Elemental and Sr-Nd-Pb-Hf Isotope Signatures of Early Cretaceous Magmatic Rocks in the Wulian Area, Eastern Shandong: Implications for Crust-Mantle Interaction at the Edge of the Sulu Collisional Orogenic Belt

Crust-mantle interaction is a key issue in understanding tectonomagmatic processes in subduction zones [1, 2]. Mantle geochemistry and its relationship to the subduction of oceanic plates into the mantle have been studied for many years [3–6]. With the discovery of ultrahigh-pressure (UHP) metamorphic minerals, such as coesite and diamond in continental supracrustal rocks [7, 8], the oceanic subduction channel model has been extended to include continental subduction zones [2, 9]. This model explains the fluid/melt metasomatism observed during continental subduction and the occurrence of residual slab trapped in the lithospheric mantle [10]. Models of crust-mantle interaction between continental crust and lithospheric mantle during subduction have been confirmed by studies on ultrahigh-pressure rocks [11–13].

The Sulu orogenic belt, bounded by the Wulian-Yantai fault and the Jiashan-Xiangshui fault (Figure 1(b)), is the product of the continental collision between the YC and the NCC during the Triassic and represents the largest exposed UHP and high-pressure (HP) metamorphic belt in the world [14–17]. Like many orogenic belts [18], postcollisional magmatism is abundant in the Sulu orogenic belt. Detailed studies on these magmatic rocks suggest that they were derived from an enriched lithospheric mantle modified by the subduction of the YC [19–21]. However, previous studies mostly assessed the orogenic belt itself or the southeastern NCC. Few studies have focused on the edge of the Sulu orogenic belt, leaving the characteristics of crust-mantle interaction along the edge of Sulu orogenic belt unclear.

The Wulian area of the Shandong province, located in the northwestern Sulu orogenic belt, comprises a significant amount of the Early Cretaceous magmatic rocks (Figure 1(b)) [22–26]. Previous studies have shown that the source region of these rocks involves both crustal and mantle materials. However, the mechanisms of crust-mantle interaction remain controversial [22–27]. It has been proposed by some researchers that the magma source may be related to the collision and deep subduction of the YC below the NCC in the Triassic [1, 21, 28–30]. Others argued that the enrichment of the magma source may be related to lithospheric delamination of the NCC [24, 27, 31, 32] or carbonate metasomatism related to oceanic crust subduction [20, 33, 34]. Therefore, the origin of the magmatic rocks in the Wulian area could provide information on crust-mantle interaction beneath the edge of the Sulu orogenic belt.

This study focuses on the Early Cretaceous magmatic rocks in the Wulian area, which are located at the intersection of the southern Sulu orogenic belt and the Jiaolai Basin. Detailed petrographic and geochemical analyses, together with zircon U-Pb dating, Hf isotope, and whole-rock Sr-Nd-Pb isotope analyses, are used to constrain the petrogenesis and geodynamic setting of the Early Cretaceous magmatic rocks in the Wulian area and to provide new insights into crust-mantle interaction during deep continental subduction.

The Dabie-Sulu orogenic belt is located in eastern China and was formed by Triassic subduction and collision between the YC and the NCC (Figure 1(a)). The Dabie-Sulu orogenic belt is the largest exposed (-30, 000 km²) and best-preserved UHP and HP metamorphic belt in the world [12, 35–37]. The belt is divided into two parts by the Tanlu fault, namely, the Dabie orogenic belt in the west and the Sulu orogenic belt in the east [38, 39] (Figure 1(a)).

The rock suites from the Sulu orogenic belt are composed of UHP and HP metamorphic rocks, as well as Mesozoic magmatic rocks [2, 12, 16] (Figure 1(b)). The ages of magmatic rocks can be divided into three periods: Late Triassic, Late Jurassic, and Early Cretaceous [1]. Early Cretaceous magmatic rocks outcrop extensively in the Sulu orogenic belt and mostly include large-scale intermediate rocks and minor mafic rocks [40–42].

The Wulian area is located in the northwestern Sulu orogenic belt, east of the Tanlu fault zone, and southwest of the Jiaolai Basin (Figure 1(b)). Mesozoic magmatic rocks bounded by the Wulian fault are widely distributed in the Sulu orogenic belt and the Jiaolai Basin [14–17]. The Qibaoshan complex (Figure 1(c)), in the northwestern Wulian area (35°52$′$N, 119°05$′$E), where most of the samples were collected, covers an area of ~12 km² and is located north of the Wulian fault (Figure 1(b)). The complex is surrounded by the Cretaceous volcanic rock series of the Qingshan Group [25, 26]. The Qibaoshan complex is a volcanic crater, which is mainly composed of intermediate-basic and felsic rocks. The volcanic facies making up this volcanic complex include intrusions, lavas, and volcano-sedimentary rocks. The intrusions and volcanic rocks are mainly composed of pyroxene diorite, granodiorite, diorite porphyrite, andesitic porphyrite, pyroxene andesite porphyrite, amphibole andesite porphyrite, and granodiorite porphyry (Figure 1(c)). The Xiaoshuhe dikes, in southeastern Wulian area (35°46$′$N, 119°11$′$E), intruded into the Mesozoic sedimentary rocks.

Samples in this study were collected from the Qibaoshan complex and the Xiaoshuhe dikes. The studied rocks are composed of trachy-andesites, pyroxene diorites, diorite porphyrites, and dolerites, which are typical of the Early Cretaceous magmatic rocks in the Sulu orogenic belt. The sampling locations are shown in Figure 1. Trachy-andesites (Figure 2(a)) from the Qibaoshan complex (samples QBS-12, QBS-14, Keng-04, and LD-03) (Figure 1(c)) comprise ~35% plagioclase phenocrysts which are commonly rimed by potassic feldspar. The matrix is composed of ~35% plagioclase, ~10% magnetite, and ~20% glassy material (Figure 2(e)). Pyroxene diorites (Figure 2(b)) from the Qibaoshan complex (samples LD-04 and LD-05) (Figure 1(c)) are characterized by a medium-grained texture, which comprises ~45% plagioclase, ~20% amphibole, ~15% clinopyroxene, ~10% alkali feldspar, ~5% biotite, and ~5% quartz (Figure 2(f)). Diorite porphyrites (Figure 2(c)) with a massive structure were collected from the Xiaoshuhe dikes (samples XSH-01 and XSH-02) (Figure 1(b)) and are composed of 45% phenocrysts. The phenocrysts comprise amphibole (~10%) and plagioclase (~35%), while the groundmass consists of microcrystalline and vitreous minerals. Dolerites (Figure 2(d)) were collected from the Xiaoshuhe dikes (samples XSH-03 and XSH-04) (Figure 1(b)) and have ophitic texture and massive structure. The dolerites are composed of ~50% alkali feldspar, ~25% clinopyroxene, ~15% biotite, and~10% glassy material (Figure 2(h)).

Zircons from all samples analyzed in this study were separated from whole-rock samples by using a combined magnetic and heavy liquid separation at the Laboratory of Hebei Provincial Institute of Regional Geology and Mineral Investigation. Zircon grains were mounted in epoxy mounts at random, which were then polished. Cathodoluminescence (CL) images of zircons were obtained by JEOL IT-500 SEM equipped with a Delmic CL system.

Zircon U-Pb and Hf isotopic analyses were conducted by LA-ICP-MS at the Beijing GeoAnalysis Co., Ltd. Laser sampling was performed using an ESI NWR 193 nm laser ablation system. An AnlyitikJena PQMS Elite ICP-MS instrument was used to acquire ion signal intensities. The laser spot diameter can reach 35 μm, and the energy density is 2.31 J/cm². The international standard zircon GJ-1 was used as the external standard reference material [44]. The ICPMSDataCal program software was used to process the raw data, and the Isoplot4.15 program was used to draw the zircon age Concordia curve. The isotopic ratio and age error are 1 σ, and the U-Pb weighted average age error is within the 95% confidence interval.

In situ zircon Lu-Hf isotopes were analyzed by a RESO 193 nm laser ablation system and a Neptune Plus MC-ICP-MS. The resolution SE 193 nm excimer laser ablation system is used to ablate zircon. The laser ablation spot beam diameter is generally 38 μm, the energy density is 7-8 J/cm², and the frequency is 10 Hz. The detailed processing process and parameters can be found in Sláma et al. [45].

Whole-rock major and trace elements were determined at the Analysis and Testing Research Center of the Beijing Institute of Geology of the Nuclear Industry. The major elements were analyzed by an AxiosmAXX ray fluorescence spectrometer, and the analytical uncertainty of this method is better than 2%. Trace elements were analyzed by inductively coupled plasma-mass spectrometry (ICP-MS) using an ELEMENT mass spectrometer. The samples were dissolved in sub-boiling distilled 1 ml HF and 0.5 ml (1 + 1) HNO₃ in Savillex Teflon screw-cap capsules and then ultrasonically stirred for 15 min. Then, the solutions were dried at 150°C, and the residues were digested with sub-boiling distilled 1.5 ml HF and 0.5 ml (1 + 1) HNO₃ in Teflon screw-cap capsules. Then, the solutions were heated to 170°C during 10 days and dried and re-dissolved in 2 ml (1 + 1) HNO₃ in the capsules. The solutions were put into plastic beakers, and then 1 ml 500 ppb was added as an internal standard. Finally, the solutions were diluted in 1% HNO₃ to 50 ml before analyses. The analytical uncertainties are lower than 10%. The detailed analytical procedures can be found in Qu et al. [46].

Whole-rock Sr, Nd, and Pb isotopic compositions were measured at the Institute of Geology and Geophysics, Chinese Academy of Sciences (IGGCAS). The 50-100 mg sample was dissolved by HF+HNO₃+HClO₄ mixed acid and then separated and purified. The isotopic compositions were measured on a Finnigan MAT262 isotope mass spectrometer. The Sr isotopic ratio was corrected to $86Sr/88Sr=0.1194$ by mass fractionation; the Nd isotopic ratio was corrected to $146Nd/144Nd=0.7219$⁠; and the Pb isotopic ratio was corrected to NBS981 standard sample. The detailed analytical procedures can be found in Li et al. [47].

The LA-ICP-MS zircon U-Pb isotopic data and the in situ zircon Hf isotopic data for the three samples from Qibaoshan complex are listed in Table S1 and Table S2.

Zircons from sample Keng-04 display magmatic oscillatory zoning in CL imaging (Figure 3(a)). All zircons are colorless, transparent, subhedral-euhedral grains, with a length range of 70 to 130 μm. The Th/U ratios of zircons are greater than 0.6 (Table S1), indicating that the zircons are magmatic. Twenty-four analyzed zircon grains yield a weighted mean ²⁰⁶Pb/²³⁸U age of $123±1$Ma (Figures 4(a) and 4(b)), representing the timing of the crystallization of sample Keng-04.

Zircons from sample LD-04 are irregular, colorless, and transparent, with a length range of 80 to 120 μm. The CL images reveal that the zircons are characterized by magmatic oscillatory zoning (Figure 3(b)). The Th/U ratios of these zircons vary from 1.28 to 2.60 (Table S1), implying that they have a magmatic origin. Twenty-four analyzed zircon grains yield a weighted mean ²⁰⁶Pb/²³⁸U age of $124±1$Ma (Figures 4(c) and 4(d)), representing the timing of the crystallization of sample LD-04.

Zircons from sample QBS-12 exhibit magmatic oscillatory zoning in CL imaging (Figure 3(c)). The grains are irregular, colorless, and transparent, with a length range of 100 to 180 μm. The Th/U ratios of zircons vary from 0.87 to 2.80 (Table S1). These features suggest that these zircons also have a magmatic origin. Twenty-five zircon grains yield a weighted mean ²⁰⁶Pb/²³⁸U age of $124±1$Ma (Figures 4(e) and 4(f)), representing the timing of the crystallization of sample QBS-12.

The Lu-Hf isotopic analyses were carried out on representative zircons that have been dated (Table S2). In general, the magmatic zircons from the samples have similar Hf isotope compositions to each other (Figure 5), which resemble those of the Dabie-Sulu orogenic belt [21]. Zircon grains from sample Keng-04 have variations in initial ¹⁷⁶Hf/¹⁷⁷Hf ratios (0.282108-0.282195), with $εHft$ values ranging from -20.8 to -17.8 and two-stage Hf model ages ranging from 2.49 to 2.30 Ga. Zircon grains from sample LD-04 have variations in initial ¹⁷⁶Hf/¹⁷⁷Hf ratios (0.282198-0.282284), with $εHft$ values ranging from -17.6 to -14.7 and two-stage Hf model ages ranging from 2.29 to 2.10Ga. Zircon grains from sample QBS-12 have variations in initial ¹⁷⁶Hf/¹⁷⁷Hf ratios (0.282246-0.282312), with $εHft$ values ranging from -15.6 to -13.6 and two-stage Hf model ages ranging from 2.18 to 2.04 Ga.

Whole-rock major and trace element data are listed in Table 1. The magmatic rocks in the Wulian area are intermediate rocks, which have a range of SiO₂ contents (50.0-60.0 wt.%). On the TAS diagram (Figures 6(a) and 6(b)), these samples are classified as monzodiorite, monzonite, basaltic trachy-andesite, and trachy-andesite. The samples have relatively high K₂O values (3.2-5.8 wt.%) and are classified as high-K calc-alkaline and shoshonitic rocks (Figure 6(c)). The AR-SiO₂ diagram shows that the samples mostly fall within the alkaline field (Figure 7(a)). The samples have A/CNK values of 0.73 to 1.85 and most plot in the metaluminous field (Figure 7(b)). They have moderate MgO (3.3-5.6 wt.%), low TiO₂ (0.4-1.2 wt.%), and Al₂O₃ and P₂O₅ ranging from 13.9 to 17.7 wt.% and 0.3 to 0.7 wt.%, respectively. The TiO₂, CaO, MgO, and P₂O₅ of the samples are negatively correlated with SiO₂ (Figure 8).

The intermediate rocks are enriched in light rare earth elements (LREEs), depleted in heavy rare earth elements (HREEs) (Figure 9(a)), and show slight negative Eu anomalies (⁠$δEu=0.61−0.93$⁠) and high (La/Yb) _N ratios (22.73-57.02). On primitive mantle-normalized multielement diagrams (Figure 9(b)), the samples are enriched in LILEs (e.g., Pb, Ba, and K) and markedly depleted in HFSEs (e.g., Nb, Ta, Zr, and Ti). These geochemical signatures are similar to those of the Early Cretaceous magmatic rocks from the Sulu orogenic belt [1, 30].

Whole-rock initial Sr-Nd-Pb isotopic compositions of the samples are shown in Table 2 and have been age-corrected to 123 Ma based on the age of the zircons. Initial ⁸⁷Sr/⁸⁶Sr (0.7079-0.7088) and $εNdt$ (-20.6 to -14.6) of the samples are lower than those of the Early Cretaceous magmatic rocks found along the Tanlu fault (⁠$87Sr/86Sr=0.7094−0.7112$⁠, $εNdt=−14.2~−11.1$⁠) (Figure 10(a)) [51, 52]. The samples have initial ²⁰⁶Pb/²⁰⁴Pb of 16.38-17.18, ²⁰⁷Pb/²⁰⁴Pb of 15.38-15.48, and ²⁰⁸Pb/²⁰⁴Pb of 37.24-37.83 and plot within a field defined by the Pb isotopic compositions of basement rocks from the lower crust of Dabie and the UHP eclogite and gneiss in the Dabie-Sulu orogenic belt (South China II) (Figure 11) [1, 10]. However, the Xiaoshuhe dikes have lower initial ²⁰⁶Pb/²⁰⁴Pb (16.38-16.85), ²⁰⁷Pb/²⁰⁴Pb (15.38-15.44), and $εNdt$ (-20.6 to -17.8) than samples from the Qibaoshan complex [$206Pb/204Pbi=16.99−17.18$⁠, $207Pb/204Pbi=15.44−15.48$⁠, and $εNdt=−15.9to−14.6$].

Most magmas undergo crustal contamination and/or fractional crystallization when they rise to the surface from the mantle through continental crust. These magmatic processes must be identified and taken into account in order to constrain the nature of the source. Generally, crustal contamination of magmas is expected to result in a positive correlation between (⁸⁷Sr/⁸⁶Sr) _i and SiO₂. The lack of both a positive correlation between SiO₂ and (⁸⁷Sr/⁸⁶Sr) _i ratios [22] in the studied rocks (Figure 12) and a negative correlation between SiO₂ and $εNdt$ indicate that crustal assimilation is unlikely to have occurred. Furthermore, no inherited zircon has been observed in the studied rocks, which supports the idea that the magmas have not been significantly contaminated by crustal materials during ascent [28, 29, 33, 34]. Therefore, it is inferred that crustal assimilation is negligible in the magma evolution of the studied rocks.

During partial melting, the ratio of La/Sm increases along with La; however, during fractional crystallization, La/Sm ratios remain constant, while La increases [53]. Figure 13 shows that the La/Sm ratio only increases slightly as La increases, suggesting that fractional crystallization might play a more important role than partial melting in the magmatic evolution of the studied rocks. Furthermore, the negative correlations between SiO₂ and MgO, CaO, TiO₂, and P₂O₅ reflect significant fractionation (Figure 8). The samples show slight negative Eu anomalies, indicating that the rocks have undergone minor fractionation of plagioclase (Table 1 and Figure 9).

As discussed above, crustal contamination is negligible in the studied samples. The magmatic rocks in the Wulian area are characterized by moderate SiO₂ (50.0-60.0 wt.%), MgO (3.3-5.6 wt.%), Cr, and Ni contents. This suggests that they were derived from a mantle source rather than the crust, as partial melting of any of the crustal rocks and lower crustal intermediate granulites in the deep crust would produce high SiO₂ (> 58 wt.%) and low MgO (< 4 wt.%) liquids [61–65]. However, the rocks in these two regions are characterized by enrichment in LREE and LILE, depletion in HFSE, and variable Na₂O (0.2-4.4 wt.%), suggesting that continental materials (granitoids, granulites, and sediments) were involved in the petrogenesis of these mantle-derived magmas. This is further supported by the high (⁸⁷Sr/⁸⁶Sr) _i (0.7079-0.7088) and low $εNdt$ values (-20.6 to -14.6) of the samples (Table 2 and Figure 10). The Sr-Nd isotopic compositions of intermediate volcanic rocks from Wulian area are different from the lower or upper crust of the NCC and YC (Figure 10), but are similar to those of the contemporaneous mafic dykes from the Sulu orogenic belt, suggesting that they have the similar magma sources [20–22, 29, 66].

Therefore, the intermediate rocks in the Wulian area are likely to be derived from the lithospheric mantle with a contribution from crustal materials, and this is consistent with the results of Lan et al. (2011). However, the source of these crustal materials is still controversial and has variously been ascribed to (1) subducted YC [41, 67, 68]; (2) delamination of the lower crust of NCC [22–24, 69–71]; and (3) fluid/melt derived from the subduction of the Palaeo-Pacific plate [20, 33, 34, 54, 72, 73].

The rocks have low Th/Yb ratios (3.15-21.41) and a range of Ba/La ratios (12.3-38.1) (Figure 14), suggesting that the origin of the crustal materials is either the subducted YC or the delamination of the lower crust of the NCC, rather than the fluid/melt derived from the subducted Palaeo-Pacific plate [33, 34, 74–77]. In addition, Pb isotopic ratios are significantly lower than the EMII-mantle endmember (usually associated with subducted seafloor sediments). This also indicates that the crustal materials in the magma source were unlikely to be derived from the Paleo-Pacific oceanic plate [78].

Pb isotope signatures can also be effective in distinguishing the origin of crustal materials [1, 21]. Mesozoic granitoids of the NCC are characterized by low initial Pb isotopic ratios, whereas the Mesozoic granitoids of the YC have high Pb isotopic ratios [21, 30, 79, 80]. The Early Cretaceous magmatic rocks in the Wulian area have relatively high initial Pb isotope ratios, similar to the lower crust of the North Dabie and South China II basement material, but distinct from the lower crust of the North China Craton (Figure 11). Therefore, it appears that the Early Cretaceous magmatic rocks in the Sulu orogenic belt have incorporated crustal material from the YC. Additionally, Nb/U ratios of magmatic rocks tend to reflect the characteristics of the source region [73]. The Nb/U ratios (0.8-11.8) in our samples are lower than the lower crust of the NCC (⁠$Nb/U=25$⁠, Rudnick and Gao (2003)) [81], indicating that the crustal materials have not been derived from the delaminated lower crust of the NCC [33]. The two-stage Hf model ages of zircon Hf isotopes are Paleoproterozoic (2.49-2.04 Ga) (Table S2, Figure 5), which are similar to those of the UHP metamorphic rocks in the Sulu orogenic belt [82, 83], also suggesting that the continental crustal materials are derived from the YC.

The age of continental crustal accretion can be used to distinguish whether the crustal materials originated from the NCC or the YC [10]. The NCC magmatism mainly occurred in the Neoarchean to Paleoproterozoic (2700-1800 Ma), with only minor Neoproterozoic magmatism in the southeastern NCC [21, 84, 85]. In contrast, the YC is characterized by widespread Neoproterozoic (860-635 Ma) magmatism, with Archean and Paleoproterozoic magmatism recorded in the northern and southwestern YC [21, 22, 83]. A large number of Neoproterozoic (542-764 Ma) inherited zircons in the Early Cretaceous magmatic rocks from the Sulu orogenic belt indicate that lower Yangtze crustal materials were involved in the Early Cretaceous magma source of the Sulu orogenic belt [21, 29, 30, 34]. In addition, low δ¹⁸O values in the Early Cretaceous granites indicate that the sources of mafic rocks from the Sulu-Dabie orogenic belt contained YC continental crust [30, 41].

In summary, the Early Cretaceous magmatic rocks were mainly derived from the lithospheric mantle with a contribution from subducted YC continental crust [21, 28–30, 51, 67, 79, 80].

As shown above, the magma source of the Early Cretaceous magmatic rocks from the Wulian area includes both crustal and mantle materials. However, the mechanisms of crust-mantle interaction remain equivocal [10, 20–25]. It has been proposed that the old lithospheric mantle of the NCC was metasomatized by fluid/melt derived from the subduction of the YC, forming enriched lithospheric mantle [20, 21, 28–30]. Yang et al. [80] further developed this idea and proposed that the degree of fluid/melt metasomatism in southeastern NCC was related to the subduction distance of the YC. In other words, the closer to the Sulu orogenic belt, the higher the initial Sr and Pb isotopic ratios in the rocks. However, the studied rocks closer to the Sulu orogenic belt have lower initial Sr and Pb isotopic ratios than Mesozoic magmatic rocks in the Tanlu fault (Figures 10(a) and 11), suggesting that the enrichment mechanism of the Early Cretaceous magma source in the Wulian area might have not been due to metasomatism by the subducted YC. Furthermore, on the basis of major and trace elements, Meng et al. [85] proposed that andesitic rocks from the Lingshan island were derived from an enriched lithosphere mantle, mixed with melts of the Yangtze crust. Based on oxygen isotopes in the granites of the Sulu orogenic belt and the characteristics of inherited zircon, Zhao et al. [86] proposed that these rocks were derived from partial melting of the lower crust of the Yangtze.

Metamorphic dehydration and partial melting of crustal materials in subduction channels are related to the P-T conditions during subduction [2, 87–92]. Subducting crustal rocks may not experience obvious dehydration at shallow depths (<80 km) but become significantly dehydrated and even partially melted at deeper depths (80-130 km) in cold continental subduction zones [10, 93]. The lower crust of the YC beneath the Sulu orogenic belt was trapped in the lithospheric mantle at shallow depths (<80 km), so it is difficult to produce fluid/melt from the YC and metasomatize the lithospheric mantle [2, 10]. However, the subduction of the Pacific plate induced lithospheric extension as well as upwelling of asthenosphere, leading to partial melting of the lower crust of the YC and the lithospheric mantle of the NCC [94–96].

Therefore, we propose that the magma sources of the Wulian area were formed by the mixing between melts of the lower crust of the YC and the lithospheric mantle of the NCC. A similar model was proposed by Fan et al. [68] who studied the Sr-Nd isotopes of Early Cretaceous magmatic rocks in the Sulu-Dabie orogenic belt. In order to test this, we used Sr and Nd isotopes to model mixing between the lower crust of the YC and the Paleozoic lithospheric mantle of the NCC (Figure 10(b)). It can be seen that Sr and Nd isotopic compositions of the Early Cretaceous magmatic rocks in the Sulu orogenic belt lie on a mixing line between Paleozoic lithospheric mantle and the lower crust of the YC. In contrast, the magmatic rocks located near the Tanlu fault zone are more likely to be derived from enriched lithospheric mantle which had been metasomatized by fluids derived from the YC slab. It is worth noting that the Xiaoshuhe dikes in southern Wulian area have lower initial Nd isotopic compositions than the Qibaoshan complex, which suggests that more YC materials were involved in the mantle source of the Xiaoshuhe dikes. It is clear that the intermediate magmatic rocks in the Wulian area can be explained by mixing between melts of the Paleozoic lithospheric mantle of the NCC and lower crust of the YC.

The main way in which crustal materials are recycled into the mantle is by plate subduction [2, 10]. However, delamination can also play a role [10]. In addition to the subduction of oceanic crust, Chopin et al. [97] proposed that continental crust can be subducted into the mantle. Over the years, the model of oceanic subduction channel has been extended to continental subduction zones and proposes that the deep subduction of continental crust can lead to the UHP metamorphism and significant crust-mantle interaction in continental subduction channels [2, 98, 99]. The extension of the subduction channel model to continental subduction zones is of particular significance to the study of the physical and chemical processes during continental crustal subduction [2, 10]. Subduction channel processes involve the following: (1) deformation, dehydration, and detachment; (2) recrystallization, growth, and melting; and (3) alteration, metasomatism, and anataxis [2].

The Wulian area experienced a complex tectonic evolution in the Mesozoic. Before large-scale Early Cretaceous magmatism, it is likely to have been subjected to subduction metasomatism by the Paleo-Asian plate in the Paleozoic [68] and the subduction transformation of the YC in the Triassic [21, 28–30, 80]. Based on the results and discussion presented above, we have identified the characteristics of crust-mantle interaction in the Wulian area. Before 220 Ma, the YC subducted deeply below the northwestern NCC, and the spatial extent of the influence of subduction on the adjacent NCC is about 200 km [1, 10, 20, 21, 28, 29], with subduction depth of more than 120 km (Figure 15(a)) [11–13]. At subduction depths shallower than 80 km, the YC slab would not have undergone dehydration or melting to metasomatize the overlying lithospheric mantle (Figure 15) [2, 10]. However, dehydration or melting would occur when the slab reached a depth of greater than 80 km in the southeastern NCC, and the old lithospheric mantle of NCC would be metasomatized by fluid/melt (Figure 15(a)). Between 130 and 110 Ma, the subduction of the Pacific plate induced lithospheric extension as well as upwelling of asthenosphere, leading to partial melting of the lower crust of the YC and the lithospheric mantle of NCC and the development of large-scale magmatism [1, 10, 20–24, 66, 100]. In the Wulian area and southern Sulu orogenic belt, mixing occurred between melts of both the lithospheric mantle of NCC and the residual lower crust of the YC (Figure 15(b)). In contrast, in the southeastern NCC, the magma was derived from the lithospheric mantle that had been enriched by metasomatic fluids derived from the YC slab [20–24, 51, 58] (Figure 15(b)).

• (1)

  Early Cretaceous intermediate rocks are widespread in the Wulian area and are mainly composed of trachy-andesites, gabbros, dolerites, and diorite porphyrites. Zircon U-Pb dating shows that these rocks formed at 123-124 Ma

• (2)

  These Early Cretaceous magmatic rocks belong to the high-K calc-alkaline series, which are enriched in LILE such as Rb and Ba, depleted in HFSEs such as Nb, Ta, Zr, and Ti, and enriched in Sr-Nd-Pb isotopes, suggesting that these rocks were derived from a lithospheric mantle source containing crustal materials

• (3)

  Formation of the Early Cretaceous magmatic rocks in the Wulian area is likely to have occurred in an extensional tectonic setting resulting from continuous slab rollback of the Pacific plate. These rocks were generated by mixing between melts derived from the lithospheric mantle of the NCC and the residual lower crust of the YC

The authors confirm that the data supporting the findings of this study are available within the article and the supplementary materials.

The authors declare that they have no conflicts of interest.

We thank Shuangrong Zhang for help with the LA-ICP-MS zircon U-Pb and Lu-Hf analyses in Beijing GeoAnalysis Co., Ltd. (Beijing, China), and Baoming Ding for help with major and trace elements and Sr-Nd-Pb isotopic compositions analyses in institution of geology and Geophysics, Chinese Academy of Sciences. This study was supported by the Marine S&T Fund of Shandong Province for Pilot National Laboratory for Marine Science and Technology (Qingdao) (2021QNLM020001-1); the Shandong Provincial Natural Science Foundation, China (ZR2021MD083); the National Natural Science Foundation of China Project (42072169; 41302102); and the Graduate Innovation Project of China University of Petroleum (East China) (YCX2021020). The manuscript benefited from the constructive comments of three anonymous reviewers.