New identification and significance of Early Cretaceous mafic rocks in the interior South China Block

Early Cretaceous mafic rocks are first reported in the northern Guangxi region from the western Qin-Hang belt in the interior South China Block. A systematic investigation of zircon U–Pb dating, whole-rock geochemistry, Sm–Nd isotopes and zircon Hf–O isotopes for these mafic rocks reveals their petrogenesis and the mantle composition as well as a new window to reconstruct lithospheric evolution in interior South China Block during Late Mesozoic. Zircon U–Pb dating yielded ages of 131 ± 2 Ma to 136 ± 2 Ma for diabase and gabbro from Baotan area, indicating the first data for Early Cretaceous mafic magmatism in the western Qing-Hang belt. These mafic rocks show calc-alkaline compositions, arc-like trace element distribution patterns, low zircon εHf(t) of − 9.45 to − 6.17 and high δ18O values of + 5.72 to + 8.09‰, as well as low whole-rock εNd(t) values of − 14.27 to − 9.53. These data suggest that the studied mafic rocks are derived from an ancient lithospheric mantle source that was metasomatized during Neoproterozoic subduction. Thus, the occurrence of these mafic rocks indicates a reactivation of Neoproterozoic subducted materials during an extension setting at Late Mesozoic in the western Qin-Hang belt, an old suture zone that amalgamates the Yangtze and Cathaysia blocks.

www.nature.com/scientificreports/ occur as stock and small intrusion that intruded into the Sibao Group ( Supplementary Fig. S1d). Different from the widespread Neoproterozoic mafic rocks that have been subjected to variable degrees of alteration and/or mineralization in this region 10,11 , the mafic rock samples collected in this study are relatively fresh and suffered only very slightly alteration. Mandong and Gaobang gabbros exhibit typical gabbro textures, mainly composed of 40% clinopyroxene, 50% plagioclase, 5% amphibole and 5% quartz ( Supplementary Fig. S2a,b). Accessory phases mainly include apatite, titanite, and rutile. Mandong and Wende diabases are of porphyritic textures with phenocrysts of clinopyroxene, plagioclase and rare amphibole ( Supplementary Fig. S2c,d). The matrix is microgranular and comprises dominant plagioclase, biotite, clinopyroxene, and minor K-feldspar and quartz. Minor clinopyroxene phenocrysts were locally replaced by an alteration assemblage of biotite ( Supplementary Fig. S2d).

Results
Zircon U-Pb dating. Cathodoluminescence (CL) images of representative zircon grains from selected mafic rock samples in the northern Guangxi region are shown in Fig. 2. The Secondary ionization mass spectrometry (SIMS) and laser ablation inductively coupled plasma-mass spectrometry (LA-ICP-MS) zircon U-Pb dating results are listed in Supplementary Table S1 and shown in Fig. 2.
Fifteen analyses were carried out from Mandong (MD-15) and Wende (WD-1) diabase, respectively. An analysis point 9 on a whitish zircon from the sample MD-15 gives a youngest 206 Pb/ 238 U age of 81 ± 1 Ma, which may be caused by lead loss. Thus, this analysis was excluded from the final age calculation. Zircons from the Mandong diabase yield a weighted mean 206 Pb/ 238 U age of 131 ± 2 Ma (MSWD = 1.5, n = 9), those from the Wende diabase show a weighted mean 206 Pb/ 238 U age of 131 ± 2 Ma (MSWD = 1.7, n = 11). Older zircons (156-2246 Ma) (Fig. 2a,c) are also determined in both samples which indicate a detrital origin that originated from assimilation of wall rocks during magma ascent or inherited grains from the magma source region 12,13 . The Gaobang gabbro also show a wide range of zircon ages with nine youngest analyses forming a tight cluster and give a weighted mean 206 Pb/ 238 U age of 136 ± 2 Ma (MSWD = 2.7) (Fig. 2e). Therefore, the age data from the above studied samples indicate an early Cretaceous age for the mafic magmatism in northern Guangxi, which is the first report of Late Mesozoic mafic rocks in this region.
Major and trace element concentrations. The Table S2). Plot of Na 2 O + K 2 O vs. SiO 2 shows that these samples belong to subalkalic gabbro and gabbroic-diorite (Fig. 3a). On the FeO T /MgO versus SiO 2 diagram (Fig. 3b), all samples plot in the calc-alkaline field.
These mafic rocks have lower total rare earth element (REE) contents (49.0-96.1 ppm) (Fig. 4a-c). They display broadly similar REE pattern showing an enrichment in light rare earth elements (LREE) and flat heavy     Table S4). The oxygen isotope data in this study are significantly higher than the typical mantle value (+ 5.3 ± 0.3‰) 16 .

Discussion
New mafic magmatic event in the western Qin-Hang belt of interior South China Block. The mafic rocks in Baotan area from the Northern Guangxi region of the western QHB are traditionally thought to be contemporaneous with widely-exposed Neoproterozoic peraluminous granitoids 9,11 . Li et al. 7 reported four mafic dykes that intruded into the Neoproterozoic Sibao Group with ages of 820-830 Ma (mean age 828 ± 7 Ma) by sensitive high-resolution ion microprobe (SHRIMP) zircon U-Pb method. Wang et al. 11 conducted LA-ICP-MS zircon U-Pb dating for a layered diabase with an age of 812 ± 5 Ma. Weighted average 206 Pb/ 238 U ages of 855 ± 6 Ma and 835 ± 9 Ma for gabbros were obtained by LA-ICP-MS zircon U-Pb dating by Yao et al. 15 and Chen et al. 9 , respectively. These ages indicate significant mafic magmatism at 855-812 Ma but also led to a widely-accepted assumption that the mafic rocks in the Northern Guangxi of the western QHB are all formed in Neoproterozoic period. www.nature.com/scientificreports/ During our field work, we noticed some mafic intrusions which are much less weathered than those Neoproterozoic mafic rocks. Three samples from the intrusions were analyzed in different laboratories with different U-Pb dating methods. They gave consistent Early Cretaceous ages ranging from 136 ± 2 Ma to 131 ± 2 Ma (Fig. 2). Thus, we believe that these Early Cretaceous ages are reliable that may reveal a previous undiscovered mafic magmatism event in this region. We noticed that Wang et al. 17 has reported a weighted average age of 100 ± 14 Ma for a lamprophyre in the adjacent region of Baotan in Northern Guangxi. Although this age is also Early Cretaceous, but the accurate data still need to be assessed because only two data points show ages of ~ 100 Ma among 30 analyses 17 . Therefore, our data have, for the first time, confirmed the presence of Early Cretaceous mafic rocks in the North Guangxi of the western QHB in spite of their small volume and unclear spatial distribution pattern.
Crustal contamination and fractional crystallization of the mafic magma. Although we observed the presence of some minor secondary minerals in some samples, most of the samples have relatively low loss on ignition (LOI) values of less than 3.0 wt%, and no obvious correlation is shown between the LOI and Nb/ La and Th/La ratios, which suggests that the effect of post magmatic alteration is not significant for these studied samples. The petrographic evidence for zircon xenocrysts in the mafic rocks emphasizes the importance of crustal assimilation during magma ascent. In addition, on the ε Nd (t) versus t diagram (Fig. 5), the studied mafic rocks exhibit negative and variable ε Nd (t) values of − 14.27 to − 9.53 and plot in the Proterozoic SCB crust field. The considerable variation of negative ε Nd (t) values suggests effective contamination of the mafic rocks by crustal materials during magma emplacement. The effect of crustal contamination and assimilation fractionation crystallization (AFC) can be evaluated by correlations between major elements (such as MgO, SiO 2 and FeO T )   [18][19][20] . Generally, La/Sm ratios decrease and ε Nd (t) values increase with increasing Nb/La ratios (Fig. 7a,b). In addition, clear positive correlations are found in the bivariation diagrams of Nb/La and ε Nd (t) versus MgO (Fig. 7c,d), similar to AFC trends. These lines of evidence indicate variable crustal contamination during their emplacement en route to the continental crust.
As presented in Fig. 6, the Wende mafic rocks show a relatively limited range of zircon ε Hf (t) values of − 9.36 to − 6.36 and δ 18 O values of + 7.03‰ to + 7.89‰. Although the Mandong samples exhibit relatively variable δ 18 O values from + 5.72‰ to + 8.09‰ for zircons, they have similar zircon ε Hf (t) values (− 9.45 to − 6.17) to the Wende samples. Furthermore, no covariation of Hf-O isotopes for the zircon grains from the Mandong and Wende mafic rocks could be observed (Fig. 6c). These facts imply that the effect of an AFC process on the Hf and O isotope composition of the zircon crystals is likely negligible.
Large variations in MgO (4.99-9.81 wt%) and compatible elements such as Cr (125-635 ppm) and Ni (5.98-125 ppm) indicate fractional crystallization of the parental magmas to varying degrees. Cr and Ni decrease sharply and SiO 2 increases with decreasing MgO, showing significant fractionation of olivine and clinopyroxene ( Fig. 8a-c). Rapid decreases in FeO T and TiO 2 with decreasing MgO at MgO > 7.0 wt% imply that Fe-Ti oxide were involved in the fractional phase at MgO > 7.0 wt% (Fig. 8d,e). The presence of a positive correlation between CaO/Al 2 O 3 and CaO demonstrates that plagioclase fractionation is likely (Fig. 8f).
To minimize the effect of assimilation and fractionation crystallization, only the least contaminated and least evolved samples are chosen to constrain the source region. Among the mafic rocks in Baotan area, Mandong gabbros and diabases have the lowest SiO 2 (50.86-54.40 wt%) and highest ε Nd (t) values (− 9.53 to − 11.23). Their Mg # values range from 65 to 68 and therefore these samples are suggested to approximate to the compositions of primary melts.
Magma source. All the mafic rocks in the Baotan area are characterized by relatively high MgO, Cr and Ni (up to 9.81 wt%, 635 ppm and 125 ppm, respectively), suggesting that these rocks originated from a lithospheric mantle source 21 . The crust-like trace element signatures, such as high LREE and large-ion lithophile elements (LILE) contents, low ratios of Nb/Ta (11.57-13.27), Ce/Pb (3.10-5.01) and Nb/U (3.99-5.53), indicate a great contribution of crustal materials in magma sources and/or during subsequent ascending processes [22][23][24] . We conducted quantitative modeling to evaluate the contribution of crustal materials based on Nd isotopic compositions by the ( 143 Nd/ 144 Nd) i -Nd plot (Fig. 9). The compositions of the Neoproterozoic metasedimentary rocks from the Sibao and Danzhou Group are taken as the crustal contaminant 25,26 . The simple modelling result suggests, that at least 30-45% crustal materials are required in the generation to achieve the observed Nd isotopic compositions of the studied mafic rocks (Fig. 9). However, the assimilation en route of such high-proportional crustal materials is obviously impossible for maintaining the mafic compositions. Moreover, the ε Hf (t) values of the zircon from these mafic rocks are negative and mostly range from − 9.47 to − 6.17, whereas their δ 18 O values in the range of + 5.72‰ to + 8.09‰, mostly higher than the mantle value (+ 5.3 ± 0.3‰) 16 . Thus, these zircon Hf-O isotopes are enriched and coherent with the Nd isotopic compositions of the least contaminated Mandong mafic rock samples. As mentioned above, the AFC process is unlikely responsible for the low ε Hf (t) values and higher δ 18 O for the zircon crystals in the mafic rocks. Thus, the observed isotopes abnormities of the zircons should be ultimately attributed to the composition of their parental magmas. Therefore, the crust-like geochemical signatures, such as high LREE and LILE contents and significant Nb-Ta-Ti negative anomalies, cannot totally be attributed to crustal contamination en route, and probably, are partially inherited from a mantle source modified by subduction-derived components in either recent or ancient metasomatism 27 . www.nature.com/scientificreports/ Subduction is the major mechanism that carries enriched crustal components into mantle source. During the long evolutionary history of the SCB, the lithosphere mantle beneath it experienced several metasomatic events 28 . Among them, the Neoproterozoic subduction related to the collision between the Yangtze and Cathaysia blocks, or the so-called Jiangnan Orogeny, is the most important 29 .
In this study, we consider that the Neoproterozoic subduction event is most likely responsible for the crustlike components of the studied mafic rocks. Several lines of evidence support this proposal. Firstly, the antiquity of the metasomatic event would require the metasomatized domain to have remained chemically isolated for a long time, which would have allowed a general isotopic homogenization of the magma source(s) 30 . Whereas, the absence of isotopic homogenization within the source(s) usually corresponds to a relatively recent metasomatism 30 . Thus, the Hf isotopic homogeneity in the samples is more consistent with an ancient enrichment event rather than a relatively recent metasomatism in which highly isotopic variability is usually expected to be observed. Furthermore, there is a broad consensus that the Yangtze and Cathaysia blocks, following the northward subduction of an oceanic lithosphere beneath the southeastern Yangtze Block, amalgamated along the Jiangnan Orogen to form a united SCB in the Neoproterozoic time [31][32][33] . Subsequently, the SCB turned into an   [34][35][36] . The uniform low ε Hf (t) values with Hf model ages ranging from 1.0 to 1.2 Ga also support a derivation from a Proterozoic metasomatized source. Finally, both the metasedimentary rocks within the Sibao Group and 860-830 Ma mafic rocks intruding into the lower Sibao Group exhibit arc-like geochemical affinity and these rocks were interpreted as the refractory source with addition of crust-derived components by oceanic subduction in an active continental margin setting 9,31,37 . The regional sedimentary sequences of Sibao and Danzhou Group were deposited at 860-832 and 803-764 Ma and the regional unconformity between them sealed the Jiangnan Orogen and was constrained to be 832-803 Ma, consistent with the presence of 833-822 Ma syn-collisional S-type granites 15,38 . These data with geological observations indicate the development of the early Neoproterozoic continental arcbasin system along the Jiangnan Orogen. Wang et al. 39 and Gan et al. 40 reported Silurian (ca. 430 Ma) arc-like gabbros and volcanics along the Yunkai-Nanling Domain and Early Jurassic olivine gabbros (ca. 191 Ma) in the Eastern Nanling Range, which have been regarded as derivation of an ancient lithospheric source inherited from the Neoproterozoic metasomatized wedge. When recalculated to 130 Ma, the previously reported Neoproterozoic mafic rocks 9,41 in the studied areas exhibit ε Nd (t) values of − 2.2 to − 14.5, covering the ranges of ε Nd (t) values of the investigated mafic rocks in this study (Fig. 5). Thus, we conclude that, the source for the studied mafic rocks was introduced by the subduction-derived components which is originally with high δ 18 O values via the Neoproterozoic subduction process along the western Qin-Hang belt.

Methods
Major, trace element and isotopic analyses. Major and trace element analyses were followed those methods described by Hou et al. 42 and Qi et al. 43 , respectively. In brief, we crushed the rock samples to ~ 200 mesh powder in an agate mill. The analyses were carried out at the National Research Center of GeoAnalysis, Chinese Academy of Geological Sciences, Beijing. Major elements were analyzed by wet chemistry and X-ray fluorescence spectrometry (XRF) with analytical uncertainties ranging from 0.5 to 1.5%. Two standards (granite GSR-1, basalt GSR-3) were used to monitor the analytical quality. Trace elements were determined with a POEMS inductively coupled plasma mass spectrometry (ICP-MS). Analytical uncertainties are 10% for elements with abundances < 10 ppm, and around 5% for those > 10 ppm.
Sm-Nd isotope analyses of whole rock samples followed those procedures described by Chen et al. 44 . The whole rock samples (MD-09, QMS-14, WD-1 and WD-4) were determined by a Finnigan Triton TI thermal ionization mass spectrometer at Tianjin Institute of Geology and Mineral Resources, Tianjin. Procedural blanks yield concentration of < 200 pg for Sm and Nd, and mass fractionation corrections for Nd isotopic ratios were    45 .
The O 2 −· primary ion beam was accelerated at 13 kV, with an intensity of ca. 10 nA. The resulted ellipsoidal spot is about 20 × 30 μm in size. Positive secondary ions were extracted with a 10 kV potential. In the secondary ion beam optics, a 60 eV energy window was used, together with a mass resolution of ca. 5400 (at 10% peak height), to resolve lead isotopes from isobaric interferences. A single electron multiplier was used in ioncounting mode to measure secondary ion beam intensities by peak jumping mode. Each measurement consists of 7 cycles with the total analytical time of ca. 12 min. U-Th-Pb ratios were determined relative to the standard zircon Plešovice ( 206 Pb/ 238 U age of 337 Ma) 46 , and the absolute abundances were calibrated to the 91500 standard zircon (81.2 ppm U and 29 ppm Th) 47 . Pb/U calibration was performed in relation to standard zircon Plešovice and analyses of standards were interspersed with unknown grains. Measured compositions were corrected for common Pb using non-radiogenic 204 Pb. An average of present-day crustal composition 48 is used for the common Pb assuming that it is largely surface contamination introduced during sample preparation. Data reduction was carried out using the Isoplot/Ex v. 3.00 Program 49 . Uncertainties on individual analyses in data tables are reported at 1 sigma level; mean ages for pooled U/Pb analyses are quoted with 95% confidence interval.
U-Pb dating of zircons from Gaobang gabbro (GB2014-1) was performed by LA-ICP-MS at the Key Laboratory of Continental Collision and Plateau Uplift, Institute of Tibetan Plateau Research, Chinese Academy of Sciences in Beijing. A NewWave UP193FX Excimer laser coupled with an Agilent 7500a ICP-MS were used for determination of zircon U-Pb ages and trace element concentrations. The ablation system operated at a wavelength of 193 nm with a spot diameter of 35 µm. Each analysis consists of 15 s background acquisition, 40 s sample data acquisition, and a 45 s washout delay at the end. During the zircon analyses, the standard samples of zircon Plešovice and glass NIST SRM 612 were analyzed first, followed by 8 unknown sample analyses. The standard zircon Plešovice was used for correction of isotope fractionation. Common Pb was corrected following the method proposed by Andersen 50 . U-Pb ages were calculated using the GLITTER 4.0 51 and age plots were carried out using the Isoplot 3.0 program 49 .
Zircon Lu-Hf isotope analysis. In situ Lu-Hf isotopic compositions were measured in previously dated zircon grains at the same domain. Zircon Hf isotope analysis was carried out on a Neptune multi-collector ICP-MS equipped with New Wave UP 213 laser-ablation system at the Institute of Mineral Resources, Chinese Academy of Geological Sciences, Beijing. Instrumental conditions and analytical procedures were comprehensively described by Hou et al. 52 and a summary is present here. A stationary spot was used for analyses, with a beam diameter of 55 μm. Helium was used as a carrier gas to transport the ablated samples mixed with Argon from the laser-ablation cell to the ICP-MS torch via a mixing chamber. The isobaric interference of 176 Lu and 176 Yb on 176 Hf was corrected using ratios of 176 Lu/ 175 Lu = 0.02658 and 176 Yb/ 173 Yb = 0.796218 proposed by Chu et al. 53 . For instrumental mass bias correction, Yb isotope ratios were normalized to 172 Yb/ 173 Yb = 1.35274 53 and Hf isotope ratios to 179 Hf/ 177 Hf = 0.7325 using an exponential law 54 . Our routine run of the zircon standard GJ1 gave a weighted mean 176 Hf/ 177 Hf ratio of 0.281979 ± 0.000009 (2σ, n = 7), which is in good agreement with the reported value (0.282015 ± 0.000019) 55 .
Zircon oxygen isotope analysis by SIMS. Zircon oxygen isotopes were analyzed using the same Cameca IMS 1280 ion microprobe at IGGCAS, following the analytical procedures of Li et al. 56 . A focused Cs + primary ion www.nature.com/scientificreports/ beam was accelerated at 10 kV with an intensity of ca. 2 nA. The analyzed spot is ca. 10 μm in diameter. In order to compensate for sample charging, a normal incidence electron flood gun was used with homogeneous electron density over a 100 μm elliptical area. Oxygen isotopes were measured using multi-collection mode on two offaxis Faraday cups (FC). The instrumental mass fractionation (IMF) during the analysis was corrected using the Penglai zircon standard with a δ 18 O value of 5.25‰ 57 . Repeated analyzes of an in-house standard Qinghu zircon during the courses of sample analysis yielded mean δ 18 O value of 5.30 ± 0.23‰ (2σ, n = 8), which is identical within errors to the reported value (5.39 ± 0.22‰) 58 . We state that our research did not carry out experiments involving human participants including the use of tissue samples. Our manuscript also did not contain information or images that could lead to identification of a study participant. www.nature.com/scientificreports/