ocean plate beneath the Erguna Massif in the north (M.J. Xu et al., 2013a; W.L. Xu et al., 2013b; Yu et al., 2012), as well as long-lasting westward subduction of the Mudanjiang Ocean beneath the Songnen block (Ge et al., 2016, 2017, 2018; Liu et al., 2017a; Zhu et al., 2017).

To further constrain the tectonic model of the Jiamusi block, we present whole-rock geochemistry, zircon U-Pb dating, and Lu-Hf isotope results of the granitoids from the LXZR. Our new data not only put new constraints on the age, source, and petrogenesis of the granitoids, but they also provide important insights into the tectonic processes related to the subduction of the Mudanjiang Ocean to form the LXZR between the Jiamusi and Songnen blocks.

GEOLOGICAL SETTING AND SAMPLE DESCRIPTIONS

Geological Background

NE China, located in the eastern part of the Central Asian orogenic belt (Fig. 1A), was likely formed by collision of multiple microcontinents. These microcontinents (also called blocks or terranes; e.g., the Erguna, Xing’an, Songnen, and Jiamusi blocks and the Nadanhada terrane from west to east) are currently separated by major faults (Fig. 1B; Ge et al., 2016; Wilde et al., 1997, 2003; W.L. Xu et al., 2013a; Zhou et al., 2009).

The Songnen block primarily is composed of the southern Great Xing’an Range, Songliao Basin, and LXZR (HBGMR, 1993; Wu et al., 2011). The southern Great Xing’an Range contains large volumes of Mesozoic volcanic rocks and granitoids (Wu et al., 2011), while the Songliao Basin (formed during the late Mesozoic) possesses a basement of Paleozoic–Mesozoic granites and Paleozoic volcanic strata with localized Proterozoic granites (Gao et al., 2007; Pei et al., 2007; Wu et al., 2001, 2011). The LXZR contain large volumes of late Paleozoic to early Mesozoic granitoids, with local occurrences of Paleozoic volcanic-sedimentary rocks (Fig. 2B; HBGMR, 1993; Meng et al., 2011; Wu et al., 2011). These granitoids mainly consist of alkali-feldspar granite, syenogranite, monzogranite, and granodiorite (Dong et al., 2017; Ge et al., 2017, 2018; Liu et al., 2017a; Wu et al., 2011; Zhu et al., 2017). These granitoids show a general north to south distribution with a widespread westward younging trend, with those in the eastern segment of the LXZR being mainly formed during the Late Permian to Early Triassic (Fig. 2A; Ge et al., 2017, 2018; Wei et al., 2012; Wu et al., 2011), whereas those in the west were mainly formed in the Late Triassic to Early Jurassic (Fig. 2A; Ge et al., 2017, 2018; Liu et al., 2017a; Wei et al., 2012; Wu et al., 2011). Recent studies also reported Early Paleozoic igneous rocks formed in an active continental margin setting (Wang et al., 2016, 2017a) and some highly fractionated I-type Jurassic granitoids in the LXZR (Wu et al., 2003).

The Jiamusi block is separated from the Songnen block to the west by the Mudanjiang-Yilan suture and from the Nadanhada terrane to the east by the Yuejinshan fault (Fig. 1B). The Jiamusi block is predominantly composed of the Mashan Group and two episodes of Paleozoic granitoids (Dong et al., 2017; Wilde et al., 1997, 2003; Wu et al., 2011; Yang et al., 2015). The Mashan Group contains series of Mesoproterozoic to Neoproterozoic khondalites that were metamorphosed to granulite facies at ca. 500 Ma (Wilde et al., 1997, 2003). Among the two episodes of Paleozoic granitoids, the early episode was mostly formed between 530 Ma and 515 Ma from the late Pan-African magmatism and experienced granulite-facies metamorphism at ca. 500 Ma (Wilde et al., 1997, 2003); the late episode was mostly formed between 270 Ma and 254 Ma, with weakly deformed to undeformed structure (Dong et al., 2017; Ge et al., 2017; Wu et al., 2011). Additionally, recent studies have identified some Neoproterozoic intrusive and sedimentary rocks indicating a ca. 560 Ma high-grade metamorphic event in the Jiamusi block (Luan et al., 2017; Yang et al., 2017), making the tectonic history of the Jiamusi block more complicated.

Located between the Jiamusi and Songnen blocks, the Heilongjiang Complex is primarily exposed in the Mudanjiang, Yilan, and Luobei areas along a rough north-to-south strike (Fig. 2B; HBGMR, 1993; Zhou et al., 2009). The Heilongjiang Complex is composed of ultramafic rocks, amphibolite, blueschist, greenschist, mica schist, quartzite, and marble, which is a rock assemblage similar to tectonic mélange. Thus, the Heilongjiang Complex likely represents the suture belt resulting from the closure of the Mudanjiang Ocean between the Jiamusi and Songnen blocks (Ge et al., 2016; Zhou et al., 2009). The Heilongjiang Complex underwent blueschist-facies metamorphism at 900–1100 MPa and 320–450 °C (Zhou et al., 2009). Blueschists, including both oceanic-island basalt (OIB)– and enriched mid-ocean-ridge basalt (E-MORB)–like sources, are the most significant components in the Heilongjiang Complex (Ge et al., 2016, 2017; Zhou et al., 2009; Zhu et al., 2015).

Sample Description

In this study, the samples were collected from six plutonic outcrops in the LXZR (Table 1), where one pluton (H15–10) was in the Tieli area (Fig. 3A) and the other five (H15–55, H15–56, H15–57, H15–58, and H15–59) were in the Sanzhancun area (Fig. 3B). In each pluton, one representative sample was selected for zircon U-Pb dating, and two or three samples were used for whole-rock major- and trace-element analyses.

Samples from the Tieli area (H15–10–01, H15–10–02, H15–10–03, H15–10–04) were predominately medium-grained syenogranites from a pluton previously mapped as Hercynian granite (HBGMR, 1993). The pluton was emplaced into the sandstone of the Late Permian Tumenling Group (Fig. 3A) and was later intruded by a mafic dike (Fig. 4A). The syenogranite samples contained quartz (31%–35% by volume), plagioclase (15%–19%), K-feldspar (42%–45%), biotite (4%–6%), and some accessory zircon and apatite (1%–3%; Fig. 4B).

Samples from the Sanzhancun area included syenogranites (H15–55–01, H15–55–03, H15–55–04, H15–55–05, H15–56–01, H15–56–02, and H15–56–03), tonalites (H15–57–01, H15–57–02, H15–57–03, and H15–57–04), granodiorites (H15–58–01, H15–58–03, and H15–58–04), and monzogranites (H15–59–01, H15–59–02, H15–59–03, and H15–59–04). Among these, the syenogranite samples showed medium-grained granitic texture, with weak gneissic to massive structures. Their mineral assemblage included quartz (21%–25%), plagioclase (18%–22%), K-feldspar (45%–50%), biotite (4%–7%), and some accessory zircon, apatite, and magnetite (2%–5%; Fig. 4C). The tonalite and granodiorite samples were both collected from the plutons previously determined as Hercynian granites (HBGMR, 1993). The tonalites showed a medium-grained granitic texture with massive structure and consisted of quartz (22%–24%), plagioclase (43%–46%), K-feldspar (8%–12%), amphibole (13%–16%), biotite (7%–8%), and minor (2%–4%) zircon, apatite, titanite, and chlorite (from alteration of biotite; Fig. 4D). Abundant mafic microgranular enclaves were preserved in the tonalite intrusion. The mafic microgranular enclaves showed round or elliptical shapes with centimeter to decimeter dimensions (Fig. 4E). The enclaves displayed a transitional contact with host rocks, and some contained plagioclase phenocrysts, which suggest the possible transfer of phenocrysts from host granitoid rocks to the mafic microgranular enclaves (Fig. 4E; Pietranik and Koepke, 2014). In contrast, the granodiorites showed porphyritic texture and massive structure. They contained substantial K-feldspar phenocrysts and were composed of quartz (30%–32%), plagioclase (34%–37%), K-feldspar (14%–17%), amphibole (5%–6%), biotite (7%–11%), and some accessary zircon, apatite, and titanite (3%–4%; Fig. 4F). Mafic microgranular enclaves were also observed in the granodiorite intrusion in the field outcrop (Fig. 4G). The monzogranites showed fine-grained granitic texture (Fig. 4H) and consisted of quartz (21%–25%), plagioclase (22%–28%), K-feldspar (44%–47%), and biotite (3%–4%) with minor zircon and apatite (2%–4%).

ANALYTICAL METHODS

Whole-Rock Major and Trace Elements

Samples were first washed and trimmed to remove altered surfaces. Fresh portions were selected and crushed to less than 200 mesh in an agate mill for whole-rock geochemical analyses. Analytical procedures for whole-rock major and trace elements have been described in detail in Ge et al. (2016). In brief, major elements were determined by an automatic X-ray fluorescence (XRF) spectrometer at the Institute of Geology and Geophysics, Chinese Academy of Sciences. International rock standards BCR-1, BCR-3, and repeated analysis of one of every 10 samples were used for quality control. The analytical precision was better than 3% (relative). Trace elements, including rare earth elements (REEs), were determined using an Agilent 7500ce inductively coupled plasma–mass spectrometer (ICP-MS) at Peking University. International standards, including GSR-1, GSR-3, GSR-10, and DZå-1, were used for quality control. The analytical precision was better than 5% (relative).

Zircon U-Pb Laser-Ablation ICP-MS Dating

Zircon grains were extracted from pulverized rock samples using combined heavy liquid and magnetic techniques and were further purified by handpicking under a binocular microscope. For each sample, over 200 grains were cast in an epoxy mount and polished to expose the grain centers. Prior to analyses, cathodoluminescence (CL) images were taken by a Quanta 200 FEG scanning electron microscope at Peking University to understand the internal structures and guide the selection of spots for U-Pb dating.

Zircon analyses of U-Th-Pb isotopes were carried out using an Agilent 7500ce ICP-MS equipped with a GeoLas 193 nm laser-ablation (LA) system at the Key Laboratory of Orogenic Belts and Crustal Evolution, Peking University. The laser spot was set to 32 µm in diameter. Zircon 91500 (ca. 1064 Ma) and Plešovice (ca. 337 Ma) were used for quality control of zircon U-Pb isotope data. NIST 610, NIST 612, and NIST 614 were used as standards for trace element (U, Th, and Pb) analyses. The 207Pb/206Pb, 206Pb/238U, and 207Pb/235U ratios were calculated using the GLITTER program (Van et al., 2001), and common Pb was corrected following Andersen (2002). The U-Pb ages and concordia diagrams were obtained using the Isoplot 3.0 program (Ludwig, 2003).

Zircon Lu-Hf LA-ICP-MS Analyses

Zircon Lu-Hf isotope compositions were measured by a 193 nm LA system coupled with a Neptune Plus multicollector (MC) ICP-MS at the Institute of Geology, Chinese Academy of Geological Science. The laser spot size was 44 μm in diameter for most of the samples, but it was reduced to 32 μm for one sample with smaller grains. Helium was used as a carrier gas to transport the ablated samples to the ICP-MS torch. Zircon 91500 was used to monitor the stability and reliability of the instrument. Detailed descriptions of analytical procedures and calculations have been provided by Hou et al. (2007) and Wu et al. (2006).

RESULTS

Zircon Morphology

The CL images of representative zircons are shown in Figure 5. All zircon grains have similar morphological characteristics, defined by subhedral short prisms or euhedral columnar shapes, with lengths between 60 and 200 μm and length/width ratios of 1:1–4:1. The grains have clear oscillatory zoning structure in the CL images. These features, integrated with the contents of Th (15.84–1246.63 ppm) and U (30.63–3206.8 ppm), and Th/U ratios (0.2–1.7; Table DR1 in the GSA Data Repository1), indicate that the zircons were magmatic in origin (Corfu et al., 2003; Wu and Zheng, 2004).

Zircon Geochronology

The U-Pb isotopic results of all analyzed grains are listed in Table DR1 (see footnote 1), but only the grains with concordant ages are plotted in Figure 6. The results of each sample are briefly described below.

Sample H15–10–01 (syenogranite from the Taoshanzhen pluton; Fig. 3A): Twenty-three of the 25 valid analyses yielded concordant ages, which defined a weighted mean 206Pb/238U age of 190 ± 2 Ma (mean square of weighted deviates [MSWD] = 3.5). This age is considered to represent the crystallization age of this syenogranite (Fig. 6A). The other two older ages (212 ± 2 Ma and 227 ± 2 Ma, respectively) are considered to be ages of inherited/captured zircons.

Sample H15–55–01 (syenogranite from ∼25 km southeast of Fengshanzhen; Fig. 3B): Twenty-two of the 25 analyses gave consistent concordant ages with a weighted mean 206Pb/238U age of 196 ± 1 Ma (MSWD = 0.60), which is considered to be the crystallization age (Fig. 6B). The other three analyses yielded apparent 206Pb/238U ages of 211 ± 2 Ma (n = 2) and 237 ± 3 Ma, respectively, which are considered to be ages of inherited/captured zircons.

Sample H15–56–01 (syenogranite from ∼5 km northwest of Fengshanzhen; Fig. 3B): Twenty-five zircon grains were analyzed, from which two analyses fell below the concordia curve, indicating possible Pb loss (Fig. 6C). Three grains captured older apparent 206Pb/238U ages of 218 ± 2 Ma, 228 ± 2 Ma, and 248 ± 3 Ma. The remaining 20 analyses yielded a weighted mean 206Pb/238U age of 196 ± 1 Ma (MSWD = 0.19), which represents the crystallization age of this sample.

Sample H15–57–01 (tonalite from ∼55 km north of Fengshanzhen; Fig. 3B): Twenty-three valid 206Pb/238U ages obtained from 25 analyses defined a weighted mean 206Pb/238U age of 186 ± 1 Ma (MSWD = 0.23), representing the crystallization age of the tonalite.

Sample H15–58–01 (granodiorite from ∼50 km northwest of Yuanmodingzi; Fig. 3B): The 206Pb/238U ages from all 25 analyses yielded a weighted mean age of 181 ± 1 Ma (MSWD = 0.20; Fig. 6E). This age is considered to be the crystallization age of the granodiorite.

Sample H15–59–01 (monzogranite from ∼55 km north of Yuanmodingzi; Fig. 3B): Twenty-four valid 206Pb/238U ages out of 25 analyses defined a weighted mean age of 182 ± 1 Ma (MSWD = 5.2; Fig. 6F). This age is considered to represent the crystallization age of the monzogranite.

Zircon Lu-Hf Isotopes

Ten representative zircon grains from each of the six dated samples were chosen for in situ Lu-Hf isotopic analysis. The results are listed in Table DR2 (see footnote 1) and plotted in Figure 7.

Zircons in syenogranites (H15–10–1, H15–55–1, and H15–56–1) had initial 176Hf/177Hf values of 0.282687–0.282864, with ɛHf(t) from + 1.2 to + 7.6 and depleted mantle model ages (TDM2) ages from 753 to 1158 Ma. Zircons in tonalites and granodiorites (H15–57–1 and H15–58–1) had initial 176Hf/177Hf values of 0.282771–0.282827, with ɛHf(t) from + 3.9 to + 6.0 and TDM2 ages from 844 to 975 Ma. Zircons in monzogranites (H15–59–1) has initial 176Hf/177Hf values of 0.282736–0.282795, with ɛHf(t) from + 2.7 to + 4.8 and TDM2 ages from 919 to 1053 Ma.

Whole-Rock Geochemistry

The major- and trace-element data for 16 granitoid samples are listed in Table DR3 (see footnote 1). These geochemically variable samples can be divided into two groups based on the total alkali-silica (TAS) classification, the chondrite-normalized REE patterns, and the primitive mantle–normalized multi-element patterns (Figs. 8 and 9).

Group 1 consists of tonalite, granodiorite, and monzogranite, which contain 62.58–74.64 wt% SiO2 and 6.46–8.46 wt% Na2O + K2O and fall across the tonalite, granodiorite, and granite fields of the TAS diagram (Fig. 8A). These samples also contain 13.56–16.27 wt% Al2O3, 0.67–4.69 wt% CaO, 0.21–0.81 wt% TiO2, 0.06–0.25 wt% P2O5, 2.65–4.41 wt% K2O, and 3.79–4.34 wt% Na2O, belonging to high-K calc-alkaline rock series in the SiO2 versus K2O diagram (Fig. 8B). Their Al2O3/(CaO + Na2O + K2O) ratios, abbreviated A/CNK, range from 0.91 to 1.07, indicating metaluminous to weakly peraluminous composition (Fig. 8C), while their MgO and TFe2O3 contents range from 0.32 to 2.50 wt% and 1.36–5.76 wt%, respectively, corresponding to Mg# values of 35.4–50.3 (Figs. 8D–8F). These samples show enrichments in large ion lithophile elements (LILEs; e.g., Ba, Rb, Th, U, and K) and light rare earth elements (LREEs), but depletions in high field strength elements (HFSEs; e.g., Nb, Ta, Ti, and P) and heavy rare earth elements (HREEs), with weakly negative to no Eu anomalies: (La/Yb)N = 11.55–22.89 and Eu/Eu* = 0.81–1.06 (Figs. 9A and 9B).

Group 2, composed explicitly of syenogranites, shows higher SiO2 contents of 75.05–80.0 wt% and Na2O + K2O contents of 7.22–9.55 wt% and plots in the granite field (Fig. 8A) and high-K calc-alkaline rock series (Fig. 8B). Compared with group 1, these group 2 syenogranites have lower Al2O3 (10.64–14.45 wt%), CaO (0.09–0.64 wt%), TiO2 (0.10–0.24 wt%), P2O5 (0.01–0.06 wt%), MgO (0.07–0.17 wt%), TFe2O3 (0.60–1.37 wt%), and Mg# (15.2–28.4) values (Figs. 8D–8G), with weakly peraluminous features (A/CNK = 1.01–1.07; Fig. 8C). Group 2 samples show a relatively flat REE pattern with strongly negative Eu anomalies, e.g., (La/Yb)N = 5.96–8.59 and Eu/Eu* = 0.23–0.60 (Fig. 9C), positive Rb, Th, U, and K anomalies, negative Nb-Ta anomalies, and strongly negative Sr, P, and Ti anomalies (Fig. 9D).

DISCUSSION

Ages of the Granitoid Rocks from the LXZR

Previous studies, mainly based on litho-stratigraphic relationships with country rocks or whole-rock K-Ar and Rb-Sr ages (HBGMR, 1993), suggested that the widely distributed granitoids (including the granitic intrusions studied here) throughout the LXZR were early Paleozoic in age. However, recent high-precision zircon U-Pb studies revealed that some intrusions previously classified as early Paleozoic were actually generated during the late Paleozoic to early Mesozoic (Ge et al., 2017; Wei et al., 2012; Wu et al., 2011). Our new results here provide further evidence that the ages of some Paleozoic rocks in the LXZR have been misclassified.

The six studied granitic plutons have crystallization ages in a limited range from 196 to 181 Ma, indicating that Early Jurassic magmatism occurred in the LXZR. Moreover, abundant coeval intrusive and volcanic rocks have also been reported in the LXZR (Fig. 10; Table DR4 [see footnote 1]; Ge et al., 2017, 2018; Gou et al., 2013; Guo et al., 2018; Hu et al., 2014; Qin et al., 2016; Tang et al., 2011; M.J. Xu et al., 2013a; W.L. Xu et al., 2013b; Wu et al., 2011; Yang and Wang, 2010; Yu et al., 2012; Zhu et al., 2017). These Early Jurassic igneous rocks vary from mafic to felsic in composition and are distributed in a roughly N–S direction in the LXZR (Fig. 2B; Wu et al., 2011; Yu et al., 2012; W.L. Xu et al., 2013b). All these factors suggest that Early Jurassic (rather than early Paleozoic) magmatism was widespread in the LXZR.

Petrogenesis

Group 1 granitoids, consisting of tonalite, granodiorite, and monzogranite, are I-type granitoids, as suggested by their relatively low A/CNK values (<1.1), negative correlation between P2O5 and SiO2, and positive correlation between thorium and rubidium (Figs. 8G and 8J; Chappell, 1999; Li et al., 2007). The mineral assemblage of hornblende, biotite, and titanite in this group is typical in I-type granites (Barbarin, 1999). The common geochemical signatures in this group, such as medium to high SiO2 (62.68–74.64 wt%) and low MgO (0.32–2.50 wt%) contents, are consistent with experimental results derived from partial melting of either pure crustal materials or basalts and amphibolites (Fig. 8D; Martin et al., 2005). Although these granitoids exhibit linear correlations between SiO2 and some major and trace elements on the Harker diagrams (Fig. 8), fractional crystallization processes may not have been predominant due to the limited occurrence of coeval mafic and intermediate igneous rocks in the field (Gao et al., 2016). Moreover, these granitoids have relatively high Eu/*Eu values (0.81–1.06) and no obvious positive correlations between Sr and Eu/*Eu ratios (Fig. 8I), suggesting that the fractional crystallization of plagioclase was not significant (Hu et al., 2016, 2017). Thus, the observed geochemical features of the granitoid rocks were more likely controlled by partial melting processes.

Multiple source materials could have been responsible for the group 1 granitoids. The monzogranites exhibit higher Sr/Y and (La/Yb)N ratios (29.9–35.1 and 21.4–22.9, respectively) and lower Yb and Y contents (0.83–0.95 ppm and 7.47–9.22 ppm, respectively) than those of tonalites and granodiorites (Sr/Y = 17.8–30.6 and [La/Yb]N = 11.6–12.8; Yb = 1.68–2.48 ppm and Y = 14.1–21.7 ppm; Figs. 11A and 11B; Martin et al., 2005). In addition, in the AMF (molar Al2O3/[FeOt + MgO]) versus CMF (molar CaO/[FeOt + MgO]) diagram, the tonalite and granodiorite samples lie in the field of partial melting of metabasaltic to metatonalitic sources, while the monzogranite samples fall in the field of partial melting of metagraywackes (Fig. 12A; Altherr et al., 2000). Similarly, the tonalite and granodiorite samples all fall in the field of partial melting of amphibolites in the CaO + FeOt + MgO + TiO2 versus CaO/(FeOt + MgO + TiO2) diagram, while the monzogranite samples all fall in the field of partial melting of metagraywackes (Fig. 12B; Patiño Douce, 1999). Experimental studies have shown that partial melting of basaltic rocks can generate melts with relatively low Mg# values (<40), irrespective of the degree of partial melting (Rapp and Watson, 1995). Thus, variably higher Mg# (42.8–50.3) values of the tonalite and granodiorite samples suggest the involvement of various (but generally small) amounts of mantle-derived material, which has a high Mg# value of >72 (Ramsay et al., 1984). This is also supported by the occurrence of mafic microgranular enclaves within the tonalite and granodiorite outcrops (Figs. 4E and 4G). Nevertheless, the relatively small ranges of ɛHf(t) values (+3.9 to +6.0) and TDM2 ages (844–975 Ma) of the magmatic zircons suggest that the tonalites and granodiorites were mainly derived from juvenile crustal materials, which could have been Mesoproterozoic to Neoproterozoic crustal materials with minor input of mantle-derived materials.

Group 2 granitoids, all composed of syeno-granite, have higher contents of SiO2 and Na2O + K2O and exhibit significant negative Ba, Nb, Sr, Eu, Ti, and P anomalies, similar to the highly fractionated I-type granites widespread in NE China (Fig. 9D; Wu et al., 2003). A highly fractionated I-type origin for the syenogranite was also supported by their high TFeO/MgO ratios and (Zr + Nb + Ce + Y) values (Fig. 8L; Whalen et al., 1987; Wu et al., 2003). As such, non-negligible fractional crystallization should have taken place during the formation of these granitoids. For example, a negative P anomaly suggests fractionation of apatite; negative Nb-Ta-Ti anomalies require separation of Ti-bearing phases (e.g., titanite and/or ilmenite); strong Eu depletion indicates fractionation of plagioclase; in the Ba versus Sr diagram (Fig. 8K), a coupled decrease in both Sr and Ba contents indicates fractionation of plagioclase and K-feldspar. In addition, the MgO and Mg# values of these highly fractionated I-type granites are much lower than the group 1 rocks, suggesting that fractionation of mafic minerals largely influenced their major elements (Fig. 8D; Martin et al., 2005). Compared with group 1 rocks, the group 2 granitoids have lower (La/Yb)N and Sr/Y ratios (5.96–8.59 and 0.95–4.96, respectively) and higher Yb and Y values (1.92–3.70 ppm and 13.2–31.4 ppm, respectively; Figs. 11A and 11B), again reflecting the influence of fractionation of feldspar (Hu et al., 2018; Martin et al., 2005; Yang et al., 2015). In the AMF versus CMF diagram (Fig. 12A), the syenogranite samples correspond to metapelitic sources. Consistently, the CaO + FeOt + MgO + TiO2 versus CaO/(FeOt + MgO + TiO2) diagram (Fig. 12B) indicates the samples could have originated from felsic pelites. In fact, these geochemical features can also be explained by high fractionation of feldspars (Hu et al., 2018). The chemical evolutionary trends of these samples (Fig. 8) suggest that they likely underwent fractional crystallization from a melt similar to the monzogranite in group 1, which is further supported by their similar ɛHf(t) values (+1.2 to +7.6) and TDM2 ages (753–1158 Ma) to those of monzogranites (Fig. 7). Accordingly, we conclude that the group 2 syenogranites were mainly derived from partial melting of juvenile Mesoproterozoic to Neoproterozoic crustal materials that underwent a relatively high degree of fractional crystallization.

TECTONIC SETTING AND IMPLICATIONS

Our results provide strong evidence that early Mesozoic granitoids are widespread in the LXZR in NE China. This has important implications for the regional tectonic model. Previous studies have proposed that the granitoids in the LXZR were a result of delamination during the postcollision stage of the eastern Central Asian orogenic belt following the closure of the Paleo-Asian Ocean (Meng et al., 2011; Xu et al., 2009, 2013; Yu et al., 2012). However, increasing numbers of studies suggest that the closure of the Paleo-Asian Ocean and the final collision of the Central Asian orogenic belt in this area occurred before the late Paleozoic, recorded by a roughly west-east magmatic belt along the Changchun-Yanji suture (Li et al., 2014; Wu et al., 2007a; Xu et al., 2014; Zhao et al., 2013). These contrasting results make it difficult to explain the formation of the early Mesozoic granitic belt along a south-north strike in the LXZR. Here, we propose that the Early Jurassic granitoids in the LXZR were likely formed in an active continental margin setting related to the subduction of the Mudanjiang Ocean. This proposal is supported by multiple lines of evidence, including early Mesozoic magmatism, metamorphism, metallogeny, and structure, as detailed below.

(1) Magmatic evidence: The early Mesozoic granitoids with ages from 250 to 160 Ma are distributed along a nearly north–south trend in the LXZR (Figs. 2 and 13A). They are mainly I-type granitic rocks and belong to high-K calc-alkaline series with enrichments in LILEs (e.g., Ba, Th, and U) and LREEs and depletions in HFSEs (e.g., Nb and Ta) and HREEs (Ge et al., 2017, 2018; Liu et al., 2017a; Zhu et al., 2017; Zhao et al., 2018), which are similar to continental arc–type granitic rocks (Wang et al., 2018). Tectonic setting discrimination (Fig. 11; Drummond and Defant, 1990; Pearce et al., 1984) indicates that these granitoids have volcanic arc field (VAG) affinity, which is consistent with an active continental margin setting. Additionally, the coeval mafic-ultramafic intrusions (e.g., hornblendite, gabbro, and gabbro-diorite) and volcanic rocks from the LXZR also show arc geochemical features of LILE (Ba, K, and Sr) enrichment and HFSE (Nb, Ta, Zr, and Hf) depletion (Wang et al., 2015; Yu et al., 2012). These signatures imply that their magma sources could have originated from partial melting of a depleted mantle wedge metasomatized by slab-derived melts during the subduction of the Mudanjiang Ocean between the Jiamusi and Songnen blocks (Yu et al., 2012).

(2) Metamorphic evidence: The Heilongjiang blueschist belt, which is distributed along a nearly north–south direction subparallel to the LXZR granitic belt studied here, is a strong evidence indicative of subduction and final closure of the Mudanjiang Ocean between the Jiamusi and Songnen blocks. The blueschists show either an OIB or E-MORB affinity. The protolith ages from 288 to 186 Ma of the blueschists (Fig. 13B; Ge et al., 2016; Zhou et al., 2009, 2013; Zhu et al., 2015) suggest that the Mudanjiang Ocean existed between the Jiamusi and Songnen blocks during the Early Permian to Early Jurassic. Moreover, published Ar-Ar metamorphic ages (202–145 Ma; Fig. 13B) of the Heilongjiang blueschist suggest that westward subduction of the Mudanjiang Ocean beneath the Songnen block occurred during the latest Triassic to Late Jurassic (Ge et al., 2017; Wu et al., 2007b; Zhao and Zhang, 2011; Zhou et al., 2013; Zhou and Li, 2017). These data exclude the possibility of an early Mesozoic postcollisional delamination of the Central Asian orogenic belt (Liu et al., 2017a).

(3) Metallogenic evidence: Large numbers of porphyry copper and molybdenum deposits occur in the LXZR. Geochronological dating, including Re-Os isochron ages from molybdenite and zircon U-Pb concordant ages from ore-bearing granitoids, has yielded Early–Middle Jurassic ages (197–161 Ma; Fig. 13C; Chen et al., 2019; Guo et al., 2018; Hou et al., 2018; Tang et al., 2011; Wang et al., 2017b, and references therein), coinciding with the magmatic activity in the LXZR. Furthermore, the LXZR metallogenic belt also displays a roughly north–south distribution (Fig. 13C), consistent with the subduction direction of the Mudanjiang Ocean. Thus, combined with the regional geological history, these porphyry copper and molybdenum deposits were likely formed in an active continental margin setting, related to the subduction of the Mudanjiang Ocean beneath the Songnen block during the early Mesozoic (Chen et al., 2019; Zhang et al., 2013).

(4) Structural evidence: Deformation structures are very well developed in the LXZR, and they provide insights into related orogenic processes. Abundant faults associated with drag folds in the LXZR have strikes of NNE 20° to 40° (HBGMR, 1993; Shao et al., 2013). Shao et al. (2013) found a large sinistral ductile shear zone with a strike similar to the main fault in this area, as well as many asymmetric folds and ocular structures. Combined with detailed geochronological studies, Shao et al. (2013) proposed that these structures were caused by oblique shear compression during the subduction of the Mudanjiang Ocean in the Early to Middle Jurassic. This is consistent with the oblique subduction of the mid-oceanic ridge between the Farallon and Izanagi plates toward Eurasia in the Late Triassic to Early Jurassic (Maruyama et al., 1997). In addition, a deep seismic reflection profile from Suihua to Hulin Country in NE China found a dipping reflection in the upper mantle under the eastern edge of the Songliao Basin, which corresponds to a low-angle westward sunk oceanic slab between the LXZR and the Jiamusi block (Wang, 2011).

In the eastern part of the Jiamusi block, previous studies have shown the early Mesozoic influence of the Paleo–Pacific Ocean in NE China. For example, Zhou et al. (2014) proposed that the accretion of the Yuejinshan Complex along the eastern margin of the Jiamusi block probably occurred between 210 and 180 Ma, suggesting that the subduction of the paleo–Pacific plate took place in the Late Triassic to Early Jurassic. This is also supported by the occurrence of north–south distributed arc-type calc-alkaline volcanic rocks with Early Jurassic ages (187–174 Ma) in the Raohe Complex (Wang et al., 2017c).

Based on these observations and discussions, we propose that, during the early Mesozoic, the Jiamusi block was located within the paleo-Pacific realm, not influenced by the Paleo-Asian realm. In the Early Jurassic (Fig. 14), the paleo–Pacific plate was subducted along the eastern margin of the Jiamusi block and produced the Yuejinshan Complex, while the Mudanjiang Ocean was subducted westward beneath the Songnen block and produced the Heilongjiang Complex. The subduction of the paleo–Pacific plate and Mudanjiang Ocean further induced partial melting beneath the Jiamusi and Songnen blocks, respectively, resulting in the emplacement of voluminous igneous rocks along these active continental margins.

CONCLUSIONS

(1) LA-ICP-MS zircon U-Pb isotopic dating revealed that the granitoids from the LXZR have crystallization ages ranging from 196 to 181 Ma.

(2) These Early Jurassic granitoids are all I-type granitoids and can be divided into two groups: Group 1 is composed of the tonalite, granodiorite, and monzogranite samples, and group 2 is composed of syenogranites with highly fractionated I-type characteristics.

(3) The magmas of the Early Jurassic granitoids were mainly derived from partial melting of three distinct Mesoproterozoic to Neoproterozoic source materials: (i) basaltic rocks with a minor addition of mantle component for group 1 tonalites and granodiorites; (ii) metagraywackes for group 1 monzogranites; and (iii) metapelites for group 2 syenogranites.

(4) The studied granitoids are high-K calc-alkaline series with arc geochemical features, and thus they were likely generated in an active continental margin setting related to the westward subduction of the Mudanjiang Ocean between the Jiamusi and Songnen blocks during the Early Jurassic.

ACKNOWLEDGMENTS

We thank Libing Gu and Fang Ma (Peking University) for help with trace element and zircon U-Th-Pb isotopic analyses, Zheng Wang (Chinese Academy of Geological Sciences) for help with zircon Lu-Hf isotopic analyses, and Hongyue Wang (Chinese Academy of Sciences) for help with the whole-rock major-element analyses. Max Lukenbach is acknowledged for improving the English of this manuscript. The manuscript benefited from constructive comments from Editor Damian Nance, Wenjiao Xiao, and two anonymous reviewers. This research was financially supported by the National Key Research and Development Project of China (grant number 2017YFC0601301), the National Natural Science Foundation of China (grant numbers 41730210 and 41830216), and Project of China Geological Survey (grant number DD20190004).

1GSA Data Repository Item 2019406, Table DR1: LA-ICP-MS U-Pb data for the early Mesozoic granitoids in the Lesser Xing’an–Zhangguangcai Ranges, NE China; Table DR2: Hf isotopic data of zircons extracted from the early Mesozoic granitoids from the Lesser Xing’an–Zhangguangcai Ranges, NE China; Table DR3: Major- and trace-element compositions of the early Mesozoic granitoids from the Lesser Xing’an–Zhangguangcai Ranges, NE China; Table DR4: Geochronological data for the Early Jurassic igneous rocks in the Lesser Xing’an–Zhangguangcai Ranges, is available at http://www.geosociety.org/datarepository/2019, or on request from editing@geosociety.org.
Gold Open Access: This paper is published under the terms of the CC-BY-NC license.