Detrital zircon populations of the South Qiangtang terrane, central Tibetan Plateau, and their implications for Tethyan evolution

ABSTRACT Detrital zircon populations of the South Qiangtang terrane (SQT) provide vital information for reconstructing the Tethyan evolution of the Tibetan Plateau but are obscured by the undefined affiliation of detrital zircon samples and the superimposition of multiple Tethyan tectonothermal events. We outlined the SQT based on geological mapping and collated Cambrian-Triassic detrital zircon samples to investigate representative detrital zircon populations. Although diachronous strata have distinctive detrital zircon age distributions, the Cambrian-Permian strata record similar ca. 2.50 Ga, 950 Ma, 800 Ma, and 550 Ma populations, which can be used in future detrital zircon studies as Palaeozoic indicators for the SQT. We analyse the above detrital zircon data and compare them to those from adjacent microcontinents (e.g. North Qiangtang terrane, NQT). We note that (1) the Cambrian-Late Ordovician detrital zircons from Upper Triassic strata yield positive εHf(t) values, suggesting juvenile arc-related magmatism in the NQT, and (2) the late Palaeozoic strata in the SQT and NQT both contain Late Devonian-early Carboniferous detrital zircons but have different detrital zircon populations. Together with previous studies on Tethyan evolution, we support that the southward subduction of the Proto-Tethys oceanic plate opened two diachronous back-arc basins between the NQT and SQT during the late Cambrian-Ordovician and Late Devonian-early Carboniferous. The former closed during the late Silurian; the latter evolved into the Palaeo-Tethys Ocean. Subsequently, the northward subduction of the Palaeo-Tethys oceanic plate produced early-middle Permian arc magmatism in the NQT, and the slab pull force from the subducting oceanic plate caused middle-late Permian intraplate magmatism on the opposite passive margin. GRAPHICAL ABSTRACT


Introduction
The Tibetan Plateau is part of the most striking continent-continent collisional orogenic system (i.e. Cenozoic India-Eurasia collision) in the world (Harrison et al. 1992;Tapponnier et al. 2001;Royden et al. 2008). Previous studies have suggested that its present-day tectonic framework has been predominantly controlled by complicated Tethyan evolution since the Cambrian (Dewey et al. 1988;Hsü et al. 1995;Yin and Harrison 2000). In this context, the Tibetan Plateau formed from the amalgamation of several microcontinents, including the Kunlun, Songpan-Ganze, Qiangtang and Lhasa blocks, and the Himalayan margin, along at least four accepted sutures (i.e. East Kunlun-A'nyemaqen, Jinsha, Bangong-Nujiang, and Indus-Yarlung-Zangbo), and these sutures record the evolutionary history of the Proto-, Palaeo-, Meso-, and Neo-Tethys Oceans (Figure 1(a); e.g. Yin and Harrison 2000;Metcalfe 2013, Metcalfe 2021Zhu et al. 2013;Huang et al. 2018). In fact, previous studies have also reported other Tethyan sutures on the Tibetan Plateau, such as the Longmu Co-Shuanghu Palaeo-Tethys suture within the Qiangtang block (e.g. Li 1987;Zhai et al. 2011;Zhang et al. 2016) and the Sumdo Palaeo-Tethys suture within the Lhasa block (e.g. Yang et al. 2009;Wang et al. 2020a). Consequently, some microcontinents have recorded the superimposition of multiple Tethyan tectonothermal events, and the reconstruction of such a complicated Tethyan evolution has been highly controversial for decades (c.f., Zhao et al. 2018;Li et al. 2019a;Xu et al. 2020a;Metcalfe 2021, and references therein).
Among these microcontinents, the South Qiangtang terrane (SQT) is a typical microcontinent of Gondwanan origin that rifted from the Gondwana supercontinent during the late Palaeozoic (e.g. Metcalfe 2013Metcalfe , 2021Huang et al. 2018;Zhao et al. 2018). To reconstruct the Tethyan evolution of the SQT, previous studies have focused on multiple stages of magmatism and metamorphism, but some key information may be missing due to the destruction of later tectonothermal events (e.g. Xu et al. 2020a, and references therein). Detrital zircon U-Pb age spectra of (meta)sedimentary units can reflect the tectonic settings of the depositional basins and determine the characteristics of the source regions (Fedo et al. 2003;Cawood et al. 2012;Gehrels 2014), and they have thus been widely used to reconstruct Tethyan evolution (e.g. Weislogel et al. 2006;Pullen et al. 2008;Gehrels et al. 2011;Zhu et al. 2011;Zhang et al. 2018;Liu et al. 2020). When comparisons are made to the SQT in detrital zircon studies, the most commonly used detrital zircon signature is roughly constrained by limited detrital zircon samples in the southern part of the Qiangtang block (e.g. Gehrels et al. 2011;Zhu et al. 2011;Ma et al. 2017;Zhang et al. 2018). However, it is uncertain whether these data are truly representative of the SQT because the terrane attribution of some detrital zircon samples in the Qiangtang block is controversial. For example, a glaciomarine diamictite sample (619051A; Figure 1(b)) that was previously considered from the NQT    Table 1. We consider that the oldest sedimentary sequences are Cambrian rather than late Neoproterozoic, as suggested by Wang et al. (2015) (see text for explanation). Detrital zircon samples within central Qiangtang (yellow dots) are from Pullen et al. (2008), Gehrels et al. (2011), Zhu et al. (2011), Zhao et al. (2014), Fan et al. (2015Fan et al. ( , 2016, Peng et al. (2019), Wang et al. (2016a), Liu et al. (2017a), Ma et al. (2017), Xie et al. (2018), Li et al. (2019b), and . The locations of the Jinsha and Bangong-Nujiang sutures are adapted from Wang et al. (2013). The locations of 'ophiolites' and high-pressure metamorphic rocks in central Qiangtang are from Xu et al. (2020a), who mark these rocks based on dozens of 1:50,000 and 1:200,000 geological maps. Two dashed lines and one thick solid line between them show the conservative and less conservative ranges of central Qiangtang, respectively. Twelve samples (four from the Southern Qiangtang terrane and eight from the Northern Qiangtang terrane) with strikethrough are excluded because of a small number of best ages (< 60, see text). '-': no error data.
likely belongs to the SQT (Xu et al. 2020a). In addition, multiple Tethyan tectonothermal events (e.g. opening, subduction, and closure of the Proto-, Palaeo-, and Meso-Tethys oceans; c.f., Metcalfe 2013Metcalfe , 2021Zhu et al. 2013;Huang et al. 2018) possibly produce characteristic detrital zircon age populations and thus complicate the detrital zircon signature of the SQT. In summary, the detrital zircon age spectra of the associated stratigraphic units in the SQT have never been systematically studied, which further prevents us from answering some questions related to the evolutionary history of the SQT. (1) Does detrital material constrain the pre-Ordovician crystalline basement of the SQT and the Ordovician back-arc basin to the north (c.f., Dan et al. 2020)? (2) Do the late Palaeozoic detrital zircon populations support or oppose the Palaeo-Tethys Ocean to the north of the SQT (e.g. Gehrels et al. 2011)? (3) Do the sedimentary strata record the initial opening of the Meso-Tethys Ocean (e.g. Fan et al. 2021)?
In this contribution, we delineated the tectonic boundaries of the SQT based on previous geological mapping for ophiolites and high-pressure metamorphic rocks and found that many detrital zircon samples belong to central Qiangtang rather than the SQT (Figure 1(b)). Then, we compiled only Cambrian-Triassic detrital zircon samples in the SQT prior to the initiation of the northward subduction of the Bangong-Nujiang Meso-Tethys oceanic plate (latest Triassic-earliest Jurassic; Zhu et al. 2013; to investigate whether variations existed over time and to provide a reference database for future detrital zircon studies on the SQT. Additionally, by comparing detrital zircon data from adjacent microcontinents (e.g. NQT, Gehrels et al. 2011;Liang et al. 2019) and incorporating previous geological observations (e.g. Xu et al. 2020a), we aim to reconstruct the Tethyan evolution involving the SQT.

Geological setting
The Tibetan Plateau consists of the Himalayan margin and several major microcontinents from south to north, including Lhasa, Qiangtang, Songpan-Ganze, and East Kunlun (Figure 1(a)). The Qiangtang block, which is located on the central-northern part of the Tibetan Plateau, is tectonically bounded by the Jinsha suture to the north and the Bangong-Nujiang suture to the south (Figure 1; e.g. Allegre et al. 1984;Dewey et al. 1988).
Initially, Li (1987) proposed that an in situ Palaeo-Tethys suture (i.e. the Longmu Co-Shuanghu suture), which consists of late Palaeozoic-Triassic low-temperature high-pressure metamorphic rocks and ophiolites, divides the Qiangtang block into the SQT and North Qiangtang terrane (NQT) (e.g. Zhai et al. 2011;Zhang et al. 2016;Liang et al. 2017;Xu et al. 2021; note: the SQT and NQT are also termed the Western Qiangtang terrane and Eastern Qiangtang terrane, respectively; e.g.; Zhu et al. 2013); subsequently, several studies on the associated metamorphic rocks favour that the blueschistbearing metamorphic mélange (i.e. the Late Triassic central Qiangtang metamorphic belt, CQMB) was underplated from the Jinsha Palaeo-Tethys suture and then exhumed by detachment faulting in the interior Qiangtang (Kapp et al. 2000(Kapp et al. , 2003Pullen et al. 2008). In summary, the petrogenesis of the late Palaeozoic-Triassic mélange has remained the subject of intense debate for a long time (see details in Xu et al. (2020a)). Recently, many researchers have also reported records of early Palaeozoic tectonothermal events in the central part of the Qiangtang block (e.g. Pullen et al. 2011;Xu et al. 2014;Zhang et al. 2014;Zhai et al. 2016). Therefore, Xu et al. (2020a) redefined the tectonic mélange that separates the SQT and NQT and consists of Cambrian-Triassic metamorphic rocks and ophiolites as central Qiangtang (Figure 1(a)). Consequently, central Qiangtang records not only Palaeo-Tethyan evolution but also possible Proto-Tethyan evolution (c.f., Xu et al. 2020aXu et al. , 2020bMetcalfe 2021).
The SQT consists of discontinuous Cambrian to Cenozoic sedimentary strata (BGMR, 1993;Li et al. 1995a;Gehrels et al. 2011;Yang et al. 2014;Wang et al. 2016a) with nearly contemporary magmatic rocks (c.f., Pullen et al. 2011;Wang et al. 2014Wang et al. , 2015Liu et al. 2016a).  first identified late Neoproterozoic strata in the Dabure area ( Figure 1(b)) based on basalt interlayers that formed at ~550 Ma . However, the dated zircon grains from the associated basalt samples are subrounded to rounded (see Figure 2 in Wang et al. (2015)), suggesting that they are more likely to be detrital zircons captured during the eruption of basic magma. Combined with their different detrital age populations from those of Ordovician detrital zircon samples (Figures 2 and 3(a), we speculate that these continental clastic rocks were deposited during the Cambrian. In summary, the lower Palaeozoic sequences are a suite of shallow sea facies carbonates and clastic rocks that were formed in a passive continental margin or an intracontinental rift (e.g. Cheng et al. 2007;Wang et al. 2016a;Liu et al. 2017b). The Pennsylvanian (upper Carboniferous)-Cisuralian (lower Permian) glacial-marine deposits, as one of the representative sedimentary units in the SQT, contain abundant cold-water biota, indicating a Gondwanan affinity during this period (e.g. Chen and Xie 1994;Metcalfe 1994Metcalfe , 2013Jin 2002), and the Gondwanan origin of the SQT is also supported by the similarity of  Table 1. The left pie diagrams show the corresponding detrital age distributions. (i) Cumulative probability diagram for integrated detrital zircon samples in the Southern Qiangtang terrane. The coloured columns show the corresponding age ranges of the deposition of compiled samples. detrital zircon populations between Ordovician-Permian strata in the SQT and those from the Himalayan margin (Leier et al. 2007;Gehrels et al. 2011;Zhu et al. 2011). Recent studies suggest that Pennsylvanian-Cisuralian strata have detrital provenance changes, representing the opening of the Meso-Tethys Ocean (Fan et al. 2021), whereas Mesozoic sedimentary sequences record the northward subduction of the Bangong-Nujiang Meso-Tethys Ocean and the subsequent continent-continent collision between the SQT and the Lhasa block (e.g. Ma et al. 2017). In conclusion, the massive late Palaeozoic-Cenozoic strata in the SQT  probably cover the older sedimentary units (e.g. Silurian-Devonian) and further obscure the early history of the SQT. Two additional remarkable features of the SQT are Ordovician granites Liu et al. 2016a;Wang et al. 2020b) and widespread late Carboniferous-early Permian mafic dikes Wang et al. 2014;Xu et al. 2016). The former was recently considered to represent the crystalline basement of the SQT (Dan et al. 2020), whereas the latter is generally attributed to intracontinental rifting on the northern margin of Gondwana Xu et al. 2016).

Data compilation and methods
Previous studies have attempted to reveal the nature of the SQT using detrital zircon geochronology; however, many of them collected detrital zircon samples within the range of central Qiangtang (Figure 1(b), e.g. Pullen et al. 2008;Zhu et al. 2011;Fan et al. 2015;Wang et al. 2016a;Liu et al. 2017a;Ma et al. 2017;Li et al. 2019b;Peng et al. 2019;Fan et al. 2021). Consequently, the detrital age populations of the SQT are still enigmatic. In this study, we ignored detrital zircon samples belonging to central Qiangtang and complied detrital zircon U-Pb isotope data from 34 'efficient' detrital zircon samples from Cambrian-Triassic units within the SQT (Figure 1(b); Table 1). Most of the compiled analyses were collected using laser ablation-inductively coupled plasma-mass spectrometry, and only those from one Ordovician quartzite sample (XZ0701, Dong et al., Dong et al. 2011) are sensitive high-resolution ion microprobe data (Table 1).
Data quality for each analysis was re-evaluated based on the following criteria (modified from Gehrels et al. 2011Gehrels et al. , 2014: (1)  Finally, these criteria resulted in rejection of ~4% of all the analyses. In addition, four detrital zircon samples (samples 6-12-06-11, 6-12-06-12, T5 and TGT4) with less than 60 best ages (c.f., Sharman and Malkowski 2020) were excluded, and the remaining 30 samples that contained 69-188 best analyses (Table 1) were used to discuss the detrital age populations of the SQT.
The detritalPy toolset (Sharman et al. 2018) was used to generate kernel density estimates (KDE; bandwidth = 25 Ma, Shimazaki and Shinomoto 2010) and nonmetric multidimensional scaling (MDS; Kolmogorov-Smirnov statistic D max , Vermeesch 2013) plots and to calculate the maximum depositional ages (MDAs), including (1) the youngest single grain age with 1σ (YSG), (2) the weighted mean age of the youngest two or more grains overlapping in age at 1σ [YC1σ(2 +)], and (3) the weighted mean age of the youngest three or more grains overlapping in age at 2σ [YC2σ(3 +)]. In this study, all three methods yielded similar MDAs (Table 1). The YC2σ(3+) method, which is the most conservative method and generally yields the oldest MDA, was applied here because it produces ages that are not younger than true depositional ages (Coutts et al. 2019).
During the discussion of Tethyan evolution, we also used a small amount of Hf isotope data from detrital and magmatic zircons. All εHf(t) values were recalculated with reference to the chondritic reservoir (CHUR) at the time of zircon growth from magmas. To calculate the εHf (t) values, we further adopted a decay constant for 176 Lu of 1.867 × 10 −11 yr −1 (Söderlund et al. 2004) and chondrite values of 176 Hf/ 177 Hf = 0.282772 and 176 Lu/ 177 Hf = 0.0332 (Blichert-Toft and Albarède 1997).

Detrital zircon geochronology compilation of diachronous strata in the SQT
The SQT contains discontinuous Palaeozoic to Triassic sedimentary strata, and sedimentary strata with different depositional ages probably have different detrital age distributions. Therefore, we present their detrital zircon signatures mainly based on stratigraphic ages, including Cambrian, Ordovician, late Carboniferous-early Permian, late Permian, and Late Triassic.
Three Cambrian clastic samples have consistent age distributions ( Figure S1), which are similar to those of Neoproterozoic-Miocene strata from the Himalayan margin (c.f., Zhang et al. 2018). They are characterized by middle ca. 2.50 Ga, 1.85 Ga and 1.60 Ga populations, a major ca. 1.00 Ga-700 Ma population that peaks ca. 950 and 800 Ma, and a smaller ca. 580 Ma population (Figure 2(a)). Note that no Precambrian rocks rich in zircon have been reported in the SQT, and the detrital zircon grains from the Cambrian samples are subrounded to rounded (Wang et al. 2016a), indicating long-distance transport from areas beyond this terrane. Due to the palaeogeographic affinity of the SQT to the Himalayan margin and the Indian plate during the Palaeozoic (e.g. Huang et al. 2018 Gehrels et al. 2006a;McKenzie et al. 2013;Wang et al. 2019a).
Previous studies have reported nine detrital zircon samples from the Ordovician sedimentary units in the SQT (Table 1). Among them, eight samples show comparable age distributions ( Figure S2), suggesting a similar detrital provenance. We thus consider that these eight samples are from Ordovician sedimentary sequences, whereas one other sample (sample 6-4-06-2b; Pullen et al. 2011; Figure 2(c) and Table 1) probably belongs to Carboniferous-Permian strata (see below). Previous geological observations have suggested that the Ordovician sedimentary sequences in the SQT generally contain Middle-Late Ordovician nautiloid fossils or are intruded by Early Ordovician granites (e.g. Cheng et al. 2007;Dong et al. 2011;Pullen et al. 2011;Yang et al. 2014;Liu et al. 2017b). Their detrital age populations were different from those of Cambrian strata (Figure 2(a,b)), and the calculated MDAs ranging from 546.6 ± 4.0 to 485.4 ± 3.3 Ma (Table 1) also support that sedimentation continued through the Ordovician. In summary, eight detrital zircon samples have internally consistent age populations ( Figure S2) of 2.52-2.45 Ga, 1.05 Ga-920 Ma, 870-770 Ma and 700-530 Ma, with a small ca. 1.15-1.05 Ga population (Figure 2(b)).
The Pennsylvanian-Cisuralian strata in the SQT are characterized by glacial-marine diamictite (e.g. Li et al. 1995b). The associated sedimentary sequences usually contain heterogeneous basic volcanic interlayers (e.g. Wang et al. 2009) and abundant marine fossils of bivalves, solitary corals, and brachiopods (e.g. Liang et al. 1983;Zhang et al. 2013). However, all eleven collated detrital zircon samples contain only late Neoproterozoic-early Cambrian MDAs ranging from 570.9 ± 3.2 Ma to 513.6 ± 3.3 Ma (Table 1), which are much older than their true depositional ages, suggesting a relatively stable depositional setting (e.g. passive continental margin; c.f., Cawood et al. 2012). They have similar detrital age distributions ( Figure S3), and the primary age groups in these samples are ca. 2.50 Ga, ca. 800 Ma, and ca. 560 Ma, with few grains that are ca. 1.88 Ga and ca. 950 Ma (Figure 2(d)). Compared with Ordovician strata, a significant variation in the detrital age population is the lack of the Grenvillian population (ca. 1.30-1.00 Ga; Figure 2(b,d)). As mentioned above, Pullen et al. (2011) collected an 'Ordovician' sample (6-4-06-2b) near the thrust fault that separates the Carboniferous-Permian strata and pre-Middle Ordovician metasedimentary rocks, and this sample has a low proportion of ca. 800 Ma zircons, similar to that of Ordovician strata (Figure 2(b,c)). However, this sample also has a late Carboniferous-early Permian detrital zircon signature, such as a high ratio of ca. 550 to ca. 950 Ma zircons (Figure 2(b,d)), which probably record the transitional features from Ordovician strata to Pennsylvanian-Cisuralian strata.
The Lopingian (upper Permian) strata are distributed only in the western part of the SQT (c.f., Zhang et al. 2013;Fan et al. 2021), and some of them contain abundant fusulinida, coral, brachiopod, gastropod and trilobite fossils, suggesting late Permian sedimentation (e.g. Wu and Lan 1990). Currently, only Fan et al. (2021) collected three samples of late Permian sandstone in the SQT, and they have MDAs ranging from 259.2 ± 2.8 Ma to 255.2 ± 1.9 Ma (Table 1), which is close to their true depositional ages (i.e. late Permian). Strangely, three samples yielded internally inconsistent age distributions (Figure 2(e-g)). Sample D18T16 yields major age populations of ca. 530 Ma and ca. 960 Ma, with minor pre-Palaeoproterozoic and post-Cambrian age groups, similar to that of late Carboniferous-early Permian sample S1921 ( Figure S3). Sample D18T17 yields similar age distributions, but it also contains a distinct Grenvillian age population (Figure 2(f)), which is generally considered a typical detrital zircon signature of the Lhasa block and Australia (Zhu et al. 2011). Sample B19T17 has a very different detrital zircon signature (Figure 2(g)), arguing against late Permian sedimentation (see below).
Late Triassic strata are sporadically distributed in the SQT   (Figure 2(h) and S4). Fan et al. (2021) collected a 'late Permian' sandstone sample (B19T17) near the angular unconformity boundary between the Late Triassic and late Permian strata, but this sample yields age distributions similar to those of Late Triassic samples (Figure 2(g,  h)), suggesting possible Late Triassic sedimentation. It seems that the reassigned depositional age of sample B19T17 is questionable because its MDA (ca. 255 Ma) is older than those of Late Triassic samples (ca. 233-210 Ma, Table 1). However, we note that one Late Triassic sample (06GT10) has only four Triassic detrital zircon grains (4%, 245-224 Ma, Gehrels et al. 2011) and has the youngest age population of ca. 265 Ma ( Figure S4), similar to that of sample B19T17 (ca. 269 Ma, Figure 2(g)). It is possible that a limited number of Triassic detrital zircon grains have not been detected by Fan et al. (2021). More importantly, the late Permian Qiangtang block (or the SQT) was an isolated microcontinent between the Palaeo-Tethys Ocean and Meso-Tethys Ocean (e.g. Yin and Harrison 2000;Xu et al. 2020a;Metcalfe 2021). Due to the lack of Precambrian magmatism in the Qiangtang block, the Precambrian detrital zircon populations of the SQT are the result of sedimentary recycling, which cannot form a major ca. 1.86 Ga but negligible ca. 970-530 Ma populations (c.f., Figure 2(a-g)). Therefore, sample B19T17 was also probably deposited during the Late Triassic.

Comparison of diachronous spectra: Representative detrital age populations of the SQT
In the SQT, the widespread stratigraphic units that contain abundant, nearly rounded Precambrian detrital zircon grains are mainly composed of coarse-grained sediment (e.g. sandstone; Table 1 and references therein), suggesting long-distance transport from areas beyond this terrane. The fact that Precambrian crystalline basement has not been found in this terrane also supports a nonproximal source outside the SQT. Previous studies have demonstrated that young strata could inherit distinct detrital zircon signatures from older sedimentary units through sedimentary recycling (e.g. Andersen et al. 2016), in which diachronous strata record similar detrital age distributions. Therefore, most studies put the diachronous sedimentary units (i.e. Cambrian-Triassic units) together to discuss the detrital zircon signature of the SQT (e.g. Gehrels et al. 2011;Zhu et al. 2011;Zhang et al. 2018). However, a recent review shows that the SQT probably underwent complicated Tethyan evolution during the Palaeozoic-early Mesozoic (e.g. Xu et al. 2020a), and it is uncertain whether the detrital zircon signature of the SQT changed over time; for example, Fan et al. (2021) identified a significant detrital provenance change between the Pennsylvanian-Cisuralian and Lopingian strata.
The Cambrian sedimentary units have six visible detrital age populations, including ca. 2.50 Ga, 1.85 Ga, 1.60 Ga, 950 Ma, 800 Ma, and 580 Ma (Figure 2(a)). The unique low abundance of the late Neoproterozoic age group in the Palaeozoic strata (Figure 2(a-f)), which is comparable to that of lower-middle Cambrian stratigraphic units in the Himalayas (e.g. Myrow et al. 2010), is possibly related to limited exposure to crystalline sources (e.g. Kuunga-Pinjarra orogen, Cawood and Buchan 2007). Additionally, note that the Cambrian clastic samples have a high ratio of ca. 800 Ma to ca. 950 Ma zircons (~2.0; Figure 2(a)), which is significantly different from late Neoproterozoic-Ordovician strata in the Himalayan margin (c.f., Gehrels et al. 2006aGehrels et al. , 2006bMyrow et al. 2003Myrow et al. , 2009Myrow et al. , 2010. Possibly, this is a result of nearby exposures due to a close connection between the Yangtze region of the South China block and the Qiangtang block (after Metcalfe 2013; Xian et al. 2019). Compared with Cambrian clastic rocks, the Ordovician strata lack late Palaeoproterozoic (ca. 1.80-1.60 Ga) detrital record (Figure 2(a,b,i)), suggesting a change in recycled sedimentary strata (e.g. North American Cordilleran strata, Schwartz et al. 2019). More importantly, these detrital zircon samples have a low ratio of ca. 800 to ca. 970 Ma zircons (~0.72; Figure 2(b,i)), excluding the provenance of the Yangtze region and probably resulting from the initial break-up of the SCB from the northern margin of Gondwana or clockwise rotation of the SCB during the Cambrian (c.f., Metcalfe 2013;Xian et al. 2019;Wang et al. 2021b). By the late Palaeozoic, the SCB separated from the northern margin of Gondwana (c.f., Metcalfe 2013; Xian et al. 2019). Therefore, the abundant ca. 800 Ma zircons in the Pennsylvanian-Cisuralian strata (~25%; Figure 2(d,i)) can only be from the Neoproterozoic igneous rocks in the Amdo terrane (e.g. Hu et al. 2021, and references therein) and/or northern India (e.g. De Wall et al. 2018, and references therein). The Neoproterozoic-Cambrian igneous rocks in the Amdo terrane include major ca. 500 Ma granitoids and minor ca. 900-800 Ma gneissic rocks (Hu et al., 2021). Unlike the Amdo terrane, the Pennsylvanian-Cisuralian strata in the SQT have similar ratios of ca. 700-500 Ma (peak ca. 561 Ma) and ca. 875-700 Ma (peak ca. 803 Ma) detrital zircons (Figure 2(d)). Notably, most Carboniferous-Permian sedimentary sequences from the Tethys Himalaya and the Lhasa block have very low ratios of ca. 800 Ma detrital zircons Zhu et al. 2011;Wang et al. 2021), and coeval strata with abundant ca. 800 Ma detrital zircons are only distributed in the western Tethys Himalaya (e.g. Kas17, Agnihotri et al. 2018) and eastern Arabia (i.e. Oman, Craddock et al. 2019). Therefore, we propose that the abundant ca. 800 Ma detrital zircons in the Pennsylvanian-Cisuralian strata in the SQT are likely from the western Tethys Himalaya and/or eastern Arabia. In contrast, the abovementioned decreasing Grenvillian zircon population from Ordovician (~14%) to Pennsylvanian-Cisuralian (<5%) strata (Figure 2(b,d,i)) should be attributed to sedimentary recycling (e.g. Schwartz et al. 2019). It is difficult to accurately assess the detrital age populations of Lopingian strata in the SQT due to limited data. Two late Permian detrital zircon samples, in comparison with the older sedimentary sequences, show us (1) the first appearance of minor post-middle Cambrian age populations (e.g. ca. 460 Ma, ca. 350 Ma and Ca. 260 Ma; Figure 2(e,f,i)), and (2) the surprising major Grenvillian age group (~22%; sample D18T17; Figure 2(f,i)). The former is considered a possible result of proximal deposition (e.g. lithic fragments with (sub)angular shapes) from the SQT itself during the rifting of the SQT from Gondwana (Fan et al. 2021), whereas the latter probably records long-distance transport either from the Lhasa block (c.f., Zhu et al. 2011).
Obviously, these diachronous strata have internally distinguishable detrital zircon signatures, such as the existence/absence and variable proportions of detrital age populations (Figure 2), indicating the significant diversity of crystalline sources (e.g. SCB) and/or recycled sedimentary rocks. Furthermore, the intersample similarities and dissimilarities of detrital age distributions could be visually highlighted by a nonmetric multidimensional scaling (MDS; Figure 3) plot, in which the 'similar' samples with short distances cluster closely together, whereas the 'dissimilar' samples with long distances plot far apart (e.g. Vermeesch 2013). In Figure 3, two plots present the same MDS results of all 30 'efficient' Cambrian-Triassic detrital zircon samples compiled in this study. The left panel is based on diachronous stratigraphic units (Figure 3(a)), and the right panel shows individual detrital zircon distributions using pie diagrams (Figure 3(b)). Similarly, the MDS-space distributions suggest that nearly only the contemporaneous detrital zircon samples have similar age spectra (i.e. short distances) and that the Carboniferous-Permian stratigraphic units, to a certain extent, inherit detrital zircon components from the Ordovician stratigraphic units (Figure 3(a)).
Theoretically, the Upper Triassic strata in the SQT should also retain significant detrital zircon components of predecessor units, such as ca. 1000-500 Ma age group (Figure 2(a-f)). However, they have totally different detrital age populations of ca. 1.85 Ga, ca. 450 Ma and ca. 270-240 Ma (Figure 2g-i)), and unique MDS-space distributions (Figure 3). The lack of associated magmatism within the SQT (Xu et al. 2020a, and references therein) excludes a proximal source. Recently, some studies have identified the ca. 169-147 Ma mid-ocean ridge-type ophiolite ) and ca. 141-135 Ma ocean Island  in the Bangong-Nujiang suture, suggesting that the Meso-Tethys Ocean to the south of the SQT did not close until the Middle Jurassic-Early Cretaceous (e.g. Zhu et al. 2013;Metcalfe 2021). Consequently, the distinctive detrital zircon spectra of the Upper Triassic strata indicate a predominantly northern detrital provenance ) rather than the southern detrital provenance of Gondwanan affinity (e.g. the Himalayas) as predecessor units. Therefore, the Upper Triassic and younger stratigraphic units should be excluded.
In summary, almost all Cambrian-Permian strata, which formed in relatively stable sedimentary basins (see Figure 11 in Xu et al. 2020a), record detrital age groups of ca. 2.50 Ga, ca. 950 Ma, ca. 800 Ma, and ca. 550 Ma (Figure 2), largely resulting from the polycycle of sedimentary rocks, and these similar age populations seemingly could be used in future studies as Palaeozoic indicators for the SQT, whereas some occasional age populations of ca. 1.85 Ga, 1.60 Ga, and ca. 1.11 Ga cannot represent the detrital zircon signature of the long-lived SQT.

Tracing the pre-Ordovician crystalline basement of the SQT (Proto-Tethys)
The identification of basement rocks is significant for revealing the origin and evolution of ancient microcontinents (e.g. Wu et al. 2019). The crystalline basement of several microcontinents composing the Tibetan Plateau (e.g. NQT, SQT, Lhasa block, Himalayan margin) is still unclear because of (1) less research limited by harsh natural conditions, (2) strong modification related to multiple tectonothermal events (e.g. reworking of basement rocks; Metcalfe 2021), and (3) coverage of young sedimentary successions (c.f., Wang et al. 2013). Consequently, the crystalline basement of the SQT has always been enigmatic. For example, previous studies on zircon Hf and whole-rock Nd model ages from Phanerozoic granitoids within the SQT suggest the presence of Proterozoic crystalline basement (e.g. Li et al. 2018a;Liu et al. 2019), whereas Dan et al. (2020) recently argued against such an interpretation and emphasized that the oldest igneous rocks (i.e. Ordovician S-type granites) to date represent the crystalline basement of the SQT. Note that some studies have reported the sporadic distribution of Palaeo-Neoproterozoic basement rocks in the microcontinents farther south (e.g. Lhasa block, Hu et al. 2018;Chen et al. 2019;Dong et al. 2020;Himalayan margin, Zhang et al. 2012a;Dong and Tian 2019). It is fascinating whether the SQT also has a pre-Ordovician crystalline basement because this terrane was separated from the northern margin of Gondwana until at least the late Carboniferous-early Permian (Zhang et al. 2012b;Xu et al. 2016;Li et al. 2019a), and the detrital zircon samples compiled in this study provide key information to trace the pre-Ordovician crystalline basement of the SQT.
In the Ordovician basement model, Dan et al. (2020) suggested that the SQT evolved from a Cambrian backarc basin with massive sediments. Subsequently, these sediments were heated by the upwelling of asthenospheric mantle-derived magma and further formed Ordovician S-type granites, which constitute the basement of the SQT (Figure 4(a)). However, the revised Cambrian stratigraphic units (the Dabure area, Section 4; Figure 1(b)) raise the possibility of pre-Ordovician basement in the SQT. In addition, the Cambrian-Ordovician stratigraphic units in the SQT mainly consist of medium-coarse clastic rocks (Table 1; e.g. sandstone, Pullen et al. 2011;Yang et al. 2014;Wang et al. 2016a), also indicating sedimentation on a continental margin rather than a mature back-arc basin.
As mentioned above, the Late Triassic detrital zircon samples along the southernmost margin of the SQT (Figure 1(b)  the late Cambrian-Ordovician granites with those of ca. 520-430 Ma detrital zircons from the Late Triassic detrital zircon samples. In Figure 4(b), magmatic zircons from the late Cambrian-Ordovician S-type granites in the SQT yield remarkably negative εHf(t) values (−20.1 to −0.8; Liu et al. 2016a;Dan et al. 2020;Wang et al. 2020b) that are distinct from the predominantly positive εHf(t) values (mostly −0.9 to +12.5) of the coeval detrital zircons from the Late Triassic sedimentary units within the Qiangtang block (Figure 4(b); e.g. Wang et al. 2016b), suggesting strongly juvenile components to the north of the SQT and probably corresponding to the coeval arcrelated magmatism in the NQT (Figure 4(a); Xu et al. 2020a). Note that (1) the late Cambrian-early Silurian supra-subduction zone (SSZ) ophiolites are reported only within central Qiangtang separating the SQT from the NQT (Figure 1(b); e.g. Zhai et al. 2016), (2) the early Palaeozoic S-type granites are distributed only along the northern margin of the SQT (Figure 1(b); i.e. Duguer, Gemuri and Bengsong Co areas, Liu et al. 2016a;Dan et al. 2020;Wang et al. 2020b), and (3) the Permian-Middle Triassic stratigraphic units in the southern margin of the NQT have a striking detrital zircon age population of ca. 450 Ma (see text below). It is thus reasonable that the southward subduction of the Proto-Tethys Ocean opened a back-arc basin between the SQT and NQT; simultaneously, the upwelling of asthenospheric mantle-derived magma resulted in the formation of the Ordovician S-type granites in the northern margin of the SQT (Figure 4(c)). From this point of view, the NQT is a typical arc terrane with more juvenile materials, whereas the SQT retains pre-Ordovician/Cambrian crystalline basement similar to the microcontinents farther south.

Constraints on the origin of the Late Triassic CQMB (Palaeo-Tethys)
As discussed above, the tectonic origin of the Late Triassic CQMB is the most controversial issue related to Palaeo-Tethyan evolution on the Tibetan Plateau, and two opposite models have been proposed. The 'in situ suture model' infers the presence of a Carboniferous-Middle Triassic Palaeo-Tethys Ocean between the SQT and NQT and the formation of the CQMB from Late Triassic northward oceanic subduction and subsequent exhumation ( Figure 5(a); Li et al. 1995a;Zhai et al. 2011;Zhang et al. 2016;Xu et al. 2021). The 'mélange underthrusting model' suggests the generation of the CQMB by mélange underthrusting and exhumation within the united Qiangtang block following the southward lowangle subduction of the Jinsha Palaeo-Tethys Ocean ( Figure 5(b); Kapp et al. 2000, Kapp et al. 2003Pullen et al. 2008. Undoubtedly, the most intuitive approach to settle this debate is the comparison of Carboniferous-Middle Triassic stratigraphic units between the SQT and NQT because there should theoretically be obvious distinctions in detrital zircon signatures if the Palaeo-Tethys Ocean separates the SQT from the NQT.
Many previous comparative studies report similar detrital age distributions of the associated stratigraphic units from the SQT and NQT (e.g. Gehrels et al. 2011;Pullen et al. 2011;Lu et al. 2017;Jian et al. 2019), and they further propose a single crustal fragment for these two terranes to support the 'mélange underthrusting model' (e.g. Gehrels et al. 2011;Pullen et al. 2011). However, the tectonic boundary between the SQT and NQT is not well-  Dan et al. 2020;Wang et al. 2020b) and Late Triassic detrital zircon samples (Wang et al., 2016b). CHUR -chondritic uniform reservoir. (c) Schematic model illustrating the petrogenesis of early Palaeozoic S-type granites on the northern margin of the SQT (after Xu et al. 2020a).
defined (e.g. Pullen and Kapp 2014). Consequently, which terrane the detrital zircon samples belong to is controversial, especially near the tectonic boundary (Xu et al. 2020a). In this study, we mark the locations of the so-called 'ophiolites' and high-pressure metamorphic rocks in central Qiangtang and outline the range of central Qiangtang to eliminate uncertainty about the affiliations of detrital zircon samples (Figure 1(b)). We further collate 17 detrital zircon samples from Cambrian-Triassic stratigraphic units in the NQT [e.g. Gehrels et al. (2011);He et al. (2011);Liang et al., 2019;Figure 1(b); note that the lower Palaeozoic strata are intruded by Middle Ordovician granitoids (e.g. Zhao et al. 2014), whereas the upper Palaeozoic-lower Mesozoic strata contain abundant fossils (e.g. Liang et al. 2020)]. After filtration based on the criteria in Section 3, nine 'eligible' detrital zircon samples (Table 1) are used to discuss the nature of the CQMB.
As shown in Figure 5( Figure 5(c)), whereas the Lower-Middle Triassic stratigraphic units are predominated by a middle Permian population ( Figure 5(c)), probably muting the presence of older zircon groups. We further omit all ages <300 Ma in these Mesozoic strata and use only older ages to plot the frequency histogram, in which the major age populations of ca. 980 Ma and ca. 450 Ma are also identified (see inserted figure in Figure 5 (c)). Overall, the Cambrian-Middle Triassic stratigraphic units in the NQT have significant lateral variability in their detrital zircon distributions ( Figure 5(d)). Specifically, the younger stratigraphic units have higher proportions of young detrital materials ( Figure 5(c,d)).
When we compare the detrital signatures of these two terranes, the Cambrian-Lower Ordovician stratigraphic units in the NQT show the same detrital zircon age distributions as those of their Ordovician counterparts in the SQT (i.e. ca. 2.50 Ga, ca. 970 Ma, ca. 810 Ma and ca. 600 Ma age groups; Figure 5(c)), and they have similar MDS results ( Figure 5(d)), suggesting a united origin for these two terranes during the Cambrian-Ordovician (e.g. Huang et al. 2018;Xu et al. 2020a). Note that some Permian detrital zircon samples in the NQT (e.g. sample 06GT43) cluster closely with some late Permian detrital zircon samples in the SQT (e.g. sample D18T16; Figure 5(d)), possibly resulting from the similarity of old age populations of ca. 2.50 Ga, 1.10 Ga and 960 Ma (Figure 5(c)). However, only upper Palaeozoiclower Mesozoic stratigraphic units in the NQT have significantly young age populations of ca. 450 Ma and ca. 260 Ma (Figure 5(c,d)), suggesting different detrital zircon age probability density functions for upper Palaeozoic-Triassic strata between the SQT and NQT ( Figure 5(c)) and arguing against previous views (e.g. Gehrels et al. 2011;Pullen and Kapp 2014;Lu et al. 2017). Therefore, the NQT was not adjacent to the SQT during the late Palaeozoic-early Mesozoic, and our comparison of detrital zircon signatures of the SQT and NQT provides solid evidence to support the 'in situ suture model' for the origin of the CQMB (Figure 5(b)).
Some recent studies, which support the 'in situ suture model', propose that the NQT records ca. 360 Ma and ca. 260 Ma arc-related magmatism, resulting from two periods of northward subduction of the Palaeo-Tethys Ocean between the SQT and NQT, whereas the magmatic gap corresponded to flat subduction (e.g. Wang et al. 2017;Zhai et al. 2018). Few studies have suggested that the Riwanchaka Formation in central Qiangtang contains Carboniferous Cathaysian warm fossils and abundant Late Devonian-early Carboniferous (ca. 360 Ma) detrital zircons and should belong to the NQT (e.g. Peng et al. 2019); in addition, ca. 260 Ma magmatism was revealed by detrital zircon populations of the Lower-Middle Triassic strata in the NQT (Figure 5(c)). It seems that the two-stage northward subduction model matches the current geological observations well. However, the Lopingian strata in the SQT (i.e. the passive margin of the Palaeo-Tethys Ocean) also have a minor detrital zircon population of ca. 358 Ma ( Figure 5(c)), arguing against the Late Devonian-early Carboniferous northward subduction of the Palaeo-Tethys Ocean. Furthermore, we suggest that the presence of ca. 360 Ma detrital zircons from the upper Palaeozoic strata in both the NQT and SQT probably support the opening of the Palaeo-Tethys Ocean as a Late Devonian-early Carboniferous back-arc basin between the SQT and NQT.

Records of subduction of the Palaeo-Tethys oceanic plate rather than opening of the Meso-Tethys Ocean by Lopingian strata in the SQT
Following the opening of the Palaeo-Tethys Ocean, the Meso-Tethys Ocean probably opened along with the northward subduction of the Palaeo-Tethys oceanic plate (c.f., Metcalfe 2013Metcalfe , 2021Zhu et al. 2013;Xu et al. 2020a). The Bangong-Nujiang suture, as a > 2000 km tectonic mélange, marks the opening and closure of the Meso-Tethys Ocean and participates in shaping the present tectonic framework of the Tibetan Plateau (Zhu et al. 2013;Li et al. 2019a;Metcalfe 2021). Despite many geological studies, the timing of the opening of the Bangong-Nujiang Meso-Tethys Ocean is debated and ranges from the early Permian to the Early Jurassic (c.f., Qu et al. 2010;Huang et al. 2012;Metcalfe 2013;Chen et al. 2017).
Theoretically, the associated stratigraphic units provide an ideal view to address this controversy because the terranes on both sides should have distinct detrital zircon signatures after their separation by continental rifting or ocean formation (Figure 6(a)). Considering the palaeogeographic linkage between the SQT and the Tethys Himalaya before the opening of the Meso-Tethys Ocean (Figure 6(a); Zhu et al. 2011, Zhu et al. 2013, we compare the Carboniferous-Permian spectra of the SQT with those of the Tethys Himalaya. As shown in Figure 6(b), we present the MDS results of seven compiled detrital zircon samples from the Tethys Himalaya. Outwardly, the MDS-space distributions of these limited samples could correspond to those of Carboniferous-Permian detrital zircon samples from the SQT to a certain extent (Figure 6(b)). In fact, the depositional ages of the Tethys Himalaya detrital zircon samples are not accurate and range from Carboniferous to Permian (e.g. Gehrels et al. 2011;Bhandari et al. 2019). In addition, the Carboniferous-Permian stratigraphic units in the Tethys Himalaya and the SQT are formed on a relatively gentle continental margin with only rift-related magmatism (e.g. Metcalfe 2013Metcalfe , 2021Zhu et al. 2013;Xu et al. 2016Xu et al. , 2020a. Consequently, most Tethys Himalaya detrital zircon samples have similar age groups of ca. 2.55-2.40 Ga, ca. 1000-875 Ma, and ca. 600-475 Ma (Figure 6(c)). We suggest that the various MDS-space distributions of the Tethys Himalaya detrital zircon samples (Figure 6(b)) mainly resulted from sedimentary  Table 1 recycling. In this sense, current comparisons of detrital zircon signatures cannot provide effective information for dating the opening of the Meso-Tethys Ocean.
Recently, Fan et al. (2021) found that, compared with Pennsylvanian-Cisuralian strata in the SQT, the Lopingian stratigraphic units in the SQT first record minor late Palaeozoic detrital age populations of ca. 360 Ma and ca. 260 Ma (Figures 2(e,f), and 5(c), and they suggested that this ~280-260 Ma provenance change marks the opening of the Bangong-Nujiang Meso-Tethys Ocean (Fan et al. 2021). Undoubtedly, the ca. 257 Ma arc magmatism in the NQT (see Xu et al. 2020a) cannot be the detrital provenance of these Lopingian strata in the SQT because (1) the Longmu Co-Shuanghu Palaeo-Tethys oceanic plate was subducted northward beneath the NQT during the late Permian ( Figure 5(a); e.g. Zhai et al. 2018), and (2) the magmatic arc rocks in the NQT display negative εHf(t) values (mostly −9.1 to −25.3), whereas the Permian detrital zircons from the Lopingian strata in the SQT yield relatively positive εHf(t) values from −8.6 to +12.5 (Figure 6(d); Liang et al. 2020;Fan et al. 2021).
Note that the ca. 320-270 Ma mafic dikes (Figure 6 (e)), which probably facilitate the opening of the Meso-Tethys Ocean, are widespread in the SQT (Zhai et al. 2013;Xu et al. 2016;Wang et al. 2019b). Fan et al. (2021) found that the Lopingian sandstone (sample B19T17) in the SQT contains ca. 299-285 Ma detrital  (Fan et al. 2021); Lower Triassic strata in the NQT (Liang et al., 2020); ca. 320-270 Ma mafic dikes in the SQT (Zhai et al. 2013;Wang et al. 2019b). zircons, which have internal zoning in cathodoluminescence images and Hf isotope compositions similar to those of ca. 320-270 Ma mafic dikes (Figure 6(d)). They thus consider these mafic dikes a significant detrital provenance for the middle-late Permian age population of the Lopingian strata in the SQT (Fan et al. 2021). However, we find that sample B19T17 has a detrital age distribution similar to that of Upper Triassic strata (Figure 2(g,h)) and that it was deposited during the Late Triassic (Section 4). In fact, the middle Permian-Early Triassic detrital zircons in the 'real' Lopingian sandstone samples (D18T16 and D18T17) have age distributions and Hf isotope compositions that are totally distinct from those of ca. 320-270 Ma mafic dikes in the SQT (Figures 5(c), and 6(d), 6(e)), arguing against the possible detrital provenance link between them. In addition, the Artinskian-Kungurian fauna with Cathaysian affinity in the SQT (e.g. Shen et al. 2016) suggests that the SQT separated from Gondwana during the early Permian (c.f., Xu et al. 2020a). Therefore, the Lopingian strata with a ca. 260 Ma age population in the SQT cannot record the opening of the Bangong-Nujiang Meso-Tethys Ocean.
The Lopingian strata in the SQT contain a minor ca. 360 Ma age population ( Figure 5(c)). As mentioned above, previous studies have shown that Late Devonian-early Carboniferous magmatic rocks are predominantly reported in the central part of the Qiangtang block, and these rocks formed in a back-arc basin between the NQT and SQT that was opened by the southward subduction of an oceanic plate to the north of the NQT (Section 5.3; see details in Xu et al. 2020a). In this sense, the associated magmatic rocks could be distributed on both the northern margin of the SQT and the southern margin of the NQT. Consequently, these Late Devonian-early Carboniferous magmatic rocks provide a northern detrital provenance for the Lopingian strata in the central part of the SQT (Figure 1(b)). We thus infer that the late Permian age population in the Lopingian strata ( Figure 5(c)) also has a northern detrital provenance.
As Fan et al. (2021) stated, the Lopingian sandstones in the SQT (i.e. samples D18T16 and D18T17 in the Jiao Co area) mainly consist of limestone and basalt fragments with angular to subangular shapes, suggesting a proximal source within the SQT. Although coeval basalts have been reported in the NQT (e.g. Wang et al. 2018;Li et al. 2018b), they cannot be the detrital provenance of the Lopingian sandstones in the SQT because these two terranes were separated by the Palaeo-Tethys Ocean during the late Permian (Section 5.3; Xu et al. 2020a). Recently,  reported middle-late Permian basalt (ca. 265 Ma) between the Lopingian strata in the SQT and the Shuanghu blueschist in central Qiangtang (Figure 1(b)). Therefore, the middle-late Permian basalt on the northern margin of the SQT is the most likely detrital provenance for the Lopingian strata in the central part of the SQT. The middle-late Permian basalt has a typical intraplate geochemical signature related to an extensional setting and is considered a middle Permian continental rift in the central part of the single Qiangtang block . However, this interpretation is inconsistent with the presence of the Late Devonian-Middle Triassic Palaeo-Tethys Ocean between the NQT and SQT (Section 5.3; see details in Xu et al. 2020a). Thus, there should be a more reasonable explanation for the formation of the middle-late Permian basalt in the northern passive margin of the SQT (Figure 6(a)).
Recent studies suggest that the northward subduction of the Palaeo-Tethys oceanic plate produced the associated blueschist and SSZ ophiolites in the earlymiddle Permian (Figure 6(a); e.g. Xu et al. 2021); subsequently, the rollback of the Palaeo-Tethys oceanic plate resulted in lithospheric extension of the upper plate and further opened a back-arc basin between the NQT and South China block in the late Permian (ca. 259-254 Ma; Figure 6(a); e.g. Liu et al. 2016b;Zhang et al. 2017b;Wang et al. 2018). Numerical modelling shows that most of the slab pull force can be transmitted through the bending area along the subducting slab during slab rollback (Capitanio et al. 2009). Consequently, the propagated slab pull force could produce important extensional deformation in the pre-existing weakness zone of the subducting plate (e.g. the Red Sea-Gulf of Aden rift system, Bellahsen et al. 2003). Considering that (1) the ca. 320-270 Ma mafic dikes related to a mantle plume are widespread in the SQT (Figure 6(a); Zhai et al. 2013;Xu et al. 2016;Wang et al. 2019b), and (2) the SQT experienced Cambrian-Early Ordovician and Late Devonianearly Carboniferous oceanic subduction (e.g. Zhai et al. 2016), the lithosphere of the SQT may have been weakened during multiple tectonothermal events. We thus propose that the middle-late Permian basalt on the northern margin of the SQT (Li and Wang 2019) represents typical passive margin magmatism driven by propagated slab pull force from the northward subduction of the Palaeo-Tethys oceanic plate (Figure 6(a)). In this context, the Lopingian strata in the SQT were deposited in an intraplate extensional setting and record the northward subduction of the Palaeo-Tethys oceanic plate.

Conclusions
After confirming the southern and northern tectonic boundaries of the SQT, we collated thirty Cambrian-Triassic detrital zircon samples from this terrane, and we found that the diachronous strata in the SQT have internally consistent detrital zircon signatures. The Cambrian-Permian strata that formed in a relatively stable continental margin have almost universal detrital zircon populations, including ca. 2.50 Ga, ca. 950 Ma, ca, 800 Ma, and ca. 550 Ma, which can be considered Palaeozoic indicators of the SQT for future studies that compare Tethyan detrital zircon data.
The synthesis of detrital zircon data from the SQT and comparison to detrital zircon distributions of adjacent microcontinents (e.g. NQT) provides key information for the complicated history of Tethyan evolution in the central Tibetan Plateau. (1) The Cambrian-Ordovician southward subduction of the Proto-Tethys oceanic plate opened a back-arc basin between the SQT and NQT and produced some Ordovician S-type granites on the northern margin of the SQT, which cannot represent the crystalline basement of the SQT. (2) Different from the detrital zircon signature of the SQT, the Permian-Middle Triassic strata in the NQT lack ca. 550 Ma age group and contain significant Palaeozoic age populations of ca. 450 Ma and ca. 260 Ma, suggesting the presence of a late Palaeozoic-early Mesozoic Palaeo-Tethys Ocean between the SQT and NQT. (3) The ca. 260 Ma detrital zircon population of the Lopingian strata in the SQT has a northern provenance, and it records passive margin magmatism related to the subduction of the Palaeo-Tethys oceanic plate rather than the opening of the Meso-Tethys Ocean, which probably opened during the early-middle Permian.