Hon Xu, Zhao-pn Ji, Shan-yin Li, Yan-qiou Yan, Shou-ji Liu, Hai-yan Zhan, c, ,Shu-shn Lu, c, , Ton-qian Shi, c, d, Mn Tao, c, , Na Qin, c, , Wi-wi Zhan, c, d, Da-pn Su,Lon-wi Qiu,*

a Qingdao Institute of Marine Geology, China Geological Survey, Qingdao 266071, China

b Key Laboratory of Marine Sedimentology and Environmental Geology, First Institute of Oceanography, Qingdao 266061, China

c Laboratory for Marine Mineral Resources, Qingdao National Laboratory for Marine Science and Technology, Qingdao 266071, China

d College of Earth Science and Engineering, Shandong University of Science and Technology, Qingdao 266590, China

e School of Geosciences, China University of Petroleum (East China), Qingdao 266580, China

f Shandong Provincial Key Laboratory of Depositional Mineralization and Sedimentary Minerals, Shandong University of Science and Technology, Qingdao 266590, China

g Beijing SHRIMP Center, Institute of Geology, Chinese Academy of Geological Sciences, Beijing 100037, China

A B S T R A C T

The Xisha Block is a minor one in the South China Sea and an important tectonic unit in the northwestern part of the region. Zircon SHRIMP U-Pb ages for three volcanic intrusive core samples from Xike-1, an exploratory well penetrating the bioherms of the Xisha Islands. The core samples are from the Miocene reef carbonate bedrock and are recognized as dark-gray biotite-hornblende gabbro, gray fine-grained biotite diorite, and gray fine-grained granite, respectively. Zircon cathodoluminescence (CL) images and trace Th, U and Pb compositions of the zircons show that these rocks are of volcanic intrusive origin. Zircon SHRIMP U-Pb dating yielded six groups of ages, ranging from 2451-1857 Ma to early Cretaceous,which indicate that the formation and evolution of the Xisha Block was affected by the evolution and closure of Neotethys Ocean, probably within its eastern extension into South China Sea. Both old, deepsourced material, including fragments from Rodina supercontinent, and recent mantle-derived magma products contributed to the emergence and formation of the Xisha block. The SHRIMP U-Pb results also proved that this process differed from that of the Kontum massif, the Hainan Block, and the South China Block, but is similar to that of the Nansha and Zhongsha blocks. The process was associated with the effects of Yanshanian magmatism induced by subduction mechanisms of the Paleo-Pacific Plate or the reworking of the multiple magmatisms since the Early to mid-Yanshanian, possibly jointly experienced by the Xisha-Zhongsha-Nansha Block.

Keywords:

Geochronology

Volcanic intrusive rocks

Zircon SHRIMP U-Pb dating

Rodina supercontinent

Xisha Block

South China Sea

1. Introduction

The Xisha Block (XB) is one of the minor blocks in the South China Sea and an important tectonic unit in the northwestern part of this area (Fig. 1). So far, however, little work has been done to document its petrologies and main formation ages. Lack of data or evidence on the relationships of XB rock composition and bedrock genesis with the adjacent blocks has prevented event or age correlation with nearby terranes. The earliest radioactive date for the Xisha islands biohermal bedrock was made 40 years ago by dating gneiss samples from well Xiyong-1, which yielded an Rb-Sr age of 627 Ma (Zeng , 1977). Since then, it has been assumed that the XB basement formed in the Neoproterozoic (Zeng, 1977;Sun JS, 1987). Sun JS, (1987) re-dated one gneiss sample from Xiyong-1 using the K-Ar method and concluded that the XB basement probably first appeared in the time interval between 96.3±1.2 Ma and 627 Ma. Liu YX and ZhanWH,(1994) reported an date of 1465 Ma derived from unrecorded samples using the Rb-Sr isochron method, which so far is the oldest basement age reported. This date is still much younger than the 2300 Ma reported for the nearby Kontum Massif(Hutchison CS, 1989). Recent zircon LA-ICM-MS dating for the granitic gneiss in Xike-1 revealed that the XB basement formed in 144-158 Ma, suggesting it is largely the product of Late Jurassic magmatism in the region (Xiu C et al., 2016).Collectively, these reported ages suggest that the XB basement most likely began to form during the Precambrian.

In recent times, new technology has become the most authoritative and most popular method for isotope dating (Wu et al, 2007). In this study, we introduce new data on gabbro, diorite, and granite samples from the cores of well Xike-1 drilled in the Xisha Islands. Using the modern technology, zircons from these rocks were analyzed for trace element compositions and ages using the SHRIMP U-Pb dating method.We attempt to understand the evolution of the Xisha Block since the Paleoproterozoic and compare its evolution with the adjacent Kontum Massif, and Hainan and Nansha blocks (Fig. 1a).

2. Geologic setting

Geographically, the Xisha Islands lie in the South China Sea between 15° 40' N, 17° 10' S and 111o E and 113o E (Fig. 1).This area is adjacent to the Hainan Island continental shelf and the Qiongdongnan basin to the northwest, the Huaguang depression of the Qiongdongnan and Zhongjiannan basins to the west and southwest, and the Xisha trough/basin to the north, east and northeast (Xu H.et al., 1998). Being part of the continental slope deep-water area in the northwestern South China Sea, this area exhibits a large carbonate-biohermal sedimentary platform standing 1300 m above the abyssal plain with surrounding water depths of 1500-2000 m. The Xisha islands are characterized by a thinned continental crust with Moho depths of 26-28 km and a basement structure intensively reshaped by Mesozoic-Cenozoic subduction and seafloor spreading activities (Huang HB. et al., 2011a, 2011b).As one of the many blocks in the South China Sea, the XB has been linked genetically to the Zhongsha-Dongsha, Nansha and Beibalawang blocks to the east or southeast (Li JB.,1999), and neighbors Kontum Massif and Hainan Block in the west and northwest formed into a united block in the geological past. The XB is one of the key components in the blockbasin-structural belt geological system in the South China Sea. In 1974, China drilled Xiyong-1 on Yongxing Island in the Xisha Islands, the first well penetrating the XB. Drilling to a depth of 1384.6 m, it penetrated through bioherms at 1251 m and revealed underlying bedrock of weathering crust and light to dark gray gneissose granite in the interval of 1251-1384.6 m(Qin, 1987). During 2012-2014, China drilled Xike-1, another continuously coring well penetrating the Xisha Islands bioherms, finalized at a depth of 1268.02 m. The petrology in this well comprises reef carbonates from 0-1257.52 m and basement volcanic intrusive rocks below that interval.Our samples were taken from the bedrocks from the Xike-1 cores.

Fig. 1. Sketch showing the geological setting of the Xisha islands in the South China Sea (modified from Zhao et al., 2016) a: Shelf basins on the northern continental margin of South China Sea; b:Sketch showing the distribution of islands and reefs in Xisha waters,in which the isocline represents the depth contour of the sea water.

As noted above, the study of XB ages started 40 years ago with zircon age analyses of granitic gneiss from Xiyong-1 using Rb-Sr dating method, suggesting the presence of Precambrian basement in Xisha block (Zeng, 1977; Liu YX and Zhan WH, 1994). Xiu C et al., (2016) has now dated the volcanic intrusive rocks from Xike-1 using the zircon U-Pb method and found Mesozoic ages (144-158 Ma) more comparable with dates of 127-159 Ma for the age of the Nansha block,possibly indicating the influence of Jurassic magmatism on both blocks induced by the subduction of Paleo-Pacific plate(Yan, 2008; Yan QS et al., 2010).

Work has been done on the petrologic compositions and the evolution of the neighboring Kontum Massif rocks, including amphibolite-granulite metamorphic facies and volcanic intrusives (Fig. 1). On the basis of similar petrologic compositions between the Kontum Massif and the Antarctic and Indian continents, the basement of the former was thought probably to have commenced to form during the Archean,which is supported by Nd model ages but lacks evidence from other radioactive dating (Hutchison CS, 1989; Lan CY et al.,2003). Over the past 20 years, many authors have conducted detailed investigations on the petrology and ages of the Kontum Massif and yielded early Proterozoic and other ages(Nam T et al., 2001; Nakano N et al., 2013; Chen JF et al.,2014), including 2700-1200 Ma using Sr-Nd and Sm-Nd dat-ing (Lan CY et al., 2003), 2300 Ma from Pb isochron age(Hutchison CS, 1989), 1810-1650 Ma from K-Ar dating(Hutchison CS, 1989), 1780±5 Ma from zircon SHRIMP and LA-ICP MS U-Pb dating (Chen JF et al., 2014), 1530-1340 Ma from monazite U-Th-Pb dating (Nakano N et al., 2013),1403±34 Ma from U-Pb concordia age of zircons (Nam T et al., 2001), 1102-903 Ma from zircon Hf model age (Hieu O et al., 2016), and 976 Ma from zircon U-Pb dating (Hieu O et al., 2016). Data inconsistency suggests that (1) an authoritative method should be determined for dating the basement of the Kontum massif, and (2) zircon SHRIMP and LAICPMSU-Pb dating results are probably more reliable than other types of data, since the earlier geological ages are most concentrated in the Mesoproterozoic, including Neoproterozoic, but also Ordovician, Permian and Cretaceous, recording multiple magmatisms, thermal events, and uplift-cooling events during the Phanerozoic (Carter A, 2001; Osanai Y et al., 2001; Nagy E et al., 2001; Nam T et al., 2001; Lan CY et al., 2003; Lepvrier C et al., 2004; Osanai Y et al., 2004;Maluski H et al., 2005; Owada M et al., 2006; Nakano N et al., 2007; Roger F, 2007; Nakano N et al., 2013).

Reported oldest ages for the Hainan block are largely based on Sm-Nd dating. Ages for this block derived from SHRIMP and LA-ICP U-Pb dating are mostly in the early Proterozoic (Xu DR et al., 2006). Zircon SHRIMP U-Pb dating of granitic rocks of the Baoban Group in the Gezhen area of northwestern Hainan Island yielded median (concordia)ages of 1455±12 Ma and 1454±12 Ma (average age). Existence of these old rocks possibly indicates that Cathaysia (including Hainan Island) used to be part of the Proterozoic Laurentia continent (Xu DR et al., 2006). Xu DR et al. (2007)dated zircons from the Mesoproterozoic-Neoproterozoic Shilu Group in northwestern Hainan Island and early Paleozoic metasedimentary rocks in central-eastern Hainan Island using the SHRIMP U-Pb method, and revealed ages of 960-1300 Ma and 470-514 Ma for the two areas, respectively. These age data tend to indicate that the central-eastern and northwest Hainan Island underwent different continental growth and accretion histories (Xu DR et al., 2007). Caledonian orogeny seems to have had less effect on Hainan Island; magmatic rocks are therefore less widely distributed except for small amounts in central Hainan: single-grain zircon Pb-Pb dating by Fu JM and Zhao ZJ (1997) indicated that the Caledonian magmatism occurred at 369 ±2.9 Ma in Hainan Island. The Hercynian-Indosinian and Yanshanian orogenies had strong impacts on this area and produced a large volume of volcanic intrusive rocks with ages mainly between 255-213 Ma, 220-180 Ma, 176-151 Ma, and 120-80 Ma (Li SX et al., 2005; Xie CF et al., 2005; Wen SN et al., 2013; Liang FG, 2013). Taken together, these previous age estimates suggest that tectonic movements and magmatisms in and around the South China Sea occurred during the Precambrian, Caledonian, Hercynian-Indosinian and Yanshanian orogenies. Products of Precambrian tectonic activities and associated magmatisms and metamorphisms are largely found in northwest Hainan.

3. Methods

In this study, volcanic intrusive rock samples were donated by the Core Laboratory of PetroChina Limited Company Zhanjiang Branch. These samples were collected at 2 m intervals, described onsite and cut for further analysis. Thin sections were made for each sample in the Qingdao Institute of Marine (QIM) Geology of China Geological Survey. Petrographic characterization and identity for the samples were conducted at QIM. For radioactive dating, carefully selected samples were crushed into powders, tabled, washed, electromagnetically separated and heavy-fluid concentrated to separate heavy minerals. Separated zircon grains were further sorted under a binocular microscope and only grains with high transparency and intact shape (without fractures) were chosen for dating analysis. The finally selected zircon grains were embedded in epoxy resin and then cut, ground and polished to expose the center of the zircons for dating. The zircon grains were imaged under cathodoluminescence microscope at QIM.Zircon SHRIMP U-Pb dating was carried out at the Beijing SHRIMP Center, Chinese Academy of Geological Sciences.International standard SL13 (417 Ma) was used for U, Th, and Pb contents. During the dating, SL13 was run for every 4 point analysis of Xisha Islands samples. The measured204Pb content was used to calibrate the level of common Pb. The error of each analysis is 1σ; the error for the weighted average value of the sample age is 2σ. The measured data were calibrated at the SHRIMP Center and processed on Isoplot (Ludwing, 2001).

4. Results

4.1. Sample characters and classification

The igneous rocks below the biohermal-carbonate rocks(257.52 m depth) in Xike-1 consist of different petrologies.Three samples were collected at 2-m intervals. Hand specimen characterization, petrographic characterization and petrological identification were performed. Zircon separation and SHRIMP U-Pb dating results are presented below.

4.1.1. The first sample

Depth: 1257.52-1258.62 m (Fig. 2: a, b, and j).

Core round trip [explain]: No. 892 (1/2-2/2), 2 boxes of samples, length 1.1-1.08 m.

Hand specimen description: Specimen No. 86. The core is iron black to dark black at the top but gets lighter down to the bottom. The core shows fine-grained texture with abundant cracks and calcite veins. The core is identified as gneiss to granitic gneiss. This round trip features high variation at the bottom, which is granite gneiss around 20 cm thick with obvious irregularly-shaped calcite veins (Fig. 2).

Fig. 2. Photographs showing the No. 892 round-trip gabbro (a, b, j); No. 893 round-trip diorite (c, d, e, k); and No. 894 round-trip quartz granite (f, g, h, l) in Xiyu-1 drill cores (explain round-trip).

Sample character: this sample was collected at the right left of the core box and immediately below the biohermal carbonate. The sample is 3.5 cm wide and 8 cm long with 1257.52-1257.60 m in depth.

4.1.2. The second sample

Depth: 1258.62-1260.58 m (Fig. 2: c, d, e, and k).

Core round trip: No. 893 (1/3-3/3), 3 boxes of samples.

Hand specimen description: Specimen No. 87. The core is generally dark to dark gray but is somewhat light or variegated at the top. The rock is dense with fine-grained texture. It is identified as amphibolite, gneiss, and granitic gneiss on site.

Sample character: the size of the sample is 3.5 cm×8 cm.This is the second volcanic rock core. It was collected 2 m from the first core and from the depth 1259.52-1259.60 m.

4.1.3. The third sample

Depth: 1260.58-1263.22 m (Fig. 2 f, g, h, l).

Core round trip: No. 894 (1/4-4/4), 4 boxes of samples.

Hand specimen description: Specimen No. 90. Apparent,dark-colored granitic gneiss appeared at the depth of 1262.8 m.It is distinct from the underlying light-colored granite. The former is very dark while the latter is obviously lighter. Its rock type is also significantly different from the rocks where the first and the second cores were collected. Again abundant calcite veins were found in the core. It was identified onsite as granite.

Sample character: The size of the sample is 3.5 cm×13 cm.The sample is collected from the depth of 1261.52-1261.65 m.

4.2. Petrography and petrological identification

Different rock types were discovered after section identification and professional identification of the three core samples.

4.2.1. The first sample

Rock-forming minerals of the sample were identified in thin sections under the microscope. Common minerals in the first sample include plagioclase, pyroxene, biotite, and hornblende (Fig. 3). Physical properties and contents of these minerals are described below.

Fig. 3. Core identification photograph of volcanic rocks closely attached to biohermal sedimentary rocks from XIKE-1: gabbro. (plagioclase (P1); hornblende (Hb1); biotite (Bi); quartz (Qz)).

Plagioclase feldspar (Pl): accounts for 40%-45% of the whole rock; euhedral to subhedral, polysynthetic and simple twins, low positive relief. Two groups of mutually perpendicular parting are detected in part of the feldspar; sericitization and epidotization are frequently observed. There is 57% An content, classified as labradorite with grain size in the 0.2-0.8 mm range.

Pyroxene (Py): accounts for 20%-25% of the whole rock;columnar, long platy, and granular in shape; no color in plane polarized light. Two groups of mutually perpendicular cleavage are detected, high positive relief, first-order gray interference color, grain size 0.2-0.6 mm, low interference color,classified as possibly orthopyroxene.

Biotite (Bi): accounts for 12%-15% of the whole rock;plate to sheet in shape, plane (010) showing reddish-brown to light yellow, apparent pleochroism and absorbent, cross section showing reddish-brown but without pleochroism, second to third-order interference color in crossed nicols, partial chloritization, 0.2-1.2 mm along long axis.

Hornblende (Hb1): accounts for 12%-15% of the whole rock; granular to columnar in shape, cross section showing regular hexagon, apparent cleavage, grass green to light yellow in plane polarized light, apparent leochroism and (strong)absorbent, second-order interference color, grain size 0.2-0.8 mm; a few well-crystallized grains with size <0.2 mm are observed.

Rock classification in thin section: biotite hornblende gabbro, sample showing some alteration.

4.2.2. The second sample

The mineral composition in the second sample is mainly plagioclase feldspar, hornblende, biotite, and quartz (Fig. 4).

Fig. 4. Telescopic core identification photograph of the second volcanic rock from XIKE-1: diorite. (plagioclase (P1); hornblende(Hb1); biotite (Bi); quartz (Qz)).

Plagioclase feldspar (Pl): accounts for 40%-45% of the whole rock; subhedral to anhedral, frequent polysynthetic,carlsbadalbite compound, pericline compound twins, few simple twins, first-order gray-white interference color, partially altered, sericitization and epidotization along cleaves,grain size 0.2-1.2 mm, 46% An content, classified as andesine.

Hornblende (Hb1): accounts for 20%-25% of the whole rock; granular to columnar in shape, dark green to light yellow in plane polarized light, apparent leochroism, two groups of cleavage, grain size 0.2-0.5 mm, classified as hornblende.

Biotite (Bi): accounts for 15%-20% of the whole rock;plate to sheet in shape, reddish-brown to light yellow in plane polarized light, one group of excellent cleavage, second-order interference color, some irregular reddish-brown biotite showing complete extinction or extreme low interference color in crossed nicols, grain size 0.2-0.6 mm.

Quartz (Q): accounts for 5%-10% of the whole sample;anhedral, granular shape, no color in plane polarized light,low positive relief, first-order interference color, grain size 0.2-0.5 mm.

Rock classification in thin section: gray fine-grained biotite diorite.

4.2.3. The third sample

The sample is mainly composed of plagioclase feldspar,quartz, and biotite (Fig. 5).

Fig. 5. Core identification photograph of the third volcanic rock from XIKE-1: granite. (plagioclase (P1); K-feldspar (Kfs); quartz(Qz); biotite (Bi)).

Feldspar: dominated by plagioclase, anhedral to subhedral, plate in shape, some dissolution harbor or arc-shaped,polysynthetic and simple twins, first-order gray interference color, sericitization, carbonicacidation and epidotization along cleavage, grain size 0.2-2.0 mm; few potassium feldspar grains with common stripe twinning and few lattice twinning,grain size 0.2-0.8 mm; plagioclase and potassium feldspar taking up 30%-35% and 10%-15% of the whole rock, respectively.

Quartz (Q): accounts for 30%-35% of the whole sample;anhedral, granular shape, some harbor-like dissolution features, low positive relief, first-order interference color, grain size 0.2-0.5 mm, occasional myrmekite texture, grain size 0.2-0.8 mm.

Biotite (Bi): sheet shape, irregular, dark brown to light yellow, apparent leochroism, one group of excellent cleavage,second to third-order interference color, some chloritization,few irregular reddish-brown biotite showing complete extinction or extreme low interference color in crossed nicols, grain size 0.2-0.8 mm along long axis, a few green grains probably the result of alteration.

Rock classification in thin section: gray white fine-grained quartz granite, sample showing some minor alteration.

5. Zircon crystal morphology

Thousands of zircon grains were physically separated from the three samples. The standard and strategy for selecting zircon grains in this study were adopted from descriptions in Rubatto D et al., (2000) and Hoskin PWO et al., (2003). A representative set of 25 euhedral to subhedral, inclusion-free zircon grains was chosen for U-Pb dating. The shapes of selected zircon grains are described below.

5.1. Zircons from the first sample (Fig. 6A)

1) Quartet long cylindrical shape

2) Dark-colored rounded multi-coned shape

3) Multi-colored irregular shape

4) Rice grain shape

5) Diamond shape

6) Other irregular shapes

5.2. Zircons from the second sample (Fig. 6B)

1) Irregular quadrilateral shape

2) Irregular triangle shape

3) Cylindrical coned shape

4) Long cylindrical shape

5) Elliptical coned shape

6) Quadrilateral shape

5.3. Zircons from the third sample (Fig. 6C)

1) Hexagonal diamond coned shape

2) Irregular shape

3) Nearly quadrilateral rounded shape

4) Nearly pentagonal coned shape

5) Nearly quadrilateral shape

5.4. Zircon internal textures

The internal textures and the origins of the selected zircons were investigated by cathodoluminescence (CL) imaging. Zircon zonation patterns reflect variations in trace element (e.g., U, Th and Pb) compositions or crystal lattice defects (Wu BY and Zheng YF, 2004). Also, zircons with different origins exhibit distinct CL patterns.

The three samples in the study are deep volcanic intrusive rocks classified as basic gabbro, intermediate diorite, and acidic quartz granite, respectively. Among them, gabbro is considered as the representative rock type for the lower oceanic crust in which primitive magmas are derived from the very lower crust or upper mantle.

Various internal textures and CL patterns were observed in zircons from Xike-1 in the XB. The overall textures of all observed CL patterns are relatively simple. For the first sample the textures include: (1) well-banded zonation (a and e in Fig. 6A); (2) sheet-like structureless zonation (b, c, f, and g in Fig. 6A); (3) dissolution relic core, (j and k in Fig. 6A); (4)weak zonation (d in Fig. 6A); (5) no zonation (i in Fig. 6A);and (6) isometric (h in Fig. 6A). For the second sample, zircons show textures of (1) weak zonation (a and b in Fig. 6B);(2) sheet-like zonation (c, d, and h in Fig. 6B); and (3) no zonation (e, f, and g in Fig. 6B). Zircon textures in the third sample are (1) oscillatory zone (a and c in Fig. 6C); (2) dissolved sheet-like texture (b and d in Fig. 6C); and (3) dissolved fan-shaped zonation (c and f in Fig. 6C).

Fig. 6. Zircon CL and SHRIMP dating results of XIKE-1. Sample A: mostly magmatic zircon; Sample B: magmatic zircon, containing some weakly metamorphic zircon; Sample C: mostly magmatic zircon.

Comparing our CL patterns and zircon textures with published data (Vavra G et al., 1996; Pidgeon RT et al., 1998;Rubatto D et al., 1998; Schaltegger U et al., 1999; Rubatto D et al., 2000; Hermann J et al., 2001; Liati A et al., 2002;Rubatto D, 2002; Rubatto D et al., 2003; Chen YL et al.,2004), we found that zircons with textures of oscillating zones and fan-shaped zonation in the granite sample are more typical and representative than others (a, b, c, e, f in Fig. 6C). Several zircons in the samples of gabbro and diorite show some degree of dissolution with alteration, as evidenced by dissolution features along zircon grain margins, such as thinned margins (i in Fig. 6A), unequal anisometric margins (b, c, g, h in Fig. 6A; f in Fig. 6B), altered textures (j in Fig. 6A; a, b, h in Fig. 6B; d in Fig. 6C), relic core (k in Fig. 6A; f in Fig. 6C),cutting zonation features, apparent magma relics in the core of zircons (c, j in Fig. 6A; d in Fig. 6B), and breakdown of the centrial axis (b in Fig. 6C). Some zircons may have been affected by metamorphism that led to weak or no zonation (Fig. 6C).

5.5. Zircon SHRIMP U-Pb dating results

Zircon SHRIMP U-Pb dating results of 25 zircon grains from the three samples from the bedrock of Xike-1 in the XB are presented in Table 1 and Fig. 6. The calculated ages are largely within early Proterozoic, late Proterozoic, early Paleozoic, late Jurassic and Cretaceous (Table 2, Fig. 7).

Fig. 7. Zircon concordia curves of the three samples.

5.5.1. Early Proterozoic ages

Analysis of samples A and B yielded two early Proterozoic ages from the basement. Sample A is classified as gabbro with a probable deep intrusive origin. Zircons in this sample A are typically of magmatic origin and thus the age of 2451±22.7 Ma obtained is probably reliable. Sample B is recognized as diorite and also belongs to deep intrusive rock.Given the trace element compositions, the zircons in sample B must be of both magmatic and metamorphic origins and the age of 1857±46.5 Ma is reliable, too. The two dates obtained are the greatest SHRIMP U-Pb ages reported so far for the basement of the XB.

5.5.2. Late Proterozoic ages

A Late Proterozoic age of 750±4.9 Ma was derived from sample A, which is classified as deep volcanic intrusive rock.

5.5.3. Late Ordovician ages

A Late Ordovician age of 445±20.7 Ma was determined from sample C, which was collected 4 m below the overlying reef carbonate rocks.

5.5.4. Jurassic ages

All three samples yielded Jurassic dates ranging from 153±0.7 Ma to 144±1.3 Ma (Middle to Late Jurassic). These ages are the dominant range from the samples, representing peak geological deformation during the Jurassic period.

5.5.5. Cretaceous ages

The samples yielded dates within 140±1.1 Ma to 131±11.9 Ma (Early Cretaceous) and 108±0.9 Ma to 106±1.5 Ma (Middle to Late Cretaceous).

5.6. Zircon REE composition

The studied samples show a large variation in REE composition, especially in the Th and U contents. The U contentof the gabbro ranges from 209×10-6to 5456×10-6and Th content from 57×10-6to 6918×10-6; Th/U values for the gabbro are from 0.23 to 1.31, with two values <0.4. Diorite shows content 60×10-6-1058×10-6for U and 24×10-6-756×10-6for Th, with Th/U values ranging from 0.10 to 0.77 and five values <0.4 but >0.1. Granite displays U content ranging from 227×10-6to 3137×10-6and Th content from 134×10-6to 4029×10-6, with Th/U values from 0.31 to 1.33 and two values<0.4 (Table 2). For all samples, the mean Th/U value is >0.65 and the minimum and maximum values are 0.1 and 1.49, respectively. Previous studies indicate that zircons with a magmatic origin commonly have Th/U values >0.4 (Rubatto D et al., 2000; Miller A et al., 2003). Thus, zircons in the study are mostly of magmatic origin. Nine zircons show Th/U values in the transition zone of magmatic to metamorphic zircons,demonstrating the influence of metamorphism in the evolutionary history of our samples.

Table 1. Zircon SHRIMP U-Pb dating results of three bedrock in Xisha block.

Table 2. Th, U, Pb element compositions of three studied volcanic intrusive samples.

6. Discussion

6.1. Previous ages of the Xisha Block

Sample A: mostly magmatic zircon; Sample B: magmatic zircon, containing some weakly metamorphic zircon; Sample C: mostly magmatic zircon.

On the basis of radioactive dating of drill cuttings petrologically recognized as granitic gneiss from Yongxi-1 well,three groups of ages have been reported for the XB:126.6±2.0-119.3.1±1.9 Ma (Jing XL, 1989), 628 Ma (Zeng,1977), and 1465 Ma using whole rock Rb-Sr dating and K-Ar dating, respectively. Xiu C et al. (2016) reported 144-158 Ma from Xike-1. These four groups reflect tectonic events in the Mesoproterozoic, Neoproterozoic, Jurassic, and Cretaceous.

6.2. Ages of the blocks surrounding the Xisha Block

The lithosphere adjacent to the XB is composed of several blocks, including Nansha, Hainan, the Kontum massif (Indochina), and South China. Published radioactive ages of these blocks are listed in Table 3. Among these ages, the greatest ages reported are 1780±5 Ma (zircon U-Pb dating) for the Archean metamorphic complex (Ngoc Linh granitic gneiss) in the Kontum massif (Chen JF et al., 2014), 1455±12 Ma (zircon U-Pb dating) for the Hainan Block, and 2137 Ma(zircon U-Pb dating) for the South China Block (Xu X et al.,2003). The abovementioned ages are close to those determined in this study and are probably reliable. In contrast, the age of 2700 Ma calculated from whole rock Sm-Nd modeling(Lan CY et al., 2003) and the that of 2300 Ma derived from U-Pb isochron dating for the Kontum massif are likely less reliable.

6.3. Geological significance

In this study, we report two new Paleoproterozoic ages for the XB, which have three-fold implications for the evolution of the block.

Firstly, our ages indicate that the XB preserves partial Paleoproterozoic crustal material, likely sourced deeply from the lithosphere. Clearly, the ages of 2451±22.7 Ma and 1857±46.5 Ma presented in this study are the earliest zircon SHRIMP U-Pb ages reported so far for the XB. Also, the revealed Neoproterozoic age of 750±4.9 Ma is probably the consequence of magmatism in the XB in response to the breakup of the Neoproterozoic Rodinia supercontinent (Zhao GC et al., 2002). The further age of 445±20.7 Ma is probably indicative of an encounter event between the XB and Pangaea supercontinent fragments during the Caledonian Orogeny.

Secondly, the host rock containing the studied zircons is recognized as a deep intrusive rock, which signifies that the zircons were transported to the near-surface during various magmatism events. According to the shapes and internal textures of the zircons, the transportation distances should not have been significant, suggesting that migmatization might have played a major role in the formation of the XB basement.

Thirdly, our SHRIMP U-Pb dating, together with previous studies, reveals that the ages peaked during the time periods of 153±0.7-144±1.3 Ma, 140±1.1-131±11.9 Ma, 108±0.9-106±1.5 Ma (Fig. 7; Table 1, 3). This may reflect that the XB was probably within the eastern extension of the Neotethys tectonic zone. The deformation and/or magmatism was at the highest level during the late Jurassic to early Cretaceous, representing the Yanshanian Orogeny. As compared to SHRIMP U-Pb results from surrounding regions, the evolutionary history of the XB is broadly compatible with the Nansha Block(Table 3) (Yan, 2008; Yan et al., 2010) and the South China Block (David et al., 1997; Li et al., 2000; Xu X et al., 2003;Zhang Y et al., 2015), but is different from the Haihan block and the Kontum Massif, which are likely related to the evolution of Pangaea until the early Cretaceous (90-87 Ma) (Chen JF et al., 2014). The close proximity between the Xisha and Nansha blocks indicates a possible coeval evolution for the two units during the formation and breakup of Pangaea, as well as for the eastern extension of Neotethys tectonic domain. During this period, the XB was incorporated into the eastern portion of the Neotethys deformation region. During the Late Triassic, the collisions and sequential accretion between the Upper Yangtze, Yunkai and Hainan blocks resulted in a connected terrane. Together with the Indochina Block,this united continent led to the formation of the Indochina-South China continent, which was later strongly deformed during the subsequent Yanshanian Orogeny, giving rise to the well-known Mesozoic subduction zone (Liu HL, 2004). It is noted that the Hainan Block also yielded several fairly old ages from zircon U-Pb and whole rock Sm-Nd dating;however, these ages are different from those in the XB presented in the study.

Table 3. Reported radioistope ages of Xisha Block and surrounding structures.

Correlation with the Southwest Basin of the South China Sea will be possible when further efforts are made in this area.

7. Conclusions

Volcanic intrusive rocks have been determined in the basement of the Xisha block in the exploratory well Xike-1.Three rock samples were collected from this well at 2 m intervals and recognized as dark-gray biotite-hornblende gabbro,gray fine-grained biotite diorite, and gray fine-grained granite.The samples are of basic, intermediate, and felsic rocks in origin, respectively. Several thousand zircons separated include euhedral grains of fifteen morphologies. They show little alteration and have been transported over relatively short distances.

Geochemical analysis and the SHRIMP U-Pb dating method revealed Th/U ratios greater than 0.4 for most zircons,indicating typical magmatic origin and Paleoproterozoic, Neoproterozoic, late Ordovician, middle to late Jurassic, and earlyCretaceous ages. The earliest SHRIMP U-Pb age of a sample is 2451 Ma, representing the oldest recorded zircon for the Xisha Block and surrounding regions. The ages recorded indicate that the formation of the Xisha Block was influenced by the evolution and closure of the Neotethys Ocean, probably within its eastern extension in the South China Sea. Both old, deep-sourced material, including fragments from Rodina supercontinent, and recent mantle-derived magma products contributed to the emergence and formation of the Xisha Block. The dominant Jurassic-early Cretaceous ages among our data suggest a maximum level of formation reached during that time span. The evolution was also controlled by the formation and breakup of Pangaea.

Our age data also suggest that the evolutionary history of the Xisha Block is broadly compatible with that of the Nansha and South China Blocks, but different from the Haihan Block and Kontum Massif. The Xisha, Zhongsha, and Nansha Blocks seem to have been strongly affected by the Yanshanian Orogeny and its associated subduction and magmatism processes.

Acknowledgments

We are grateful to Jianhui Liu at the Beijing SHRIMP Center of the Chinese Academy of Geological Sciences for his assistance in zircon SHRIMP U-Pb dating and CL analysis. We also want to thank the CNOOC Zhanjiang Branch for providing samples to make professional identification of the samples. We are most grateful to the reviewers and editors who gave help to improve the paper. The research was financially supported by the Technology Basic Resources Investigation Program of China (2017FY201407); China Geological Survey and Major Oil and Gas Technology Projects (GZ H201200510; 2011ZX05025-002-04).