Abstract

Widespread Paleozoic and Mesozoic granites are characteristics of SE China, but the geodynamic mechanisms responsible for their emplacement are an issue of ongoing debate. To shed new light on this issue, we present an integrated geochronological and isotopic study of detrital zircon and monazite from Cambrian metasandstones and modern beach sands in the Yangjiang region, SE China. For the Cambrian metasandstone sample, detrital zircon displays a wide age range between 490 and 3000 Ma, while monazite grains record a single age peak of 235 Ma. The results suggest that a significant Triassic (235 Ma) metamorphic event is recorded by monazite but not zircon. For the beach sand sample, detrital zircon ages show six peaks at ca. 440, 240, 155, 135, 115, and 100 Ma, whereas detrital monazite yields a dominant age peak at 237 Ma and a very minor age peak at 435 Ma. Beach sand zircon displays features that are typical of a magmatic origin. Their Hf–O isotopes reveal two crustal reworking events during the early Paleozoic and Triassic, in addition to one juvenile crustal growth event during the Jurassic–Cretaceous. The beach sand monazite records intense Triassic igneous and metamorphic events with significant crustal reworking. Such early Paleozoic and Triassic geochemical signatures of detrital zircon and monazite suggest they were derived from granitoids and metamorphic rocks which formed in intraplate orogenies, i.e., the early Paleozoic Wuyi–Yunkai Orogeny and Triassic Indosinian Orogeny. The Jurassic–Cretaceous signature of detrital zircon may reflect multistage magmatism that was related to subduction of the Paleo-Pacific Plate beneath SE China.

1. Introduction

Extensive Paleozoic and Mesozoic granites distributed in SE China are crucial for understanding the Phanerozoic geodynamic evolution of the South China Block. Three magmatic events have been recognized during the early Paleozoic, Triassic, and Jurassic–Cretaceous in SE China (e.g., [1]), and the geodynamic mechanisms that are responsible for these remain hotly debated. The early Paleozoic granites have previously been considered to be the product of an intraplate orogeny [2, 3] or a collisional orogeny [4, 5]. The Triassic Orogeny is considered to be related to either continental collision between the South China, Indochina, and North China blocks [1, 6] or flat-slab subduction of the Paleo-Pacific Plate beneath SE China [7]. While it is generally accepted that the widespread Jurassic granites in SE China formed as the Paleo-Pacific Plate was subducted beneath eastern Asia, the origin and genesis of these granites are still controversial [8, 9].

Zircon (ZrSiO4) has been widely used in research of the magmatic evolution of SE China (e.g., [3, 10]) because it is a ubiquitous accessory mineral in Si-saturated igneous and high-temperature metamorphic rocks [11]. Its U–Pb ages and Hf–O isotopic compositions have been successfully used for tracing protolith of host rock, magmatic process, and tectonic events (e.g., [10, 12]). For example, zircon from early Paleozoic and Triassic granites in SE China shows a dominant crustal reworking process while zircon from Jurassic–Cretaceous granites shows more juvenile crust involved in magma formation (e.g., [3, 13]).

Zircon has also been used in a few studies to constrain the timing of amphibolite facies to granulite facies metamorphism in interiors of these orogenic belts (e.g., Wuyi and Yunkai massifs; [2, 14]). However, it is hard to date low- to medium-grade metamorphic rocks which are common rock types in the external domains of orogenic belts because it seldom crystallizes under low-temperature conditions [15]. Consequently, the metamorphic timing in the external portion of mountain belts and distribution range of the polyphase Phanerozoic metamorphism in SE China remain loosely constrained. Recent studies have emphasized the utility of monazite [(La,Ce,Nd)PO4] in geochronological studies, as it is far more susceptible to metamorphism (down to greenschist facies; e.g., [16]). It is a common accessory mineral in a wide range of igneous and metamorphic rocks, including peraluminous granite, pegmatite, metapsammite, and metapelite [17]. It can grow under low-temperature and fluid-enhanced conditions (e.g., via dissolution-reprecipitation; [18]). It has been widely referred to as a good geochronometer due to its high U and Th contents, low common Pb content and high closure temperature of U–Th–Pb system (e.g., [1921]). Therefore, monazite ages in metamorphic rocks can better characterize the timing of metamorphism.

On the other hand, zircon and monazite are common detrital minerals in sedimentary rocks and sediments from modern rivers or beach placers (e.g., [22, 23]). Hf isotopes of detrital zircon and Nd isotopes of detrital monazite provide a good estimate for the average Hf and Nd isotopic composition of large continental areas that supplied the sediments [23, 24]. Oxygen isotopes of detrital zircon and monazite can be used to characterize the host rock and investigate provenance (e.g., [25, 26]). Thus, a combined study of U–Th–Pb age and Hf–Nd–O isotopes for detrital zircon and monazite can provide a comprehensive view of the magmatic and metamorphic events that have affected a region (e.g., [23, 27]).

In this study, we use monazite to date the metamorphic age in low- to middle-grade rocks in the external part of orogenic belts in SE China. We also investigate the detrital zircon and monazite from modern beach sands to delineate the nature and distribution range of the polyphase Phanerozoic magmatism and metamorphism in SE China. Finally, we compare the geodynamic model of SE China with other well-known orogens based on integrated zircon and monazite data.

2. Geological Background and Sample Descriptions

2.1. Geological Setting of SE China

The borders of the South China Block are marked by the Qinling–Dabie–Sulu orogenic belt to the north, Longmenshan Fault to the northwest, and Red River Fault to the southwest; these structures define the boundaries between the South China Block and North China Craton, Songpan–Ganzi terrane, and Indochina Block, respectively (Figure 1(a)). The South China Block consists of the Yangtze Block to the NW and Cathaysia Block to the SE. The Cathaysia and Yangtze blocks amalgamated to form the South China Block during the early Neoproterozoic [2833]. The NE–SW-trending Jiangshan–Shaoxing Fault lies along the boundary between these two blocks (Figure 1(a); [34]), although it is poorly exposed along its southwestern section.

The Cathaysia Block contains the two oldest crystalline units: the 1.89–1.77 Ga metagranitoids and meta-amphibolites of the Wuyi domain (e.g., [3539]) and the 1.43 Ga metagranitoids and volcanic rocks of Hainan Island (e.g., [40, 41]). Various 1.0–0.8 Ga igneous rocks are sparsely distributed within the Wuyi and Yunkai massifs (e.g., [4245]). Early Paleozoic (~450–420 Ma) granitoids are widespread in SE China (Figure 1(b)), with scarce coeval volcanic and mafic rocks (e.g., [2, 46]). Minor amounts of late Paleozoic (~260–250 Ma) igneous rocks are exposed on Hainan Island (e.g., [47, 48]) and in the Wuyi domain (e.g., [49, 50]). Mesozoic (~250–90 Ma) granites and volcanic rocks crop out extensively in SE China (Figure 1(b)). The Mesozoic granites and volcanic rocks are usually divided into three age groups: Triassic (Indosinian), Jurassic (early Yanshanian), and Cretaceous (late Yanshanian; e.g., [6, 7]). ~460–420 Ma and ~250–220 Ma amphibolite- to granulite-facies metamorphic rocks crop out within the Wuyi and Yunkai massifs ([1] and references therein). ~130–100 Ma amphibolite-facies metamorphic rocks are confined to the Pingtan–Dongshan Metamorphic Belt along the southeast coast (e.g., [51]).

2.2. Regional Geology of Guangdong Province

Granitic rocks associated with the early Paleozoic, Triassic, and Jurassic–Cretaceous tectonic events crop out widely in Guangdong Province (Figure 1(b)). Moreover, the early Paleozoic Wuyi–Yunkai Orogeny has generated a regional metamorphism and migmatization of early Paleozoic sedimentary rocks, with a regional angular unconformity between the pre-Devonian and overlying Devonian strata [52, 53]. However, high-grade metamorphic rocks are mainly exposed in the Yunkai massif (e.g., [14, 54, 55]; Figure 1(b)) while low- to middle-grade rocks show planar distribution outside the Yunkai massif [52]. The Triassic Orogeny resulted in an angular unconformity between pre-Triassic and late Triassic–early Jurassic strata [52, 53]. It introduced middle- to high-grade metamorphism around the Yunkai massif (e.g., [14, 55, 56]) and low-grade to middle-grade metamorphism in east Guangdong ([52, 57]). Jurassic–Cretaceous tectonic event only affected the Pingtan–Dongshan Metamorphic Belt along the SE coast area (Figure 1(b); [51]).

The studied area is located in Yangjiang City, Guangdong Province, which is the external part of the three Phanerozoic metamorphic belts in SE China. The oldest rock unit exposed in the Yangjiang area is the Cambrian Bacun Group ([52]; Figure 2(a)). This group was deposited during middle to late Cambrian according to the discovery of inarticulate brachiopod fossils [52, 58]. It has been divided into three formations based on rock assemblages ([58]; Figure 2(b)). The bottom one is the Xiazhai Formation, consisting of quartz sandstone interlayered with carbonaceous shale and coal units. The middle one is called the Oujiadong Formation, interlayered by quartz sandstone and sericite slate. The upper one is named as the Laoshuzhai Formation, consisting of arkosic sandstone and muddy shale [58].

The Bacun Group is unconformably overlain by or in fault contact with Devonian and Mesozoic strata in Guangdong ([52, 53]; Figure 2). The Devonian strata are dominated by sandstones, shales, and conglomerates with limestone interlayers on the top. The Devonian strata were intruded by Paleozoic granites and unconformably overlain by Mesozoic strata [52, 53]. The Mesozoic (late Triassic–early Jurassic) strata contain mainly interbedded conglomerates, sandstones, and shales, intruded by Mesozoic granites. The Cambrian and Mesozoic strata were partly covered by Quaternary sediments [53].

2.3. Sampling

A Cambrian metasandstone sample (10GD48; 22°0428.2N, 111°2644.9E) from the upper part of the Bacun Group was collected in northwestern Yangjiang (Figures 1(c) and 2(a)). This sample consists mainly of quartz (~65%), muscovite (~15%), and altered minerals (~20%). Quartz is medium to coarse in size and subhedral to anhedral (Figure 3). Muscovite develops along fractures with a subhedral shape and is partly resolved. The altered minerals show a platy and elongated shape which might represent dissolved feldspar. The Cambrian metasandstone was metamorphosed under epidote amphibolite facies to low amphibolite facies estimated by lack of chlorite and partly resolved muscovite. LA-ICPMS and SIMS U–Pb ages were reported for the detrital zircon grains from this sample [59]. An aliquot of beach sands about 2 kg (NSH01; 21°323.44N, 111°3651.41E) was also collected from the backshore area of Nanshanhai beach in southwestern Yangjiang (Figures 1(c) and 2(a)).

3. Analytical Methods

3.1. Sample Preparation

Zircon and monazite concentrates were separated from the two samples using conventional magnetic and heavy liquid techniques. Zircon and monazite grains were handpicked from each concentrate under a binocular microscope. Zircon and monazite grains, together with zircon (Plešovice, 91500, Penglai, and Qinghu) and monazite standards (UNIL1-Mnz1, USGS-44069, Itambe, RW-1, and Manangoutry), were mounted on double-sided tape. Standard 2.5 cm diameter epoxy casts were made and then polished to expose the interiors of crystals. Zircon and monazite grains were documented using transmitted and reflected light photomicrographs, followed by cathodoluminescence (CL) imaging to study their internal structures and mineral inclusions. Cathodoluminescence images were obtained using a Nova (NanoSEM 450) field emission scanning electron microscope equipped with Gatan MonoCL4 detector. After imaging, the zircon and monazite mounts were coated with high-purity carbon and gold prior to electron probe microanalysis (EPMA) and SIMS analysis, respectively. The imaging and chemical analyses were performed at the Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing, China.

3.2. SIMS Zircon and Monazite Dating

Measurements of zircon U, Th, and Pb isotopes for sample NSH01 were conducted using a CAMECA IMS-1280 SIMS following the protocol of Li et al. [60]. During the analysis, the intensity of the primary ion beam was ca. 10 nA and the spot size was 20×30μm. The monocollector mode was used with the mass resolving power (MRP) fixed at 7000 (50% peak height definition here and after). Each spot measurement consists of seven cycles with a total analytical time of ca. 12 min. The calibration of Pb/U fractionation is relative to the Plešovice zircon standard (206Pb/238Uage=337Ma; [61]), while U and Th concentrations were calibrated against the 91500 zircon standard (Th=29ppm and U=81ppm; [62]). A long-term uncertainty of 1.5% (1 RSD) for 206Pb/238U measurements was propagated to the unknown analyses [63], although measured 206Pb/238U uncertainties in a specific session were generally ≤1% (1 RSD). Common Pb correction was based on nonradiogenic 204Pb, using the average present-day crustal composition [64]. Data reduction was carried out using the Isoplot program [65]. In order to monitor the external uncertainties of SIMS U–Pb zircon dating (calibrated against the Plešovice standard), a second zircon standard (Qinghu) was analysed as an unknown together with other unknown zircon crystals. Eleven measurements of the Qinghu zircon yielded a concordia age of 159.3±1.4Ma (2 SE), which is identical (within error) to the reported value of 159.5±0.2Ma (2 SE; [66]).

Monazite U–Th–Pb isotopes for samples NSH01 and 10GD48 were measured using the same CAMECA IMS-1280 SIMS, following the protocol of Li et al. [67]. The intensity of the O2 primary ion beam was ca. 1.0 nA, and the ellipsoidal spot was about 10×15μm in size. The monocollector mode was used with MRP fixed at 7000. Each spot analysis consisted of seven cycles with a total analytical time of ca. 12 min. U–Th–Pb isotopic ratios and absolute abundances were calibrated against the USGS-44069 monazite standard (206Pb/238Uage=424.9Ma; [68]). An external uncertainty of 0.7% (1 RSD) for 208Pb/232Th measurements was propagated to the unknown analyses [67]. Because of isobaric interference of 204Pb+ (232Th144Nd16O2++) during SIMS analyses on monazite [67], the 207Pb-based method [69] was used for common Pb correction by assuming the terrestrial Pb isotope composition [64]. During this study, the U–Pb and Th–Pb ages were calculated based on the decay constant values of 238U, 235U, and 232Th reported by Steiger and Jäger [70]. A second monazite standard (RW-1) was analysed to monitor the external reproducibility. Twenty-one analyses of the RW-1 monazite yield a 206Pb/238U age of 904.4±14.3Ma (2 SD) and 208Pb/232Th age of 905.1±13.6Ma (2 SD), which is identical (within error) to the reported value of 904.15±0.26Ma (2 SD; [71]).

3.3. SIMS Zircon and Monazite Oxygen Isotopes

After SIMS age dating, the sample mounts were reground and repolished to ensure that any oxygen implanted in the sample surfaces from the O2 beam was removed. Oxygen isotopic compositions were measured on the same domains that were analysed for age dating using the same CAMECA IMS-1280 SIMS. The zircon oxygen isotope was analysed following standard analytical procedures [72, 73]. The spot size was typically 15 μm in diameter. Uncertainties on single analyses were mostly ca. ±0.2‰ (2 SE). The instrumental mass fractionation factor (IMF) of zircon was corrected using the Penglai zircon standard with a δ18O value of 5.31±0.10 (2 SD; [72]). Measured 18O/16O ratios were normalized using the Vienna Standard Mean Ocean Water composition (VSMOW; 18O/16O=0.0020052) and then corrected for the IMF. A second zircon standard Qinghu was analysed as an unknown to determine the accuracy of the IMF. Twenty-six analyses of the Qinghu zircon standard yield a mean δ18O=5.40±0.31 (2 SD), which is consistent (within error) with the reported value of 5.39±0.22 (2 SD; [66]).

The oxygen isotopes of monazite were measured using a similar protocol to that of Didier et al. [74]. Four monazite O isotope standards (UNIL1-Mnz1, USGS-44069, Itambe, and Manangoutry; Th=0.612.1wt.%) were used to calibrate the matrix effects based on the correlation between IMF and Th content. However, instead of the linear regression [74, 75], we used a power-law fitting between IMF and Th content, which gave a better fit and accounted for errors of the standards. The RW-1 standard was analysed as an unknown to determine the accuracy of the composition-dependent IMF. Seventeen measurements of RW-1 during this study yielded a mean δ18O=6.33±0.49 (2 SD), which is consistent with the recommended value of 6.30±0.16 (2 SD; [76]).

3.4. LA-MC-ICPMS Zircon Lu–Hf and Monazite Sm–Nd Isotopes

Zircon Lu–Hf isotopic analyses were performed on the same zircon domains that were used for U–Pb dating and oxygen isotopes. The analysis was conducted using a Geolas-193 laser ablation system in conjunction with a Thermo Neptune Plus multicollector inductively coupled plasma mass spectrometer (LA-MC-ICPMS), following the protocol of Wu et al. [77]. The spot size for zircon Lu–Hf isotopic analyses was 65 μm in diameter, with an ablation time of 26 s, repetition rate of 8 Hz, and a laser energy density of 10 J/cm2. Isobaric interference of 176Lu on 176Hf was corrected by measuring the intensity of the interference-free 175Lu isotope and using a recommended 176Lu/175Lu value of 0.02655 [78]. Isobaric interference of 176Yb on 176Hf was corrected using independent mass bias factors for Hf and Yb. The contribution of 176Yb to 176Hf was corrected by applying ratios of 176Yb/172Yb=0.5887 and 173Yb/172Yb=0.73925 [77]. Measured 176Hf/177Hf ratios were normalized to 179Hf/177Hf=0.7325. The Mud Tank zircon standard was analysed alternately with the unknowns. During the analytical session, we obtained a 176Hf/177Hf ratio of 0.282495±0.000024 (2 SD; n=42) for Mud Tank, in good agreement (within error) with the recommended value of 0.282507±0.000006 (2 SD) by solution MC-ICPMS measurements [79].

Monazite Sm–Nd isotopic analyses were performed after the SIMS oxygen isotope analyses, on the same monazite domains for samples NSH01 and 10GD48. The same Geolas-193 laser ablation system was used for the analyses, in conjunction with a Thermo Neptune Plus MC-ICPMS. The analytical procedure was similar to that of Liu et al. [80]. The spot size for each monazite Nd isotopic analysis was 16 μm in diameter, with a repetition rate of 4 Hz, laser energy density of 10 J/cm2, and total acquisition time of ca. 60 s. A 147Sm/149Sm ratio of 1.08680 [81] and 144Sm/149Sm ratio of 0.22332 [82] were used for correcting the isobaric interference of 144Sm on 144Nd in this study. Measured 143Nd/144Nd ratios were normalized to 146Nd/144Nd=0.7219. During the analytical session, we obtained a 143Nd/144Nd ratio of 0.512182±0.000038 (2 SD; n=35) for the USGS-44069 monazite standard, which is in good agreement within error with the recommended value of 0.512175±0.000040 (2 SD; [80]).

3.5. EPMA and LA-Q-ICPMS Monazite Compositions

Elemental abundances of the monazite grains, including P, La, Ce, Pr, Nd, Sm, Gd, Dy, Y, Si, Ca, and Th, were determined near the monazite domains that were used for U–Th–Pb dating using a CAMECA SXFiveFE EPMA equipped with thallium acid phthalate (TAP) and large LiF and polyethylene terephthalate (PET) diffraction crystals (Supplementary Table S3). Operating conditions were as follows: an accelerating voltage of 15 kV, beam current of 100 nA, and defocused beam size of 5 μm. The peak counting time for each element was ca. 20–30 s, and the total acquisition time for a single data set was ca. 6 min. For the calibration procedure, a natural apatite [Ca5PO4(F,Cl,OH)] is used as the P and Ca standard (Supplementary Table S3), a natural wollastonite (CaSiO3) for the Si standard, and ten Si–Al–Ca glass standards for Th and rare earth elements (REE).

The concentrations of U, Pb, middle rare earth elements (MREE), and heavy rare earth elements (HREE) were determined using an Agilent 7500a quadrupole inductively coupled plasma mass spectrometer (LA-Q-ICPMS) in conjunction with a 193 nm ArF excimer laser system. The analytical procedure was similar to that of Wu et al. [83], with a laser beam diameter of 24 μm, pulse repetition rate of 3 Hz, and beam flux of 5 J/cm2. The element concentrations were calibrated against the National Institute of Standards and Technology (NIST) Standard Reference Material (SRM) 610 as an external reference material, with the Ce content of the monazite (determined by EPMA) used as an internal standard. Concentrations were calculated using the GLITTER 4.0 program [84].

4. Results

4.1. Zircon U–Pb Ages and Hf–O Isotopes

SIMS zircon U–Pb analysis was conducted on 150 grains for the beach sand sample NSH01 (Supplementary Table S1). 206Pb/238U ages are adopted for zircon grains younger than 1000 Ma, while 207Pb/206Pb ages are used for zircon grains older than 1000 Ma. Excluding a few crystals with ages older than 1100 Ma, zircon grains are subhedral to euhedral in shape and have oscillatory zoning under CL imaging (Supplementary Figure S1). Twelve zircon analyses are discordant and high in 204Pb. The remaining 138 analyses are concordant (within 90–110% of concordance; Supplementary Figure S2). Of the 138 concordant analyses, 107 zircon ages (accounting for 82% of the total analyses) are between 95 and 500 Ma, with six major peaks at 440, 240, 155, 135, 115, and 100 Ma (Figure 4(a)). The remaining 31 analyses yield variable zircon ages between 576 and 2751 Ma without any significant age peaks.

All zircon grains have δ18O values from 3.3‰ to 12.6‰ and εHft values from −17.3 to +10.1 (Supplementary Table S2). For zircon with an age peak of 440 Ma, the corresponding δ18O values are 7.9‰ to 11.0‰ (Figure 5(a)) and the εHft values are −10.8 to −0.8 (Figure 5(b)). For zircon with an age peak of 240 Ma, the corresponding δ18O values are 8.0‰ to 10.8‰ and the εHft values are −12.7 to −8.0. Zircon grains with Jurassic–Cretaceous ages show a clear decrease in δ18O values (from 7.9‰ to 4.8‰), but an increase in εHft values (from −17.0 to +4.2) with younger ages (Figure 5).

4.2. Monazite U–Th–Pb Ages, Nd–O Isotopes, and Chemical Compositions

4.2.1. Cambrian Metasandstone Sample (10GD48)

A total of 150 monazite grains from sample 10GD48 were selected for SIMS U–Th–Pb age analysis (Supplementary Table S1). Monazite crystals are mostly subhedral to euhedral in shape, with patchy to homogeneous internal structures under CL imaging (Supplementary Figure S1). 143 grains are concordant (within 90–110% of concordance) for U–Pb and Th–Pb ages (Supplementary Figure S3), of which Th–Pb ages display a conspicuous age peak at 235 Ma, with a weighted mean of 235.3±0.5Ma (2 SE; MSWD=1.2; Figure 6(a)). Thirty-three grains were selected for O and Nd isotopic microanalyses (Supplementary Table S2), and they show fairly homogeneous O–Nd isotopes, with an average δ18O=11.4±0.8 (2 SD) and εNdt=12.3±0.6 (2 SD; Figure 7).

These monazite grains are also relatively homogeneous in terms of chemical composition, with high ThO2 (5.84±1.48wt.%, 2 SD), low SiO2 (0.38±0.44wt.%, 2 SD), and low CaO (1.52±0.43wt.%, 2 SD) contents. The total REE contents vary from 50.98 to 53.12 wt.%, with the Y2O3 content ranging between 0.86 and 3.60 wt.%. All the monazite grains are light rare earth element (LREE) enriched (Supplementary Figure S4), with an average (La/Dy)N ratio of 20 (Supplementary Table S3).

4.2.2. Beach Sand Sample (NSH01)

Monazite grains from the beach sand sample (NSH01) are mostly subhedral to euhedral and display oscillatory and patchy zoning under CL imaging (Supplementary Figure S2). A total of 161 monazite grains were selected for SIMS U–Th–Pb age analysis (Supplementary Table S1), of which 157 are concordant (within 90–110% of concordance) for U–Pb and Th–Pb ages (Supplementary Figure S2). 140 grains show a major age peak at ca. 237 Ma and 16 grains form a subordinate age peak at ca. 435 Ma (Figure 4(b)). One single grain (NSH-m03@12) yields an age of 153 Ma.

Monazite δ18O values for sample NSH01 are highly variable (8.5‰–11.5‰) for both the 435 and 237 Ma groups of monazite (Figure 7(a)). Furthermore, the 435 Ma group of monazite yields εNdt values between −15 and −8, while the 237 Ma group yields εNdt values between −18 and −7 (Figure 7(b)). The single monazite grain with an age of 153 Ma yields a δ18O value of 6.0‰ and εNdt value of −8.

In terms of chemical composition, all monazite shows highly variable ThO2 (1.93–14.60 wt.%), SiO2 (0.12–3.30 wt.%), and CaO (0.38–2.43 wt.%) contents. The total REEs vary from 45.51 to 55.98 wt.%, with Y2O3 contents from 0.36 to 5.06 wt.%. Monazite grains are characterized by LREE enrichment (Supplementary Figure S4) with (La/Dy)N ratios ranging from 5 to 255 (Supplementary Table S3).

5. Discussion

5.1. Detrital Zircon and Monazite Records from the Cambrian Metasediments

Yao et al. [59] investigated the provenance of the Cambrian metasediments in the Yangjiang area based on the study of detrital zircon from the same sample (10GD48). They suggested that this metasandstone characterized by a 1000 Ma detrital zircon group (Figure 6(b)) was most likely derived from north India during the Cambrian. This idea is supported by a lack of locally exposed ~1000 Ma igneous rocks in South China and comparable zircon age patterns and εHft values with north India ([59] and references therein). They proposed that after breaking away from central Rodinia, the South China Block collided with north India, causing the early Paleozoic Wuyi–Yunkai Orogeny and transporting materials from north India to the Yangjiang area.

Monazite extracted from the Cambrian metasandstone sample (10GD48) yields a unimodal age population of 235 Ma (Figure 6(a)) which is extremely younger than the deposition age of this sedimentary rock. As monazite could have crystallized during the magmatic, metamorphic, and hydrothermal events, specific geochemical characteristics are used to determine the origin(s) of the monazite (e.g., [85]). In general, hydrothermal monazite has lower Th contents and Th/Ce ratios than igneous and metamorphic monazite (Figure 8(a); [76, 86]). Igneous monazite falls in the centre of the Si/Ce–Ca/Ce–Y/LREE diagram, while metamorphic monazite is scattered in the low Si/Ce and low Y/LREE fields (Figure 8(b); [76]). Nearly all the 235 Ma monazite grains from the Yangjiang Cambrian metasandstone show geochemical features of metamorphic origins (Figure 8). This suggests that, after deposition in the Yangjiang area, the Cambrian sedimentary rocks experienced intense metamorphism during the Indosinian Orogeny, coeval with the metamorphic age peak in the Yunkai massif of the orogenic belt (Figure 9(a)). The high monazite δ18O values (ca. 11.4‰) are comparable with those of high-grade metapelitic monazite [75]. All the monazite grains have highly negative εNdt values (ca. −12.3), indicating significant involvement of old supracrustal components when metamorphic monazite formed.

5.2. Detrital Zircon and Monazite Records from Modern Beach Sands

Detrital zircon grains from the Yangjiang beach sands preserve three major zircon age groups of the early Paleozoic (ca. 440 Ma), Triassic (ca. 235 Ma), and Jurassic–Cretaceous (ca. 155, ca. 135, ca. 115, and ca. 100 Ma; Figure 4(a)). All of them display oscillatory zoning under CL images and have high Th/U ratios (>0.1; Supplementary Table S1), which are indicative of a magmatic origin.

The current topography of SE China generally shows a high altitude in the northwest and low in the southeast. Therefore, modern rivers in SE China flow eastward and southeastward into the Pacific Ocean and South China Sea (Figure 1(b)) and supply continental pieces for the beach sands. Phanerozoic igneous rocks are widespread in SE China (e.g., [87, 88]) and potentially supply igneous zircon to the Yangjiang beach sands through coastal rivers. Zircon grains from the beach sands are subhedral to euhedral in shape suggesting transport by a short distance. They have comparable age patterns with detrital zircon from Pearl River Mouth Basin [89, 90] and similar δ18O and εHft values to zircon from granites in the Nanling Region (within the Pearl River system; Figure 5). Specifically, εHft values of the early Paleozoic and Triassic zircon reveal crustal reworking, consistent with the enriched Nd isotopes of regional granites (Figures 5(b) and 7(b)). The Jurassic–Cretaceous zircon δ18O and εHft values indicate a trend, from supracrustal reworking (older) to juvenile involvement (younger; Figure 5), as do the regional Mesozoic granites [91, 92]. Therefore, most zircon grains from the beach sands were likely derived from weathering and erosion of proximal Paleozoic and Mesozoic granitoid rocks and recycled from adjacent strata (e.g., through Moyangjiang River in the Yangjiang area; Figure 2(a)) which should have similar zircon features to regional rocks in Guangdong.

Monazite from the Yangjiang beach sands is classified into a main age group of ca. 237 Ma and a minor group of ca. 435 Ma (Figure 9(a)). Geochemically, the 435 Ma monazite is dominantly of metamorphic origin while the main group of Triassic monazite is mixed with both igneous and metamorphic grains (Figure 9(b)). The igneous group displays a continuous age ranging from 230 to 235 Ma, while the metamorphic group has a dominant age peak at ca. 237 Ma (Figure 9(b)). It seems that a major thermotectonic event took place at ca. 235 Ma in SE China, with magmatism and metamorphism that produced the monazite (i.e., Indosinian Orogeny; [1]). It is noticed that εNdt values for some ca. 235 Ma monazite grains are very similar to the whole-rock εNdt values of the Indosinian peraluminous granites from the Pearl River drainage area (Figure 7(b)), indicating that the detrital monazite was partly derived from the regional Indosinian peraluminous granites. Some 235 Ma monazite grains have lower εNdt values than regional granites and similar δ18O values to the Cambrian metasandstone suggesting a contribution from regional metamorphic rocks. Thus, it appears that proximal Triassic rocks were the source of igneous and metamorphic monazite. This is further supported by similar monazite age patterns in the modern Pearl River sands [93] and the Yunkai massif ([94]; Figure 9(a)) where the Triassic monazite is much more abundant than the early Paleozoic one.

5.3. Contrasting Records of Detrital Zircon and Monazite

As mentioned previously, detrital zircon is indicative of the high-temperature magmatic events that affected the source area. Detrital monazite provides additional information on metamorphic events that might be missed by the detrital zircon record. Studying sedimentary provenance by combination of detrital zircon and monazite can therefore produce different but complementary datasets.

For the Yangjiang Cambrian metasandstones, detrital zircon provides a good record of the major and minor magmatic events during the Archean to late Neoproterozoic in their “external” source region (Figure 6(b)). In contrast, monazite records a strong metamorphic imprint from the regional Indosinian Orogeny (Figure 6(a)). The lack of ~440 Ma metamorphic monazite records suggests that the early Paleozoic Wuyi–Yunkai Orogeny has insignificant effects in the Yangjiang area or has been entirely removed by the subsequent Triassic event. We consider that the latter is unlikely the main reason because (1) the ~440 Ma metamorphic monazite survived during the Triassic Orogeny seen from the beach sands and (2) the Cambrian metasandstone has experienced metamorphism of low amphibolite facies with temperature much lower than the closure temperature of ~900°C for monazite U–Th–Pb system [19].

The detrital zircon and monazite records are more complicated for the modern beach sands from Yangjiang. Detrital zircon records three major magmatic events in South China, i.e., the early Paleozoic Wuyi–Yunkai Orogeny, Triassic Indosinian Orogeny, and Jurassic–Cretaceous Yanshanian tectonic event (Figure 4(a)). However, detrital monazite mainly records the magmatic and metamorphic events during the Triassic Orogeny, with a weak imprint of early Paleozoic Orogeny (Figure 4(b)).

Specifically, the Jurassic–Cretaceous event is almost absent in the detrital monazite record, apart from one 153 Ma grain (Figure 4(b)) of igneous origin (Figure 8(b)) which is very similar to the small, highly differentiated granitic plutons associated with polymetallic deposits in SE China (e.g., [67, 95]). On a P2O5 versus SiO2 plot (Figure 10), the Jurassic–Cretaceous granites exhibit a clear I-type and fractioned I-type trend, while the Triassic and early Paleozoic granites show both S- and I-type trends. Because monazite crystallization is favoured in peraluminous S-type and highly fractionated I-type granites [96], the Jurassic–Cretaceous granites would not generate abundant igneous monazite. Furthermore, the Jurassic–Cretaceous metamorphism required for generating metamorphic monazite is scarcely documented in the interior of SE China. The Pingtan–Dongshan (or Changle–Nan’ao) Metamorphic Belt along the coastal region (Figure 1(b)) is the only exception, as it hosts Jurassic–Cretaceous amphibolite-facies metamorphic rocks (e.g., [51, 97, 98]). As a result, the metamorphic Jurassic–Cretaceous monazite is more abundant in SE coast areas proximal to the Pingtan–Dongshan Metamorphic Belt (e.g., the Hanjiang, Jiulongjiang, and Jinjiang rivers; the Xiamen and Jinmen regions) than in distal areas (e.g., the Oujiang, Minjiang, and Pearl rivers; the Yangjiang and Wuyi–Yunkai massifs; [93, 94, 99102]). Thus, the scarcity of Jurassic–Cretaceous monazite in the Yangjiang beach sands is attributed to the dominance of Jurassic–Cretaceous I-type granites and a lack of Jurassic–Cretaceous metamorphism in the interior of Southeast China.

It is evident that early Paleozoic monazite is significantly less abundant than Triassic monazite in the Yangjiang beach sands (Figure 4(b)), although the zircon populations for these two age groups are nearly equal (Figure 4(a)). As discussed above, monazite grains are mainly derived from regional peraluminous granites and metamorphic rocks. Three factors may account for the difference in detrital monazite and zircon records. (1) The exposed area of the Triassic granites in the Yangjiang area is larger than that of the early Paleozoic granites (Figure 2(a)), which may have brought more Triassic igneous monazite. However, the quantity of early Paleozoic and Triassic zircon is comparable which makes this factor less important. (2) In the Yunkai massif, the quantity of Triassic metamorphic monazite is significantly larger than that of the early Paleozoic metamorphic monazite ([101]; Figure 9(a)), emphasizing more contribution of regional Triassic metamorphic rocks to the Yangjiang beach sands. (3) Early Paleozoic Orogeny had less impact in the adjacent area and possibly introduced more low-temperature metamorphism seen from the abundant Triassic monazite and absence of Paleozoic metamorphic monazite in the Cambrian metasandstone. The last two factors suggest that the provenance of the Yangjiang beach sands is far away from the interiors of the early Paleozoic Wuyi–Yunkai orogenic belt and is closer to the interiors of the Triassic orogenic belt.

5.4. Implications for Phanerozoic Polyphase Thermotectonic Events in SE China

Different types of orogens have distinct magmatic and metamorphic processes which can be archived in zircon and monazite records (e.g., [27]). For example, zircon can record extensive arc magmatism during oceanic subduction (Figure 11), while monazite can record syncollisional metamorphism that is often missed by zircon (e.g., [27]). This is because little contemporaneous magma (e.g., migmatites and leucogranites) will be generated when the continental crust is thickened during continental collision with only minor heat input from the mantle [103, 104]. On the other hand, abundant zircon can be produced by numerous late-orogenic granites (e.g., [105, 106]). Therefore, comparison of age peaks between zircon and monazite can elucidate tectonic environments.

It is generally accepted that the widespread Jurassic–Cretaceous Yanshanian magmatic rocks in SE China are genetically related to the subduction of the Paleo-Pacific Plate (e.g., [6, 7]). Large amounts of igneous rocks are characterized by increasing inputs of juvenile component (e.g., [107, 108]), as shown by the detrital zircon Hf–O isotopes of Yangjiang beach sands (Figure 5). In sharp contrast with the multimodal zircon records, late Mesozoic monazite records are nearly absent from the Yangjiang beach sands (Figure 4). Such decoupled zircon and monazite records are typical of arc magmatism comprising a predominantly silicic member of I-type granites and their volcanic counterparts [27, 85]. Although some late Mesozoic metamorphic monazite was reported from the schists and gneisses at the Jinmen Island (e.g., [99]), these rocks are only locally distributed within the Pingtan–Dongshan Metamorphic Belt in the SE coast area.

As discussed above, both the early Paleozoic Wuyi–Yunkai Orogeny and the Triassic Indosinian Orogeny are well archived by the detrital zircon and monazite grains from the Yangjiang beach sands (Figure 4). The detrital zircon from the beach sands records extensive granitoid magmatism during these two periods. Taken together with published CHIME monazite dates [94, 101], the early Paleozoic and Triassic monazite records extensive metamorphism (Figure 9(a)). The overlapping igneous zircon and metamorphic monazite age peaks reveal that granitoid plutonism and regional metamorphism were mainly coeval during the early Paleozoic and Triassic orogenies. This is distinguished from the syncollisional stage in typical collisional orogens where metamorphic monazite records are abundant and igneous zircon records are rare (Figure 11; [27]). Furthermore, the early Paleozoic and Triassic granitoid plutons have similar planar distributions (Figure 1(b)), with dominantly peraluminous and subordinately metaluminous granites ([1]; Supplementary Table S4). These features are different from those of granitoids formed in an accretionary orogen where granitoid plutons usually occur in an elongated orogenic belt and are predominantly I-type [109111]. Therefore, the early Paleozoic and Triassic intracontinental igneous rocks are unlikely to have been generated in an accretionary or collisional orogen (e.g., [4, 6, 112]). Intraplate orogenesis possibly caused by flat-slab subduction ([2, 7]) can better account for the early Paleozoic and Triassic plutonism, which is consistent with the detrital zircon and monazite records in the Yangjiang beach sands.

6. Conclusions

We draw the following conclusions based on our integrated isotopic and elemental microanalyses of detrital zircon and monazite from the Cambrian metasandstone and modern beach sands in the Yangjiang area of SE China.

  • (1)

    Monazite from the Cambrian metasandstone records intense Triassic metamorphism which is coeval with the metamorphic age peaks in the interiors of the Triassic orogenic belt

  • (2)

    Detrital zircon from the beach sands records three major magmatic events in the early Paleozoic, Triassic, and Jurassic–Cretaceous. The early Paleozoic and Triassic zircon records an event dominated by crustal reworking, while the Jurassic–Cretaceous zircon reveals an increasing input from juvenile crustal material with time

  • (3)

    Detrital monazite from the beach sands mainly records the Triassic Orogeny with a weak imprint of early Paleozoic Orogeny. Intense Triassic metamorphism contributed to the abundance of Triassic monazite. Detrital zircon and monazite were derived from weathering and erosion of proximal Paleozoic and Mesozoic granites, metamorphic and sedimentary rocks

  • (4)

    Ages and isotopic data from detrital zircon and monazite place new constraints on the thermotectonic evolution in SE China during the Phanerozoic. By integrating these data with the spatiotemporal distribution and chemical compositions of granitoid rocks in the region, the Jurassic–Cretaceous granitoids were produced in an active continental margin setting, due to subduction of the Paleo-Pacific Plate, while the early Paleozoic and Triassic rocks are more likely formed from reworked continental crust in the intraplate orogen, rather than in an accretionary and/or collisional orogen

Data Availability

All the data used in this study are provided in the supplementary materials. No other data from any other source was used in the research described in the article.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

We thank Y. Liu, G.Q. Tang, Y.H. Yang, S.T. Wu, and Q. Mao for the assistance with SIMS, LA-MC/Q-ICPMS, and EPMA analysis. J.L. Zhou, Y. Li, and C. Yang are thanked for useful discussions. We are grateful for Dr. Villa I.M. who helped to improve the content and presentation. This study was supported by the National Natural Science Foundation of China (Grant 41673018). This is a contribution to IGCP 648.

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