Middle–Late Triassic sedimentary provenance of the southern Junggar Basin and its link with the post-orogenic tectonic evolution of Central Asia

Due to the unknown Triassic volcanism in the Junggar Basin, the Middle–Late Triassic sedimentary provenance in the southern Junggar Basin (SJB) has long been controversial. Detrital zircon grains from 13 samples of the Middle–Upper Triassic Xiaoquangou Group in the SJB were analyzed using zircon U–Pb geochronology to constrain the provenance of Triassic sedimentary rocks and to further understand their source-to-sink system. Comparison of detrital zircon U–Pb age distributions for 13 samples reveals that the Triassic age populations predominate in sediments of the northern Bogda Mountains, with subordinate in the southern Bogda Mountains, and no or minimal in the North Tianshan (NTS). Coupled with sandstone petrological, sedimentary geochemical and paleocurrent data, the Triassic detrital zircon grains of the Xiaoquangou Group in the SJB were probably input from the Bogda Mountains. As Pennsylvanian and Mississippian zircon grains are mainly derived from the NTS and Central Tianshan (CTS), the provenance of the Xiaoquangou Group includes the NTS, CTS and Bogda Mountains. But the different samples in different sink areas have different provenances, originating from at least four source-to-sink systems. The supply of sediments from the Bogda Mountains started in the Late Triassic, suggesting initial uplift of the Bogda Mountains.

Detrital zircon U-Pb geochronology, petrography, paleocurrent measurements and sandstone petrography techniques are widely exploited for source-to-sink analysis [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18] . Of these, the rationale for detrital zircon U-Pb geochronology is the similarity of zircon U-Pb age distribution patterns from the source and sink areas 6,14 . However, comparison limitations emerge when direct evidence of related detrital zircon U-Pb ages is missing for potential source areas or when the source areas are poorly preserved 14 . This situation is common for deeptime sediment provenance analysis due to intense denudation in source areas and poor exposure of stratigraphic records 14,[19][20][21] . In recent years, researchers have established provenance analysis tools (e.g., probability density functions, kernel density estimates, multidimensional scaling, and cumulative distribution plots) to study the source signature when its source characteristics are unknown 4,11,22 . Some researchers also prove the effectiveness of these methods for inferring source characteristics, particularly when the source area is poorly preserved 18 .
In the southern Junggar Basin (SJB) of Central Asian Orogenic Belt (CAOB) (Fig. 1a,b), many detrital zircon grains with Triassic ages are found in sediments of the Middle-Upper Triassic Xiaoquangou Group in the Bogda Mountains 16,17 . These zircon grains show similar ages/epochs to the sedimentary rocks of the Xiaoquangou Group, indicating that the Xiaoquangou Group was deposited with syndepositional magmatism or volcanism. Moreover, the Xiaoquangou Group is subdivided into the Kelamayi Formation (Middle Triassic), Huangshanjie Formation (Upper Triassic) and Haojiagou Formation (Upper Triassic) from bottom to top (Fig. 1c). Compared to underlying strata, the detrital zircon U-Pb age-distribution pattern of the Upper Triassic Haojiagou Formation shows a significant multimodal distribution and presents a dramatic change with the first appearance of zircons from sub-contemporaneous Middle-Late Triassic magmatic sources 16,17 . These phenomena highlight the importance of the provenance of Triassic detrital zircon grains for understanding the source-to-sink system of www.nature.com/scientificreports/ half graben or Jimusaer graben 17,35 (Fig. 2). The Triassic strata of the SJB contain mudstone, sandstone and conglomerate with minor coal layers, which are interpreted as fluvial and lacustrine facies sweeping into an intracontinental basin 39 (Fig. 2). Of these, the Middle-Upper Triassic Xiaoquangou Group is marked by mudstone, argillaceous siltstone and sandy mudstone interbedded with conglomerate. Paleocurrents were mainly northward-directed in the western part of the SJB 12,16,17,41 , while in the eastern part of the SJB, paleocurrents were mainly northward-directed in the northern Bogda Mountains and southwestward-directed in the southern Bogda Mountains [42][43][44] (Fig. 3). The NTS is composed of the Devonian to Carboniferous arc-related igneous and volcaniclastic rocks, which have been interpreted as products of the southward subduction of NTS oceanic crust during the late Paleozoic 30,45,46 . The Central Tianshan (CTS) was a continental block with Precambrian basement, and is consist of the Paleozoic northern and southern magmatic belts which generated by the NTS oceanic crust subducting southwards below the CTS Block and the northward subduction of the STS oceanic crust 45 . The tectonostratigraphic units of the CTS comprise Proterozoic basement and early Paleozoic arc-type rocks intruded by late Paleozoic granitoids 32,46 . The Bogda Mountains, like the northern part of the NTS (Fig. 1c), contains mainly the Devonian to Quaternary sedimentary and igneous rocks 45,47 . Among these, Carboniferous bimodal volcanic rocks are widely distributed 48,49 , while the sedimentary rocks primarily include siliciclastics, carbonates, and volcanic breccia, and are characterized by the absence or minor quantities of quartzose sandstone or ophiolite 25 . The East Junggar is dominated by the Ordovician to Permian volcanics, siliciclastic rocks, limestones, and cherts, with an early Paleozoic Zhaheba-Aermantai ophiolite belt extending through its middle part 30,50 . The Silurian-Carboniferous arc magmatism was intense in the Yemaquan area along the southern part of the East Junggar systems, possibly resulting from the northward subduction of Kalamaili oceanic crust 51 . The West Junggar is characterized by Paleozoic volcanic-sedimentary sequences and granitoid intrusions. The Ordovician and Silurian rocks consist of low-grade schists, phyllites, and tuffaceous siliciclastic rocks. The Devonian to early Carboniferous strata is dominated by siliciclastic rocks and felsic-mafic volcanic rocks that mainly formed in marine environments. Above a late Carboniferous depositional hiatus lie Permian terrestrial volcanic and clastic successions 50 .

Sampling and analytical methods
Medium-to coarse-grained sandstones were collected from twelve sections and one well of the Middle-Upper Triassic Xiaoquangou Group in the SJB for detrital zircon U-Pb analysis (Fig. 1b) 16 , Wang et al. 17 , and Liu et al. 8 , respectively. Detrital zircon grains were separated using standard heavy-liquid and magnetic techniques, each sample followed by hand-picking of 250 random zircon grains under a binocular microscope. The grains then underwent cathodoluminescence (CL) imaging and U-Pb isotope analyses by LA-ICP-MS at Rockman Technology Co., Ltd microanalysis lab, Beijing, with detailed procedures following Thompson et al. 53 . The Resolution SE model laser ablation system (Applied Spectra, USA) was equipped with an ATL (ATLEX 300) excimer laser and a Two Volume S155 ablation cell. The laser ablation system was coupled to an Agilent 7900 ICPMS (Agilent, USA). The isotope ratios and element concentrations of the zircon grains were calculated using the Iolite software package 54 . Zircon 91500 and GJ-1 were used as primary and secondary reference materials, respectively. Correction for common lead used the method of Andersen 55 and the ages were computed using ISOPLOT 3 with uncertainties quoted at the 1σ standard deviation and 95% confidence levels 56 . The probability density plots and kernel density estimates of detrital zircon U-Pb ages were conducted by IsoplotR 22 .
The detrital ages of thirteen samples also collected for classical Multidimensional scaling (MDS) analysis which were conducted by IsoplotR 22 . MDS is a dimension-reducting technique that takes a matrix of pairwise dissimilarities between detrital age distributions as input and produces a configuration of two (or higher-) coordinates as output, so that the Euclidean distance between these coordinates approximates the dissimilarities of the input matrix. Thus, an MDS-configuration serves as a map in which similar samples cluster closely together and dissimilar samples plot far apart 22 .  The detrital zircon U-Pb age patterns for samples of the Haojiagou section and Xidalongkou sections in the northern Bogda Mountains, and Dabancheng and Yuergou sections, and well C in the southern Bogda Mountains are similar, with principal age peaks and relative proportions of age groups comparable for various locations (Fig. 5). The detrital zircon grains of the samples from the well C and Haojiagou section yield age patterns dominated by Triassic populations, with subordinate populations of Pennsylvanian and Cambrian-Devonian ages (Fig. 5). The detrital zircon grains from the Xidalongkou section yield age patterns dominated by Triassic populations, with subordinate populations of Mississippian and Cambrian-Devonian ages (Fig. 5). The detrital zircon grains from the Dabancheng section display a relatively lower proportion of the Triassic age group, yielding age patterns dominated by Pennsylvanian, accompanied by subordinate Triassic and Cambrian-Devonian populations (Fig. 5). The detrital zircon age group for the Yuergou section sample is dominated by Mississippian ages, with minor Cambrian-Devonian constituents (  www.nature.com/scientificreports/  www.nature.com/scientificreports/ 10,178 ppm, mostly 400-900 ppm), with average concentrations significantly above those for chondrite 59 (2.65 ppm). Most of the zircon grains, irrespective of the age groups, are depleted in light REEs and enriched in heavy REEs, as displayed in the chondrite plots 59 (Fig. 6). The majority of the analyzed zircon grains exhibit negative Eu and positive Ce anomalies ( Fig. 6; Supplementary Information; Eu/Eu * = 0.01-0.72; Ce/Ce * = 1.1-357.8), oscillatory zoning, low luminescence, and high Th/U values (> 0.1), which are diagnostic of a magmatic origin 2,3,57 (Fig. 7). The Triassic zircon grains (Fig. 6a) show REE concentrations and chondrite-normalized patterns similar to those of the Cambrian-Permian zircon grains (Fig. 6b-f; Supplementary Information). On the zircon U-Y, Eu/Eu*-Ce/Ce*, Yb/Sm-Y, and Ce/Ce*-Y diagrams, most of the zircon grains with age groups of Cambrian-Devonian, Mississippian, Pennsylvanian and Permian populations plot in granitoid (acidic rock) areas 3 (Fig. 8). Zircon with Triassic ages shows features similar to those of granitoid (acidic rock) and syenite (intermediate rock) 3 (Fig. 8).

Results
Multidimensional scaling. The detrital zircon U-Pb age probability and multidimensional scaling (MDS) plots were used in conjunction to establish a Middle-Late Triassic source-to-sink system of the SJB 22 (Figs. 5 and  9). In the MDS plot, 13 samples from the Middle-Upper Triassic Xiaoquangou Group in the SJB show distinct differences that generate four groups according to their detrital zircon age distributions (Fig. 9). The similar www.nature.com/scientificreports/ samples cluster closely together relative to their similar provenances, while the dissimilar samples plot far apart revealing that they have different source areas (Fig. 9). Samples 15C-33, 16DLK-49 and 17HJG-01 (group 1) from the well C and Xidalongkou and Haojiagou sections differ significantly from the others (Fig. 9). Samples XJ12-06, 16TDG-06, and 17DDLK-09 (group 2) from the Aiweiergou, Taodonggou and Dongdalongkou sections are assigned similar sediment sources owing to their close links in the MDS plot (Fig. 9). Samples 12SKS-07, 17JGMK-06, and 16DBC-04 (group 3) from the Sikeshu, Jiangongmeikuang and Dabancheng sections generally display a limited age variation and show proximity in the MDS plot (Fig. 9). Samples 15HXG-95, 17JJZG-10, 17BYH-61, and HX-01 (group 4) from the Haxionggou, Jingjingzigou, Baiyanghe and Yuergou sections show a mixed age variation and have proximity in the MDS plot (Fig. 9).    (Fig. 5). In particular, constraining the provenance of Triassic detrital zircon is difficult because reports suggest that there is relatively sparse direct field evidence of Triassic magmatic activity in the Junggar Basin and adjacent areas.
Although limited field evidence of Triassic magmatic or volcanic activity exists, numerous studies have indicated that Triassic magmatism or volcanism is extensively distributed both in the Junggar Basin and in the surrounding areas. In the Altai Mountains, the Late Triassic magmatism (zircon U-Pb age of 222 ± 2 Ma) was found based on the Huoremudeke granitic pluton 26  To constrain the provenance of Triassic detrital zircon grains in the SJB, detrital zircon U-Pb geochronological, sandstone petrological, sedimentary geochemical, and paleocurrent data were integrated. First, analysis of the depositional environment reveals that the Junggar Basin contains mainly lacustrine facies in the Middle-Late Triassic strata 16,17,25,39 (Fig. 2), and sedimentary rocks of many areas have high sediment textural maturity. However, in the Bogda Mountains, sediments of the Middle-Upper Triassic Xiaoquangou Group show low sediment textural maturity with angular to subangular lithic fragments and detrital zircon grains 16 (Fig. 7), suggesting a proximal source area. Second, paleocurrent observations show that the drainage system is perpendicular to the Bogda Mountains, with northward-directed paleocurrents in the northern Bogda Mountains, and southwestward-directed paleocurrents in the southern Bogda Mountains (Fig. 3) This suggests that the Triassic detrital zircon grains are derived from the Bogda Mountains and are restricted to the Urumqi-Jimusaer area and probably the Haojiagou-well C-Xidalongkou areas. In summary, the provenance of Triassic detrital zircon grains from sediments of the Xiaoquangou Group in the SJB is the nearest Bogda Mountains. Note that not all the Triassic zircon grains are derived from the Bogda Mountains; for example, the Balikun area may have received material from the Barkol Tagh (Fig. 2).
In addition, the trace element composition of detrital zircon grains is sensitive to the source rock type 2,3,57,63 . The variations in representative elements and ratios of REEs for zircon grains indicate that the Triassic detrital zircon originated from intermediate-acidic igneous rocks, with probable dominance of intermediate igneous rocks 2,3 (Fig. 8). We attribute the Triassic volcanic rocks are likely interlayered in the Middle-Upper Triassic Xiaoquangou Group of the Bogda Mountains, as suggested by BGMRXUAR 25 .
This study corroborates the utility of the comparison of detrital zircon U-Pb age distributions for identifying source locations, especially when the potential source areas lack direct zircon field evidence (e.g. 11,22 ). It also confirms that this provenance analysis tool is an effective approach for deciphering source characteristics in deep-time source-to-sink system.
Source-to-sink systems of the SJB in the Middle-Late Triassic. The systematic analysis of detrital zircon grains preserved in the Xiaoquangou Group allowed us to discriminate their source areas and establish the source-to-sink systematics in the Middle-Late Triassic for the SJB. The provenance analysis of thirteen samples partitions the SJB and adjacent areas into four groups (Fig. 10), which is consistent with groups from the MDS plot (Fig. 9). The four groups are derived from at least four source-to-sink systems in the Middle-Late Triassic time (Fig. 10). www.nature.com/scientificreports/ Group 1 from the well C, and Xidalongkou and Haojiagou sections has high contents of Triassic zircon grains (Figs. 5 and 9), indicating a predominant contribution from the Bogda Mountains. Moreover, group 1 contains subordinate Carboniferous and Cambrian-Devonian zircon grains, revealing that the NTS and/or CTS also provided sediments to the northern Bogda Mountains (Figs. 5 and 9). That is, the provenance of sediments for group 1 is primarily attributed to the Bogda Mountains, with secondary contributions from the NTS and/or CTS. Group 2 from the Aiweiergou, Taodonggou and Dongdalongkou sections includes lots of Mississippian and Cambrian-Devonian zircon grains, suggesting that the CTS are likely major sources (Figs. 5 and 9). Group 3 from the Sikeshu section displays a limited age variation that is dominated by Pennsylvanian (Fig. 5), combined with the compositionally and texturally immature detrital grains, indicating that group 3 is originated from the igneous rocks of the NTS with zircon U-Pb ages of 320-300 Ma (Fig. 5). Group 4 from the Haxionggou, Jingjingzigou, and Baiyanghe sections shows a mixed age variation of Pennsylvanian (320-300 Ma) and Mississippian (360-320 Ma), revealing that the major provenance of group 4 is from the NTS and CTS (Figs. 5 and 9). Summarising, the Bogda Mountains, CTS, and NTS appear to be the dominant sources for group 1, group 2, and group 3, respectively (Fig. 9); a mixture of the NTS and CTS appears to be the dominant source for group 4 (Fig. 9).
Although the source areas of the SJB include the Bogda Mountains, NTS, and CTS in the Middle-Late Triassic, the western (front of the NTS) and eastern (Bogda Mountain areas) parts of the SJB exhibit inconsistencies in their provenance patterns. In the Bogda Mountains, the provenance is attributed to the Bogda Mountains itself and the NTS (Figs. 9 and 10). Of these, the Bogda source mainly provides syndepositional and recycled sediments. In the NTS areas, the NTS piedmont (e.g., Sikeshu and Jiangongmeikuang sections) received material mainly from the NTS, with minor input from the CTS and no input from the Bogda Mountains. In the transition zone of the western and eastern parts of the SJB, such as the Haojiagou section area, the basin received material from both the NTS and Bogda Mountains. In conclusion, the NTS accounts primarily for the provenance of the western part of the SJB (front of the NTS), while sediments from the eastern part of the SJB (Bogda Mountains areas) are principally derived from the Bogda Mountains (Figs. 2,5,9,10).

Implication for the tectonic evolution of the southern Central Asian Orogenic Belt. The Mid-
dle-Upper Triassic Xiaoquangou Group in well C, and Haojiagou and Xidalongkou sections yields a complex age population with many syndepositional Triassic detrital zircon grains from the Bogda Mountains (Fig. 5), indicating the initiation of sediment supply from the Bogda Mountains in the same area. We attribute this abrupt change to the initial uplift of the Bogda Mountains 17 (referred to as the uplift stage). Moreover, the Xiaoquangou Group in the Haojiagou and Xidalongkou sections contains lithic fragments 16,40 , with sediments characterized by moderate chemical weathering and sedimentary recycling 17 . This suggests that the Bogda Mountains provided syndepositional volcanics and recycled siliciclastics to the basin. In addition, according to the tectonic cycle identified by previous works 17,33,35 (the SJB entered in the tectonic evolution of post-orogenic since the Pennsylvanian, of these, the Pennsylvanian to Guadalupian referred to as the syn-rift stage which characterized by faulting, the Lopingian to Early Triassic referred to as the post-rift stage which is characterized by depression with little faulting, and the Middle-Late Triassic referred to as the uplift stage due to the uplift of the Bogda Mountains), we draw cumulative frequency curves of the lag time 5 (lag time = crystallization age -deposition age; Fig. 11) for the Cisuralian Shirenzigou Formation to the Upper Triassic Haojiagou Formation. In the cumulative frequency curves of the lag time 5 (Fig. 11), the sediments of the uplift stage show a near "bimodal lag time", including low and high lag time groups, compared to mainly low lag time groups for syn-rift and post-rift sediments 5 (Fig. 11). The high lag time sediments probably arose from sedimentary recycling, while the low lag time sediments are attributed to syndepositional magmatic activity. www.nature.com/scientificreports/ Although the proportion of Cambrian-Devonian, Mississippian, Pennsylvanian, and Permian populations diminishes, these populations do not vanish compared to the underlying units 16,17 . This implies that both the NTS and CTS continued providing sediments for sedimentation in the Bogda Mountains. This is further evidenced by the detrital zircon U-Pb age cumulative frequency curves, probability plot and kernel density estimates, which exhibit zircon age groups such as syn-rift and post-rift sediments (Figs. 5 and 11). We attribute this to the Late Triassic uplift of the Bogda Mountains, but the relatively smooth relief of its topography allowed the scouring by rivers and consequently the supply of detritus from the NTS and CTS (Figs. 2 and 10).
On the other hand, the data of this study has suggests that the Triassic magmatism or volcanism is extensively distributed both in the Junggar Basin and in the surrounding areas. The available thermal data including zircon (U-Th)/He, biotite 40 Ar- 39 Ar, titanite fission track, and apatite fission track dates suggest that the Central Asian has undergone extensive cooling and exhumation in the Triassic 64 . Summarizing we can show that the Triassic event, which has not received much attention, is widespread in the southern CAOB. The Triassic event in the southern CAOB coincides with the Qiangtang-Eurasia collision 65 . Thus, we suggest that the Qiangtang Block collision with Eurasia probably induced fault reactivations and basement exhumation within the southern CAOB 64 . The Late Triassic uplift of the Bogda Mountains resulted from the far field effect of the Qiangtang-Eurasia collision (e.g. 42,66 ). We conclude the Triassic event is a tectonic reactivation (induced by Qiangtang-Eurasia collision) within the post-orogenic tectonic evolution of the southern CAOB.

Conclusions
Comparison of detrital zircon U-Pb age distributions enabled the constraining of the location of Triassic volcanic activity in the SJB area. The comparison of the new detrital zircon U-Pb age dataset for the Middle-Upper Triassic Xiaoquangou Group of the SJB shows that Triassic volcanic activity occurred mainly in the Urumqi-Jimusaer area. The REE abundance of zircon suggests that Triassic volcanism resulted in mainly intermediate-acidic rocks. The geochronology data combined with provenance analysis indicate that provenance varies for different areas in the SJB, resulting in four sink areas. The Bogda Mountains started to supply sediments in the Late Triassic, marking the initial uplift of the Bogda Mountains. This study proves the effectiveness of the comparison of detrital zircon U-Pb age distributions for inferring source signatures in deep-time source-to-sink system.

Data availability
All data generated or analysed during this study are included in this published article and its Supplementary Information files. Figure 11. Cumulative frequency curves of the lag time (crystallization age-depositional age) for the Cisuralian Shirenzigou Formation to the Upper Triassic Haojiagou Formation. The data and tectonic evolution stages are from 16,17 . P 1 s, Shirenzigou Formation; P 1 t, Tashikula Formation; P 2 w, Wulapo Formation; P 2 jj, Jingjingzigou Formation; P 2 l a , lower Lucaogou Formation; P 2 l b , upper Lucaogou Formation; P 2 h, Hongyanchi Formation; P 3 q, Quanzijie Formation; P 3 wt, Wutonggou Formation; P 3 -T 1 g, Guodikeng Formation; T 1 j, Jiucaiyuanzi Formation; T 1 s, Shaofanggou Formation; T 2 k, Kelamayi Formation; T 3 hs, Huangshanjie Formation; T 3 hj, Haojiagou Formation. www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.