The origin and tectonic regime responsible for the inland Jurassic granites in Southeast (SE) China remain controversial. This study presents zircon secondary ion mass spectrometry (SIMS) U-Pb ages, in situ zircon Hf-O isotopes, and whole-rock geochemical and Sr-Nd isotopic data for the Fogang and Xinxing Batholiths in central Guangdong. Mineralogical and geochemical features indicate that these granites are high-K (>4.8 wt% K2O at 72 wt% SiO2), calc-alkaline I-type granites. SIMS U-Pb analyses on magmatic zircons yield consistent ages ranging from 158 to 163 Ma, suggesting that the Fogang and Xinxing granites were emplaced in the period of 163–158 Ma. In addition, these granites have whole-rock initial Sr87/Sr86 ratios of 0.6802–0.7072 and negative εNd(t) values of −9.5 to −8.2, zircon negative εHf(t) values of −12.34 to −0.56, and high δ18O values of 7.64‰–10.08‰. The above features imply that the granites were most likely generated through the mixture of supracrustal sedimentary components with minor addition of mantle-derived magmas. Granites from the Fogang and Xinxing Batholiths in SE China should be derived from the Proterozoic crustal reworking due to asthenosphere upwelling or underplating and intrusion of mafic magmas. These Jurassic granites reflect anorogenic magmatism probably formed in an intraplate extensional setting resulted from the foundering of the flat slab beneath SE China.

Granite is a primary component of continental crust, preserving abundant information about the formation, evolution, and accretion of crust, as well as interactions between the crust and mantle. Multiperiod Mesozoic granites are widely distributed in Southeast (SE) China, with a concentration in the Triassic, Jurassic, and Cretaceous, respectively Figure 1(a) [1-3]. Among them, the Nanling region is mainly characterized by Jurassic granite, while the coastal areas are dominated by Cretaceous granite (see Figures 1(a) and 1(b)) [4]. The coexistence of multiperiod rocks from different origins is of great significance for understanding the genesis of the granite, crust-mantle interaction, magma differentiation, and mixing processes [5-8]. Previous researchers have reported the geochronology, petrology, mineralogy, and geochemistry of the Nanling granites. However, there has been ongoing debate regarding their petrogenesis and tectonic mechanism.

The Fogang and Xinxing Batholiths represent Late Mesozoic basements in the Nanling region, with Fogang Batholith being the largest and most representative granite basement in the region (Figures 1(c) and 1(d)) [4, 7]. Due to intense fractional crystallization, the batholiths exhibit complex geochemical characteristics, making their genetic types difficult to determine [9]. Different scholars have classified the Fogang Batholith as I-type [8, 9], A-type [5], S-type [6], or high-fractionated I-type granites [10]. Similarly, there are different views on the genetic type of the Xinxing Batholith, such as I-type [11], A-type [12], or S-type [13]. Thus, it is necessary to further determine the nature of these granites; the results can provide valuable insights on establishing the classification criteria for the petrogenesis of these granites [14].

Furthermore, various models have been proposed over the past few decades regarding the tectonic mechanism of Jurassic granites in SE China: (1) the subduction of the Paleo-Pacific plate beneath the Eurasia plate [3, 15]; (2) the intraplate extension and/or rifting regime [1, 16]; and (3) the flat-slab subduction and slab-foundering model [12, 17]. Accurately understanding the petrogenesis of these granites is of great significance for understanding the Mesozoic tectonic history of SE China.

To restrict the petrogenesis and tectonic significance of Jurassic granites in SE China, this study systematically conducted zircon U-Pb geochronology and Hf-O isotope, whole-rock geochemical, and Sr-Nd studies on Jurassic granites in the Fogang and Xinxing Batholiths, further providing new insights on Late Mesozoic geodynamics in SE China.

The Cathaysia Block is located in SE China, which collided with the Yangtze Block along the Jiangnan belt to form the South China Block in the Neoproterozoic (Figure 1(a)) [18-22]. The Cathaysia Block consists of a Precambrian basement and a Sinian to Mesozoic sedimentary and volcanic cover [23]. Of those, the Precambrian basement rocks with protolith ages ranging from Paleoproterozoic to Mesoproterozoic are mainly outcropped beneath Mesozoic volcanic rocks in the northeastern region [24]. Since the Neoproterozoic, extensive granitic intrusions and contemporaneous volcanic rocks have developed in the southeastern Cathaysia Block (Figure 1(b)). Notably, there are three age peaks of granitoid magmatism during the Mesozoic, concentrating at Triassic, Jurassic, and Cretaceous, respectively [4, 25, 26]. Among them, Jurassic granitoids occupy 22% of the entire exposed granitoids in SE China (Figure 1(b)) [26].

The Triassic (250–200 Ma) granitoid-volcanic rocks, composed primarily of monzonite, syenogranite, and granodiorite, have a widespread outcrop range and exhibit a planar distribution pattern in the current inland areas of SE China [9]. In contrast, the volcanic rocks are mainly rhyolite and are less extensively exposed [26]. The Jurassic (200–140 Ma) granitoids and volcanic rocks are distributed in the present inland of SE China. They consist mainly of monzonite, syenogranite, and granodiorite, associated with the coeval volcanic rocks of rhyolite and tuff [9, 26]. The Cretaceous (140–80 Ma) granitoids and volcanic rocks are predominantly found in the current southeastern coastal areas and the Lower Yangtze region, with a total exposed area of 120,000 km2 [26]. Of those, volcanic rocks are widespread, primarily consisting of rhyolites; while monzonite, syenite, and granodiorite remain the dominant rock types for contemporaneous granitoids [9, 26].

The Fogang Batholith (Figure 1(c)) is the largest and most representative granitoid batholith in central Guangdong Province of SE China, with a total exposed area of over 6000 km2 [9, 27]. It consists of coarse-grained biotite monzogranite and K-feldspar syenogranite, with minor subordinate granodiorite, fine-grained biotite granite, two-mica granite, and granite porphyry [14, 28]. The Fogang Batholith intruded into the late Neoproterozoic to early Jurassic sedimentary rocks and was in turn intruded by Cretaceous syenite [29], without deformation and metamorphism. Previous studies showed that the Fogang Batholith was emplaced during the period of 158–168 Ma [5, 6, 14, 29]. However, the genetic type of these granitoid batholith remains controversial, including the S-type [6], I-type [8], and aluminous A-type [5]. The Xinxing Batholith (Figure 1(d)) is a large granitic pluton (ca. 1500 km2) in the central and southern Guangdong Province of SE China. It parallels to the regional NE-trending faults and is generally considered to have been generated in late Mesozoic [4]. The Xinxing pluton consists of alkali-feldspar granite, syenogranite, monzogranite, granodiorite, and biotite granite, which intruded into Cambrian and Devonian sedimentary rocks [11, 13, 30]. Previous studies revealed that the genetic type and timing of the Xinxing Batholith remain controversial, including S-type [13], I-type [12], and A-type [11] from Triassic (~230–240 Ma) to Jurassic (~160 Ma) [12, 13].

A total of thirty representative samples were collected from the Fogang Batholith and Xinxing pluton, respectively. Of those, twenty samples (samples 23CH01-1~23CH01-10 and 23CH02-1~23CH02-10) were collected from two different plutons of the Fogang Batholith, located at the Shunxing quarry in Conghua (Figures 1(c) and 2(a)). Samples 23CH01-1 to 23CH01-10 are gray-white medium-grained biotite granites with a massive structure and porphyraceus texture and consist of biotite (5%), plagioclase (25%), alkali feldspar (35%), quartz (34%), and minor muscovite (1%; Figure 2(b)). The phenocrysts are biotite, alkali feldspar, plagioclase, and quartz (Figures 3(a) and 3(b)). Samples 23CH02-1 to 23CH02-10 are red coarse-grained biotite granites with a massive structure and porphyraceus texture and are composed of biotite (5%), plagioclase (25%), alkali feldspar (35%), quartz (34%), and minor muscovite (1%; Figure 2(c)). The phenocrysts are alkali feldspar (Figure 2(c)), biotite (Figures 3(c) and 3(d)), plagioclase, and quartz. Though these samples from two different plutons of the Fogang Batholith display different specimen colors and grain sizes, they exhibit similar mineral compositions under a microscope. In addition, ten samples 23JM03-01 to 23JM03-10 were collected from the Xinxing pluton, located at the Taisheng quarry in Jiangmen (Figures 1(d) and 2(d)). These samples are medium- to coarse-grained biotite granites (Figure 2(e)) with a massive structure and equigranular texture and are locally intruded by red fine-grained moyite vein (Figure 2(f)). These granites consist of biotite (15%), alkali feldspar (45%), plagioclase (20%), and quartz (30%; Figures 3(e) and 3(f)), showing higher contents of alkali feldspar and biotite than those from the Fogang Batholith.

Whole-rock major compositions were determined by X-ray fluorescence (XRF) spectrometry using a ZSX Primus II XRF spectrometer. FeO contents were determined by Fe2+ titration, and Fe2O3 contents were calculated by difference. Trace elements were analyzed using an Agilent 7900e inductively coupled plasma mass spectrometer, following the detailed analytical procedure described by Gao et al. [31]. The measured whole-rock compositions and trace elements of the samples are listed in online supplementary Table S1 and S2.

Three samples 23CH01-1, 23CH02-1, and 23JM03-1 were selected for zircon U-Pb dating using CAMECA IMS-1280 HR SIMS (secondary ion mass spectrometry) at the State Key Laboratory of Isotope Geochemistry, Guangzhou Institute of Geochemistry, Chinese Academy of Sciences. The ellipsoidal spot is about 20 × 30 μm in size. Standard zircons of Plěsovice and Qinghu were used for calibrating U-Th contents and Pb/U ratios. The results are listed in online supplementary Table S3. Data processing was carried out using the Isoplot 4.15 program [32].

Zircon O isotope analysis was conducted using the CAMECA IMS 1300-HR3 at Nanjing University, China. A primary ion beam of 133Cs+ with a current of 2–3 nA and a total impact energy of 20 keV was focused on the surface of the sample. A 5 × 5 μm raster was used for the study, and charge compensation was achieved using a normal-incidence electron gun. The total analytical time was about 4.5 minutes. The results are listed in online supplementary Table S3.

Zircon Hf isotope analysis was conducted on the same zircon spots that were previously analyzed for SIMS U-Pb ages and O isotope at the Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai). This analysis was carried out using the Neptune XT/Thermo Scientific, which was equipped with an NWR FemtoUC/Elemental Scientific Lasers. The spot size used for analysis was 50 µm, with a repetition rate of 5 Hz and a laser power of ~3 J/cm2. Zircon 91500, Penglai, and Plesovice samples were used as the reference standard in this study. The analytical procedures, calibration method, and referred standard were described by Wu et al. [33] and Li et al. [34]. The measured 176Hf/177Hf ratios were normalized to a value of 179Hf/177Hf = 0.7325. To calculate εHf values, the decay constant value for 176Lu was set at 1.865 × 10−11/y [35], and the present-day 176Hf/177Hf and 176Lu/177Hf ratios for chondrite and depleted mantle were used (0.28277 and 0.0332, 0.28325 and 0.0384, respectively) [36]. The single-stage model ages were computed relative to the depleted mantle, using a present-day 176Lu/177Hf ratio of 0.28325 and a 176Lu/177Hf ratio of 0.0384 [37]. Additionally, the two-stage model Hf ages were calculated assuming the average continental crust 176Lu/177Hf value of 0.015 [38]. The results are listed in online supplementary Table S4.

Sr-Nd isotope analyses were conducted using the classical two-step ion-exchange chromatographic method and measured using a Thermo Fisher Scientific Triton Plus multicollector thermal ionization mass spectrometer at the Wuhan Sample Solution Analytical Company. Details of analytical procedures were described by Li et al. [39]. Whole-rock powders for Nd isotopic analyses were dissolved in a Savillex Teflon screw-top capsule after being spiked with the mixed 149Sm-150Nd tracers prior to HF + HNO3 + HClO4 dissolution. The whole procedure blank was 100 pg for Sm-Nd. The isotopic ratios were corrected for mass fractionation by normalizing to 146Nd/144Nd = 0.7219. The international standard sample JNdi-1 was employed to evaluate instrument stability during the period of data collection. The measured values for the JNdi-1 Nd standard were 143Nd/144Nd = 0.512108 ± 0.000006 (n = 2, 2SD). The 143Nd/144Nd data of JNdi-1 show good agreement with previously published data by thermal-ionisation mass spectrometry (TIMS) technique [39]. Results of whole-rock Nd isotope analyses are presented in online supplementary Table S5.

4.1. Whole-Rock Major and Trace Element Compositions

Thirty samples of the three types of granites, namely, 23CH01-1 to 23CH01-10, 23CH02-1 to 23CH02-10, and 23JM03-1 to 23JM03-10, were selected for whole-rock major and trace element analyses. The results are listed in online supplementary Tables S1 and S2.

4.1.1. Major Element Compositions

Samples 23CH01 are characterized by high SiO2 (72.89, 74.46 wt%), high K2O (5.42, 5.99 wt%), comparatively low CaO (1.04, 1.41 wt%), and Na2O (2.65, 2.99 wt%) contents. They have high K2O/Na2O ratios (1.86, 2.16) and plot in the shoshonite series on a diagram of SiO2 versus K2O (Figure 4(a)). According to Na2O + K2O versus SiO2 total alkali silica (TAS) diagram, these samples are subalkaline, and all of them plot in the field of granite (Figure 4(b)). In addition, they reveal an Al2O3 content of 13.18%–13.75% with Aluminium Saturation Index (ASI) index (A/CNK = [Al2O3/(CaO + Na2O + K2O) mol%]) of 1.03–1.06, indicating a metaluminous feature (Figure 4(c)). Fe# numbers (FeOT/[FeOT + MgO]) of 0.76–0.81 span the magnesium and ferroan granites (Figure 4(d)). The 23CH02 samples have 73.18–75.63 wt% SiO2, 4.96–5.83 wt% K2O, relatively low (0.62, 1.64 wt%) CaO, and Na2O (2.66, 3.43 wt%) contents, which also display high K2O/Na2O ratios (1.45, 2.20) and plot in the high-K (calc-alkaline) to shoshonite series on the SiO2 versus K2O diagram (Figure 4(a)). They all exhibit subalkaline granite character on the TAS diagram (Figure 4(b)). The wide ASI index range of 0.92–1.13 as presented on the A/CNK versus A/NK diagram (Figure 4(c)) suggests a metaluminous to strongly peraluminous characteristic. Besides, the relatively low Fe# number (0.59, 0.76) of these samples shows affinities to magnesium granites (Figure 4(d)). Compositional spectra of the 23JM03 samples exhibit the highest SiO2 (74.85, 76.91 wt.%) and lowest CaO (0.74, 0.90 wt%). The 4.75–7.78 wt% K2O and 3.16–3.41 wt% Na2O display high K2O/Na2O ratios of 1.43–1.76, also implying high-K (calc-alkaline) to shoshonite plutons (Figure 4(a)). On the TAS discrimination diagram (Figure 4(b)), all samples are plotted in the field of subalkaline granite. In addition, these samples have A/CNK values of 0.90–1.02 and are plotted in the metaluminous field on the A/CNK-A/NK diagram (Figure 4(c)). The relatively high Fe# number of 0.88–0.90 (Figure 4(d)) further marks the characteristics of ferroan granites (Figure 4(d)).

On Harker plots, samples 23CH01 show negative correlations between Al2O3, Fe2O3, MgO, CaO, Na2O, P2O5, TiO2, and SiO2, the 23CH02 samples display negative correlations between Al2O3, Fe2O3, MgO, CaO, K2O, P2O5, TiO2, and SiO2, and the 23JM03 samples reveal negative correlations between Al2O3, K2O, and SiO2 (Figure 5).

4.1.2. Trace Element Composition

Chondrite-normalized rare earth element (REE) patterns and primitive mantle-normalized trace element variation diagrams for the investigated samples are presented in Figure 6. Samples 23CH01 and 23CH02 show similar trace element characteristics with high total REE contents (ΣREE) of 240.73–312.25 ppm and 234.71–266.83 ppm, respectively. The heavy rare earth element (REE) patterns are all shaped by enriched light rare earth element (LREE), high fractionation between LREE and HREE ([La/Yb]N = 17–30), and negative Eu anomalies (Eu/Eu* = 0.33–0.40) on the chondrite-normalized REE plots (Figures 6(a) and 6(c)). In addition, similar trace element-normalized patterns for these rocks show enrichment in large-ion lithophile elements (LILEs; e.g., Rb, Th, U, and Pb) and depletion in high-field strength elements (HFSEs; e.g., Nb, Sr, and Ti; see Figures 6(b), and 6(d)). In contrast, samples 23JM03 have comparatively low ΣREE of 148.50–223.46 ppm and slight fractionation between LREE and HREE ([La/Yb]N = 2–6). Chondrite-normalized REE patterns exhibit relatively flat right-declining but conspicuous negative Eu anomalies (Eu/Eu* = 0.07–0.12; Figure 6(e)). Besides, 23JM03 samples also show similar enrichment in LILEs (e.g., Rb, Th, U, and Pb) but much stronger depletion in HFSEs (e.g., Nb, Sr, and Ti; Figure 6(f)).

4.2. Zircon CL Imaging and SIMS U–Pb Dating

The cathodoluminescence (CL) images (Figure 7) show that most zircons from three granite samples in this study are euhedral, prismatic with lengths of 100250 μm and length/width ratios of 1:13:1. All zircon grains are characterized by obvious concentric oscillatory zoning in CL images and have Th/U ratios of 0.181.65, indicative of a magmatic origin.

A total of twenty analyzed spots were made on zircon grains of fine-grained granite in the Conghua area (Sample 23CH01-1), and the results are shown on the concordia diagram (Figure 8(a)). One spot was made on a zircon core that displayed an apparent 206Pb/238U age of 836.7 ± 9.0 Ma, which was interpreted as the age of inherited magmatic zircon. The rest nineteen analyses yielded concordant ages from 156.6 to 167.6 Ma, with a weighted mean 206Pb/238U age of 161.6 ± 1.6 Ma (n = 19; MSWD=4.5; see Figure 8(a)), interpreted as the crystallization age of the fine-grained granite in the Conghua area. A total of twenty-one analyzed spots were made on zircon grains of medium-grained granite in the Conghua area (Sample 23CH02-1). Sixteen analyses gave concordant ages of 156.5171.8 Ma and a weighted mean age of 163.3 ± 2.2 Ma (n = 16; MSWD = 5.5; see Figure 8(b)), which is consistent with the crystallization age of fine-grained granite in the Conghua area. Four spots yielded 206Pb/238U ages ranging from 186 to 434 Ma, which probably represents a mixture of the core and rim domains. Data for spot 6 are missing. For twenty-one analyzed spots of medium-grained granite in the Jiangmen area (Sample 23JM03-1), except four discordant spots (σ > 3 or concordance < 90%), the rest seventeen spots yielded concordant ages of 154.5162.7 Ma and a weighted mean age of 157.7± 1.0 Ma (n = 17; MSWD = 1.1, Figure 8(c)), interpreted as the crystallization age of the medium-grained granite in the Jiangmen area.

4.3. Zircon Hf-O Isotope Analyses

Zircon in situ O isotope data for samples 23CH01, 23CH02, and 23JM03 are listed in online supplementary Table S3. All samples have zircon δ18O values ranging from 7.64‰ to 10.80‰, which are significantly higher than those of igneous zircons in equilibrium with mantle magmas (δ18O = 5.3 ± 0.3‰) [40, 41]. Sample 23CH01 shows a wider range of δ18O values, varying from 8.23‰ to 10.80‰, with an average of 9.21‰ (online supplementary Table S3). In contrast, sample 23CH02 has relatively uniform δ18O values ranging from 8.37‰ to 9.47‰, with an average of 8.87‰ (online supplementary Table S3). Sample 23JM03 exhibits lower δ18O values of 7.64‰–8.37‰ with an average of 8.00‰ (online supplementary Table S3).

Magmatic zircons from sample 23CH01-1 display nearly identical 176Hf/177Hf ratios with a narrow range of 0.2823240.282661, corresponding to negative εHf(t) values ranging from −12.34 to −0.56 and hafnium model ages (TCDM) of 0.76–1.17 Ga (online supplementary Table S4). In contrast, the εHf(t) value of 23CH01-1 detrital zircon is +2.98, which corresponds to a hafnium model age (TCDM) of 1.22 Ga. The 176Hf/177Hf ratios of 23CH02-1 zircon exhibit a relatively wide range from 0.282462 to 0.282936, corresponding to εHf(t) values from +9.05 to −11.11, and hafnium model ages (TCDM) from 0.41 to 1.13 Ga (online supplementary Table S4). The 176Hf/177Hf ratios of the zircons from 23JM03-1 are relatively consistent, ranging from 0.282420 to 0.282579, corresponding to εHf(t) values ranging from −9.14 to −3.50, and hafnium model ages (TCDM) ranging from 0.86 to 1.06 Ga (online supplementary Table S4).

4.4. Whole-Rock Sr-Nd Isotopic Composition

Six granite samples 23CH01-1, 23CH01-2, 23CH02-1, 23CH02-5, 23JM03-1, and 23JM03-4 were selected for whole-rock Sr-Nd isotopic analysis (online supplementary Table S5). Samples 21CH01-1 and 23CH01-2 possess initial 87Sr/86Sr ratios of 0.70430.7039 and an identically negative εNd (t) value of −9.4, corresponding to Nd model ages of 1529 Ma and 1532 Ma. Samples 21CH02-1 and 21CH02-5 show initial 87Sr/86Sr ratios of 0.69730.7072, with negative εNd(t) values of −9.5 and −9.2 and Nd TDMC ages of 15341518 Ma. In contrast, samples 23JM03-1 and 23JM03-4 exhibit relatively higher initial 87Sr/86Sr isotopic ratios (0.6802, 0.6880) and εNd (t) values (−8.3, −8.2), with Nd model ages of 1445 and 1440 Ma.

5.1. Petrogenesis and Magma Process

Granites from the Fogang and Xinxing Batholiths are hornblende-free, highly siliceous (SiO2 > 72%), and subalkaline, of which Fogang granites are magnesian with relatively lower Fe# values, while Xinxing granites show ferroan features with high Fe# values (Figures 3 and 4). It is noticeable that both Fogang and Xinxing granites exhibit low A/CNK ratios (<1.1; see Figure 4(c)), in contrast to the highly felsic S-type granites that are strongly peraluminous with A/CNK >1.1 [42]. In addition, these granites are absent of Al-rich minerals and show a negative correlation between P2O5 and SiO2 (see Figures 5 and 9) [42, 43]. These features are important criteria for distinguishing I-type granites from S-type granites because apatite reaches saturation in metaluminous to middle peraluminous magmas but is highly soluble in strongly peraluminous melts [9, 42, 44]. Moreover, granites of the Fogang and Xinxing Batholiths have high SiO2 and alkali contents and low 10,000*Ga/Al and Zr + Nb + Ce + Y, which indicate that they are not A-type granites (Figure 9(a)). Additionally, combining other geochemical characteristics and their relatively higher Na2O+K2O/CaO ratios, they show a similarity to highly fractionated calc-alkaline I-type granites (see Figure 9(b)) [45, 46]. Thus, the Fogang and Xinxing granites should be fractionated I-type, rather than A-type or fractionated S-type [46].

Generally, three potential models accounting for the origin and magmatic process of I-type granites have been proposed: (1) fractional crystallization from mantle-derived basaltic magma [47], (2) mixing of mantle-derived and crustal materials [48], and (3) partial melting of mafic crust [49]. In this study, these high-K calc-alkaline I-type granites did not show an intermediate to felsic trend (SiO2 > 69%) in the Harker plots (Figure 5). In addition, most magmatic zircons of these granites display nearly identical 176Hf/177Hf ratios with a narrow range of 0.282324 0.282936, corresponding to negative εHf(t) values ranging from −12.34 to −0.56, and hafnium model ages (TCDM) of 0.86–1.17 Ga (Figure 10(a)). This suggests that their parental magma was derived from the partial melting of a 0.86–1.17 Ga old crust. This is also supported by the Sr-Nd isotopic data that the initial 87Sr/86Sr ratios of these granites have a narrow range of 0.6802–0.7072, negative εNd(t) values of −0.95 to −8.2 (Figure 11) and Nd model age of 15341440 Ma (online supplementary Table S5). Thus, the I-type granites in the Fogang and Xinxing Batholiths could not be derived directly from the fractional crystallization of mantle-derived magma or partial melting of mafic crust [50]. Thus, they could be a mixture of magmas derived from metaluminous infracrustal and old supracrustal rocks. Moreover, all zircons of these samples have δ18O values ranging from 7.64‰ to 10.80‰ (Figure 10(b)), which are significantly higher than those of igneous zircons in equilibrium with mantle magmas (δ18O = 5.3 ± 0.3‰) [40, 41]. These Hf-O isotopic results indicate involvements of supracrustal sedimentary components in the Fogang and Xinxing Batholiths in SE China.

Therefore, geochemical and isotopic data in this study imply that the ca. 160 Ma granites from the Fogang and Xinxing Batholiths were dominantly derived from the reworking of old crust [9, 15, 51], with a small amount of asthenosphere upwelling and underplating of mafic magmas [14]. Those mafic magmas could be a major heat source for the Jurassic granitic magmatism [14, 51].

5.2. Tectonic Implication

The Mesozoic magmatic rocks are widespread in SE China, which are essential to understanding the growth and evolution of the continental crust in SE China [9, 14]. However, the tectonic regime accounting for the generation of these granites is still controversial [1, 17]. It is considered that the Cretaceous magmatism along the coastal region occurred in an active continental margin due to the subduction of the Paleo-Pacific plate beneath the Eurasia plate [52-57], while the tectonic regime accounting for Jurassic magmatism in the interior has been long debated [9, 14, 17]. Tectonic models of Mesozoic magmatic rocks in SE China mainly include (1) the subduction of the Paleo-Pacific plate beneath the Eurasia plate [3, 15], (2) the intraplate extension and/or rifting regime [1, 16], and (3) the flat-slab subduction and slab-foundering model [12, 17].

Regarding the subduction of the Paleo-Pacific plate, the low-angle subduction had been proposed by Zhou and Li [3] to explain the petrogenesis of widespread Jurassic magmatism in SE China. In addition, Jiang et al. [15] proposed that Jurassic magmatic rocks were formed due to the slab roll-back. However, it is difficult to interpret the extensional basin-and-ranges province [16] and the contemporaneous intraplate, rifting-related igneous rocks [58-61]. An alternative intraplate extension and/or rifting regime model [1, 16] has been proposed to account for the development of intraplate Jurassic magmatism in an extensional setting. Nevertheless, the tectonic process responsible for the intraplate extension remained enigmatic. Later, Li and Li [17] proposed a flat-slab subduction model to account for the formation of the 1300-km-wide intracontinental orogen and postorogenic magmatic province in SE China.

The element and isotopic features of the 163158 Ma I-type granites from the Fogang and Xinxing Batholiths suggest that the parental magma was derived from the reworking of the old crust [9, 15, 51], accompanied by a small amount of asthenosphere upwelling and underplating of mafic magmas [14]. These mafic magmas could be the major heat source for the Jurassic granitic magmatism [14, 51]. Moreover, these ca. 160 Ma fractionated I-type biotite granites from the Fogang and Xinxing Batholiths show relatively high Rb, Ta, Yb, and Nb + Y contents, and most of them fall into the within-plate granite and syncollision granite fields (Figure 12), suggesting a syncollision or postcollision tectonic setting, rather than arc setting.

In conclusion, this study lends further support to the flat-slab subduction/slab-foundering model of Li and Li [17]. The flat slab of Paleo-Pacific plate subducted to the Eurasia plate started to break up at Early Jurassic (∼190 Ma) [17]. This process was further enhanced during the period of 180170 Ma, causing the generation of small amounts of basalts in SE China [14]. The flab-slab foundering occurred during the middle Jurassic, resulting in the upwelling of asthenosphere mantle and underplating of mafic magma, leading to the formation of large amounts of granites in an extensional setting at ca. 160 Ma.

  1. SIMS U-Pb results suggest that the Fogang and Xinxing granites were emplaced in the period of 163158 Ma.

  2. Mineralogical and geochemical features indicate that these granites are high-K (>4.8 wt% K2O at 72 wt% SiO2), calc-alkaline I-type granites.

  3. Geochemical features and isotopic data imply that these granites were most likely generated through the mixture of supracrustal sedimentary components with a minor addition of mantle-derived magmas due to asthenosphere upwelling or underplating and intrusion of mafic magmas.

The data used in this study are available in the Supplementary Materials.

The authors declare that there is no conflict of interest regarding the publication of this paper.

We gratefully acknowledge Lai Jiang for help with sampling in the field and Fu Liu for help with zircon SIMS dating and in situ LA-ICP-MS Lu-Hf isotopic analysis. This research was funded by the Mining Engineering Grant of Guangdong Dongsheng Holding Group Co., Ltd (20210113).

Exclusive Licensee GeoScienceWorld. Distributed under a Creative Commons Attribution License (CC BY 4.0).

Supplementary data