Introduction

Granitoid magmas are commonly generated throughout the evolution of an orogen from subduction to collision-related orogeny and post-orogenic extension (Chappell and White 1974, 1992; Pitcher 1983; Pearce et al. 1984; Frost and Mahood 1987; Sylvester 1989; Brown 1994). Understanding the petrogenesis of such magmas provides important insights into the geodynamic processes that operate in the deeper crust, including mantle-derived magmatism and the anatexis of pre-existing crustal material (Barbarin 1999, 2005; Bonin 2007; Koteas et al. 2010).

The Sanjiang orogenic belt incorporates the Jinshajiang–Ailaoshan, Lancangjiang, and Nujiang tectonic belts, and is located in the southeastern segment of the eastern Paleo-Tethys tectonic domain (Zhong 1998). The orogenic belt varies in orientation from the WNW–ENE trending Himalayan–Tethyan segment of the belt to the N–S trending southeastern Asian segment (Hutchison 1989; Metcalfe 1996a, b, 2002, 2013; Zhang et al. 2008, 2012, Wang et al. 2013, 2014; Deng et al. 2014). The Lancangjiang tectonic zone (also known as the Changning–Menglian suture belt) within this belt represents the eastern Paleo-Tethys main ocean that separated the Baoshan–Tengchong Block to the west from the Simao Block to the east (Mahawat et al. 1990; Cong et al. 1993; Metcalfe 1996a, b, 2002; Mo et al. 1998; Zhong 1998; Feng et al. 2005; Peng et al. 2008, 2013; Hennig et al. 2009; Wang et al. 2010; Yang et al. 2014). The magmatism in the east of this area generally occurred during orogenesis in the Paleo-Tethyan region between the late Paleozoic and the early Mesozoic. In addition, the westernmost Tengchong Block forms the southwestern part of the Sanjiang orogenic belt, which is thought to represent the southeastern extension of Lhasa Block (Xu et al. 2008, 2012; Li et al. 2011; Huang et al. 2013; Wang et al. 2014). Recent research has also identified Triassic granitoids within the Tengchong Block (Cong et al. 2010; Li et al. 2010, 2011; Zou et al. 2011; Huang et al. 2013), also suggesting that records tectonomagmatism associated with the evolution of the eastern Paleo-Tethys. However, it remains unclear whether these granitoids were derived from partial melting of crustal sources during collisional orogenesis as part of the evolution of the Sanjiang orogenic belt (Cong et al. 2010; Li et al. 2010; Zou et al. 2011) or whether they were derived from the partial melting of mantle-derived sources during the late Permian-to-Early Triassic subduction- and collision-related orogeny between the Lhasa and Northern Australia blocks (Huang et al. 2013), an event that also caused crustal melting during the late Triassic (Li et al. 2011). Consequently, the evolution of both the petrogenesis and the geodynamic setting of the late Permian-to-Late Triassic tectonic-associated magmatism in the Tengchong Block remains poorly constrained.

Here, we present new precise geochronological (zircon U–Pb ages), bulk-rock geochemical, and Sr–Nd–Pb isotopic data for late Paleozoic-to-early Mesozoic quartz diorite, granitoid, and metagabbroic units within the Tengchong Block. These data provide new insights that further constrain the Paleo-Tethyan evolution of this block and the southeastern Tibetan Paleo-Tethys belt (Yang et al. 2014).

Geological setting and petrography

The Sanjiang orogenic belt is part of the eastern Tethyan tectonic domain and represents the southeastern extension of the Himalayan Orogen (Kou et al. 2012; Fig. 1a–b). From east to west, this region is divided into the Yangtze, Lanping–Simao–Indochina, Baoshan–Shan–Thai, and Tengchong blocks (Wang et al. 2006, 2010, 2013, 2014; Chen et al. 2007; Metcalfe 2013; Deng et al. 2014), which are separated by the Jinshajiang–Ailaoshan, Changning–Menglian, and Bangong–Nujiang sutures, respectively.

Fig. 1
figure 1

a Distribution of main continental blocks of SE Asia in the Eastern Tethys Domain after Fan et al. (2010). b Distribution of principal magma and strata of Sanjiang Tethys tectonic Domain, SW China after YNBGMR (1990). All the age data in the diagram are from above reference therein. Geological map of Pianma area (c), Lianghe area (d), and western Yingjiang area (e) in the Tengchong block, SW China after YNBGMR (1990)

Full size image

The Tengchong Block forms the northern part of Gondwanan Sibumasu and is located in the southwestern segment of the Sanjiang orogenic belt (Fig. 1b). The presence of Permian–Carboniferous glacio-marine deposits and overlying post-glacial black mudstones and Gondwana-like fossil assemblages suggests that this block was derived from the margin of western Australia within the eastern Gondwana supercontinent and remained in a platform-type setting at a passive continental margin (Jin 1996; Xiao et al. 2003; Chen et al. 2006; Jin et al. 2002). The block contains Mesoproterozoic metamorphic basement material of the Gaoligong Mountain Group, which is overlain by late Paleozoic clastic sedimentary rocks and carbonates, and Tertiary–Quaternary volcano–sedimentary sequences, intruded by Mesozoic–Tertiary granitoids (YNBGMR 1990; Zhong 1998; Zhao et al. 2016). The Gaoligong Mountain Group contains quartzite, two-mica–quartz schist, feldspathic gneiss, migmatite, amphibolite, and marble units, and zircons from paragneiss and orthogneiss of this group yield zircon U–Pb ages of 1053–635 and 490–470 Ma, respectively (Song et al. 2010). The Paleozoic sedimentary units in this area are dominated by Carboniferous clastic rocks, upper Triassic to Jurassic turbidites, Cretaceous red beds, and Cenozoic sandstones (YNBGMR 1990; Zhong 2000).

The Tengchong Block contains voluminous granitic gneiss, migmatite, and leucogranite units that were previously thought to have formed during the Proterozoic (YNBGMR 1990). However, recent geochronological data indicate that this block also contains numerous late Paleozoic-to-early Mesozoic granitoids with zircon U–Pb ages of 245–206 Ma (Cong et al. 2010; Li et al. 2010, 2011; Zou et al. 2011; Huang et al. 2013). All these granitoids were emplaced into Paleozoic and Mesozoic units and some emplaced into the Gaoligong Mountain group. The long axis of these granitic plutons is parallel to the suture belt (Fig. 1b), but some were controlled by the regional fault (Fig. 1d). The granites at Gaoligong lie to the west of the Lushui–Luxi–Ruili Fault, have undergone strong shearing, and possess a schistosity dipping at a shallowly angle to the north and a subhorizontal mineral lineation (Wang et al. 2006; Zhang et al. 2011). Based on similar stratigraphy, paleobiogeography, and magmatism between the Tengchong and Lhasa Terranes, both have experienced similar tectonomagmatic history since the Early Paleozoic (Xie et al. 2016): as part of the northern margin of the Australian Gondwana in the Early Paleozoic, with moveward to the Eura-Asian continent from the Trassic and collision with Qiangtang–Baoshan during the Cretaceous.

In comparison, the Pianma quartz diorites (Fig. 1c) are fine-grained (Fig. 2a), are undeformed, and contain quartz veins. These intrusions contain 0.1–0.4 mm grains of quartz (15–20%), plagioclase (45–55%), K-feldspar (15–20%), biotite (10–15%), and hornblende (8–12%), as well as accessory apatite, zircon, ilmenite, titanite, and magnetite (total 3–5%; Fig. 3a). The Lianghe two-mica granite was emplaced into the Gaoligong Mountain group (Fig. 1d), is underformed coarse-grained and porphyritic (Fig. 2b), and contains quartz (35–40%), K-feldspar (28–33%), plagioclase (21–28%), biotite (5–10%), muscovite (~5%), and accessory apatite, zircon, and ilmenite (total ~1%; Fig. 3b). The Yingjiang metagabbros were emplaced into the Gaoligong Mountain Group (Fig. 1e), are fine-grained (Fig. 2c), are undeformed, and contain plagioclase (50–60%), K-feldspar (10–15%), biotite (10–15%), hornblende (8–12%), clinopyroxene (3–5%), quartz (<1%), calcite (<3%), and accessory apatite, zircon, ilmenite, titanite, and magnetite (total 3–5%; Fig. 3c). The Lianghe biotite granite was also emplaced into the Gaoligong Mountain Group (Fig. 1d), is coarse-grained and porphyritic (Fig. 2d), and records a little deformation. The granite contains quartz veins and consists of 0.2–0.8 mm grains of quartz (30–38%), K-feldspar (30–35%), plagioclase (25–30%), biotite (5–8%), and accessory apatite, zircon, ilmenite, and magnetite (total 1–3%; Fig. 3d).

Fig. 2
figure 2

Field petrography of quartz diorite in the Pianma area (a), two-mica granite (b), metagabbroic rock in the Yingjiang area (c), and biotite granite (d) in the Lianghe area of the Tengchong block, SW China

Full size image
Fig. 3
figure 3

Microscope petrography of quartz diorite in the Pianma area (a), two-mica granite (b), metagabbroic rock in the Yingjiang area (c), and biotite granite (d) in the Lianghe area of the Tengchong block, SW China

Full size image

Analytical methods

Major and trace elements

Prior to analysis, bulk-rock samples were trimmed to remove weathered surfaces, cleaned with deionized water, crushed, and powdered using a tungsten carbide ball mill to pass through a 200 mesh screen. Major element concentrations were determined using a Rikagu RIX 2100 X-ray fluorescence (XRF) spectrometer at the State Key Laboratory of Continental Dynamics, Northwest University, Xi’an, China and the Guizhou Tuopu Resource and Environmental Analysis Center, China. Repeat analyses of USGS and Chinese national rock standards (BCR-2, GSR-1, and GSR-3) indicate that major element analytical precision and accuracy are generally better than 5%.

Trace element concentrations were determined using inductively coupled plasma–mass spectrometry (ICP–MS; Bruker Aurora M90) at the State Key Laboratory of Continental Dynamics, Northwest University, and the Guizhou Tuopu Resource and Environmental Analysis Center using the methods of Qi et al. (2000). Prior to analysis, sample powders were dissolved using an HF + HNO3 mixture in a high-pressure PTFE bomb at 185 °C for 36 h. The ICP–MS analyses are estimated to have accuracies better than ±5 to ±10% (relative) for most elements.

Sr–Nd–Pb isotopic analyses

Bulk-rock Sr–Nd–Pb isotopic data were obtained using a Nu Plasma HR multi-collector (MC) mass spectrometer at the State Key Laboratory of Continental Dynamics, Northwest University, and the Guizhou Tuopu Resource and Environmental Analysis Center, using an approach similar to that of Chu et al. (2009). Sr and Nd isotopic fractionation was corrected to 87Sr/86Sr = 0.1194 and 146Nd/144Nd = 0.7219, respectively, and a Neptune MC–ICP–MS instrument was used to measure 87Sr/86Sr and 143Nd/144 Nd isotope ratios. NIST SRM-987 and JMC-Nd certified reference standard solutions were used for 87Sr/86Sr and 143Nd/144Nd isotopic ratios, respectively, and BCR-1 and BHVO-1 were used as reference materials.

Bulk-rock Pb was separated using an anion exchange and HCl–Br columns, with Pb isotopic fractionation corrected to 205Tl/203Tl = 2.3875. Thirty measurements of the NBS981 standard during the analytical period yielded average values of 206Pb/204Pb = 16.937 ± 1 (2σ), 207Pb/204Pb = 15.491 ± 1 (2σ), and 208Pb/204Pb = 36.696 ± 1 (2σ), and repeat analyses of the BCR-2 standard yielded values of 206Pb/204Pb = 18.742 ± 1 (2σ), 207Pb/204Pb = 15.620 ± 1 (2σ), and 208Pb/204Pb = 38.705 ± 1 (2σ). Total procedural Pb blanks were 0.1–0.3 ng.

Zircon U–Pb analyses

Prior to analysis, zircons were separated from four samples (each ~4 kg) taken from various sampling locations within the Tengchong Block, using the conventional heavy liquid and magnetic techniques. Representative zircons were then handpicked and mounted in epoxy resin disks before polishing and carbon coating. The internal morphology of these zircons was then imaged using cathodoluminescence (CL) prior to U–Pb analysis.

Laser ablation–ICP–MS (LA–ICP–MS) zircon U–Pb analyses were undertaken using an Agilent 7500a ICP–MS instrument equipped with a 193 nm laser at the State Key Laboratory of Continental Dynamics, Northwest University, following Yuan et al. (2004). The resulting 207Pb/206Pb and 206Pb/238U ratios were calculated using the GLITTER program and corrected using external calibration by analysis of a standard 91,500 Harvard zircon. The resulting factors were then applied to each sample to correct for both instrumental mass bias and depth-dependent elemental and isotopic fractionation, following Yuan et al. (2004). Common Pb contents were evaluated using the approach of Andersen (2002), and age calculations and concordia diagrams were made using ISOPLOT version 3.0 (Ludwig 2003). Uncertainties are quoted at the 2σ level.

Results

The metagabbroic rocks and granitoids of the Tengchong Block include late Permian quartz diorites in the Pianma area, Late Triassic and earliest Jurassic granites in the Lianghe area, and latest Triassic metagabbroic rocks in the Yingjiang area. The data obtained in this study are given in Tables 1 (major and trace elements), 2 (Sr–Nd isotopic data), and 3 (Pb isotopic data).

Table 1 Major (wt%) and trace element(ppm) analysis result of granitoids and metagabbroic rocks from Tengchong block
Full size table
Table 2 Whole-rock Rb–Sr and Sm–Nd isotopic data for granitoids and metagabbroic rocks from Tengchong block
Full size table
Table 3 Whole-rock Pb isotopic data for granitoids and metagabbroic rocks from Tengchong block
Full size table

Zircon U–Pb ages

The complete data set for the zircon U–Pb analysis is given in the Supplementary Data set Table, with sampling locations, lithologies, and results summarized in the Supplementary Data set Table and in Fig. 1, 2. The zircons are generally euhedral, up to 100–400 μm long, and have length:width ratios of 2:1–4:1 (Fig. 4). The majority of these zircons are colorless or light brown, prismatic, transparent to translucent, have clear oscillatory zoning visible during CL imaging, and Th/U ratios >0.4, all of which are indicative of a magmatic origin (Hoskin and Schaltegger 2003).

Fig. 4
figure 4

Zircon CL images for the quartz diorite (a PM48), two-mica granite (b LH02), metagabbroic rock (c NB30), and biotite granite (d LL128) in the Tengchong block, SW China

Full size image

The 36 spot analyses of zircons from quartz diorite (sample PM48) included seven inherited zircons with ages of 1581–478 Ma and five discordant analyses that are not discussed here. The youngest 10 analyses are scattered and have a wide range of 206Pb/238U ages (240 ± 5–197 ± 4 Ma) that may reflect Pb loss. Fourteen coherent analyses of zircons from this sample yield 206Pb/238U ages of 251 ± 4.87–271 ± 5.31 Ma, with a weighted mean age of 263.3 ± 3.6 Ma (MSWD = 1.4, n = 14, 2σ) (Fig. 5a).

Fig. 5
figure 5

LA-ICP-MS U–Pb zircon concordia diagram and CL images of representative zircon grains for the quartz diorite (a PM48), two-mica granite (b LH02), metagabbroic rock (c NB30), and biotite granite (d LL128) in the Tengchong block, SW China

Full size image

Some 42 spot analyses were undertaken on zircons from two-mica granite (sample LH02). Four of these yielded unreliable and discordant results with further 10 yielding anomalous Th/U values (<0.1), none of which are discussed further. Six inherited zircons within this sample yielded ages between 2393 ± 41 and 507 ± 10 Ma, all of which are thought to represent inherited zircons derived from ancient crustal material. The remaining 32 coherent analyses yield 206Pb/238U ages from 201 ± 4 to 242 ± 5 Ma and a weighted mean age of 218.5 ± 5.4 Ma (MSWD = 1.7, n = 22, 2σ) (Fig. 5b).

Of 36 analyses on zircons from metagabbros (sample NB30), 14 are discordant and are not discussed further. One analysis yielded a slightly old 206Pb/238U age of 232 ± 2 Ma, with a further three spots yielding slightly young 206Pb/238U ages (181 ± 2–189 ± 2 Ma) that may reflect radiogenic Pb loss. Eighteen coherent analyses of zircons from this sample yielded 206Pb/238U ages from 195 ± 2 to 216 ± 2 Ma and a weighted mean age of 205.7 ± 3.1 Ma (MSWD = 2.5, n = 20, 2σ) (Fig. 5c).

Thirty-six analyses were undertaken on zircons from biotite granites (sample LL128), with a single analysis (spot 25) yielding a concordant 206Pb/238U age of 982 ± 10 Ma that most likely reflects an inherited zircon, further four spots yielding scattered 206Pb/238U ages that range from 208 ± 2 to 252 ± 2 Ma, and two other spots yielding slightly young and scattered 206Pb/238U ages of 174 ± 2 and 179 ± 2 Ma, both of which likely reflect radiogenic lead loss. A total of 26 coherent analyses yielded 206Pb/238U ages from 187 ± 2 to 205 ± 2 Ma, with a weighted mean age of 195.5 ± 2.2 Ma (MSWD = 0.68, n = 26, 2σ) (Fig. 5d).

Major and trace elemental geochemistry

On an FeOt/MgO vs. SiO2 diagram (Fig. 6a), the metagabbros are typical high-Fe tholeiites, whereas the quartz diorite, two-mica granite, and biotite granite are medium- to low-Fe calc-alkaline rocks. On an A/CNK vs. A/NK diagram (Fig. 6b), the quartz diorites are metaluminous to slightly peraluminous, whereas the two-mica granite and biotite granite are peraluminous and the metagabbros are metaluminous. These samples are also classified as gabbros, diorites/quartz diorites, and granites in both TAS and R1–R2 diagrams (Fig. 6c, d).

Fig. 6
figure 6

FeOt/MgO vs. SiO2 diagram after Miyashiro (1974) (a); A/NK vs. A/CNK diagram after Frost et al. (2001) (b); TAS diagram for classification of the rocks Middlemost (1994) (c). R1 vs. R2 diagram for classification of the rocks De La Roche et al. (1980) (d)

Full size image

Pianma quartz diorites (263 Ma)

The Pianma quartz diorites have SiO2 concentrations of 60.71–64.32 wt% and relatively high concentrations of Na2O (3.03–3.49 wt%) with Na2O/K2O ratios of 1.09–1.88, but contain relatively low concentrations of TiO2 (0.73–0.90 wt%), CaO (3.99–6.05 wt%), Fe2O3T (5.41–6.35 wt%), and MgO (2.10–2.56 wt%), with Mg# values of 47.5–48.4. These samples also contain 15.92–17.50 wt% Al2O3 and have A/CNK (molar Al2O3/CaO + Na2O + K2O) ratios of 0.92–1.04. They contain 107.7–209.3 ppm light rare-earth elements (LREE), 33.47–48.40 ppm heavy REE (HREE), have LREE/HREE ratios of 2.91–4.79, and, when plotted on a chondrite-normalized rare-earth element (REE) variation diagram (Fig. 7a), have insignificant negative Eu anomalies (δEu = 0.70–0.75). These samples also have high (La/Yb) N (11.04–15.88) and (Gd/Yb) N (1.83–1.97) values. On a primitive-mantle-normalized trace element variation diagram (Fig. 7b), the samples are enriched in large ion lithophile elements (LILE; Rb, Th, U, and Pb), have positive Nd and Sm anomalies, and are relatively depleted in Ba (429–610 ppm), Nb, P, and Ti, with Sr = 274–372 ppm, and Y = 21.4–30.8 ppm.

Fig. 7
figure 7

Chondrite-normalized REE patterns and primitive-mantle-normalized trace element spider diagram (a–h) for the rocks in the Tengchong block. The primitive-mantle and chondrite values are from Sun and McDonough (1989). The high Al TTG and island arc volcanic rocks field from Feeley and Hacker (1995) and collision-related granites from Hennig et al. (2009) and Peng et al. (2013)

Full size image

Lianghe two-mica granite (218 Ma)

The samples from the Lianghe area contain high concentrations of SiO2 (73.23–74.35 wt%), have K2O/Na2O ratios of 1.86–2.19, and contain low concentrations of Fe2O3T (1.18–1.33 wt%) and MgO (0.27–0.36 wt%). They have Al2O3 concentrations of 14.04–14.64 wt% and have high A/CNK ratios (1.18–1.23). These samples contain 77.3–118.7 ppm LREE, have LREE/HREE values of 2.78–4.63, (La/Yb)N values of 11.07–24.44, (Gd/Yb)N values of 2.45–4.77, and have chondrite-normalized REE patterns (Fig. 7c) that show sharply negative Eu anomalies (δEu = 0.43–0.59). They have primitive-mantle-normalized trace element variation diagrams (Fig. 7d) that show positive Rb, Th, U, and Pb anomalies along with enrichments in Ta and/or the other high field strength elements (HFSE), and clear depletions in Ba, Nb, Sr, and Eu.

Yingjiang metagabbros (205 Ma)

Samples from the Yingjiang area contain low concentrations of SiO2 (50.17–50.96 wt%) and high concentrations of Na2O (3.89–4.13 wt%), have Na2O/K2O values of 1.50–1.70, and contain 9.78–10.13 wt% Fe2O3T, 6.88–7.12 wt% CaO, and 19.31–19.64 wt% Al2O3, with A/CNK ratios of 0.87–0.90. These samples contain elevated total REE concentrations (268.0–539.8 ppm) with slightly elevated (La/Yb)N values (8.70–25.88) and have chondrite-normalized rare-earth element (REE) patterns (Fig. 7e) with insignificant negative Eu anomalies (δEu = 0.74–0.81). They also have primitive-mantle-normalized trace element patterns (Fig. 7f) that are enriched in the LILE and depleted in Nb, P, Hf, and Ti, and contain elevated concentrations of Sr (869–894 ppm), Ba (837–1100 ppm), and Y (41.1–43.5 ppm).

Lianghe biotite granite (195 Ma)

The samples from the Lianghe area have variable concentrations of SiO2 (69.78–76.32 wt%), elevated concentrations of K2O (4.96–5.99 wt%), K2O/Na2O values of 2.23–2.55, low concentrations of Fe2O3T (1.14–1.36 wt% barring one sample with 2.57 wt%), moderate concentrations of Al2O3 (12.41–12.89 wt% barring one sample with 15.12 wt%), and A/CNK ratios of 1.05–1.10. They contain 74.8–121.9 ppm total REE, have (La/Yb)N values of 9.55–19.12, and have chondrite-normalized REE patterns (Fig. 7g) characterized by variable Eu anomalies (δEu = 0.76–1.13). These samples have primitive-mantle-normalized trace element distribution patterns (Fig. 7h) that are enriched in Rb, Th, U, Pb, and the HFSE, and depleted in Ba and Nb.

Bulk-rock Sr–Nd–Pb isotopes

The bulk-rock Sr–Nd–Pb isotopic data for the granitoids and metagabbros in the study area are given in Tables 2 and 3, with all initial 87Sr/86Sr isotopic ratios (I Sr) and εNd(t) values calculated for the time of magma crystallization.

The Pianma quartz diorite has a low I Sr ratio (0.706467) but a stable εNd(t) value (−5.7) and a T 2DM age of 1.31 Ga. This unit has initial 206Pb/204Pb, 207Pb/204Pb, and 208Pb/204Pb ratios of 18.501, 15.727, and 38.979, respectively.

The Lianghe two-mica granite has an elevated I Sr value (0.729082), an εNd(t) value of –11.4, a T 2DM age of 1.67 Ga, and initial 206Pb/204Pb, 207Pb/204Pb, and 208Pb/204Pb ratios of 18.742, 15.770, and 39.089, respectively.

The Yingjiang metagabbro has relatively low I Sr values (0.707724–0.707805), εNd(t) values of –3.4 to –4.0, T 2DM ages of 1.15–1.10 Ga, and initial 206Pb/204Pb, 207Pb/204Pb, and 208Pb/204Pb ratios of 18.674–18.706, 15.696–15.707, and 38.378–38.880, respectively.

The Lianghe biotite granite has an elevated I Sr value (0.720066), an εNd(t) value of –6.8, a T 2DM age of 1.33 Ga, and initial 206Pb/204Pb, 207Pb/204Pb, and 208Pb/204Pb ratios of 18.641, 15.801, and 39.281, respectively.

Discussion

Petrogenesis of the multi-stage late Permian to Late Triassic magmatism within the Tengchong Block

Late Permian quartz diorites

The quartz diorites in the study area are characterized by low concentrations of SiO2 (60.71–64.32 wt%), are metaluminous to slightly peraluminous, and are Na-rich with Na2O/K2O ratios of 1.09–1.88, similar to those intermediate to granitic rocks generated by the partial melting of basaltic rocks (Rapp and Watson 1995). The low (Na2O + K2O)/(FeOt + MgO + TiO2) and high Na2O + K2O + FeO + MgO + TiO2 values of these samples, combined with their slightly elevated Mg# values (47.5–48.4), are typical of magmas from the basaltic sources (metamafic rocks) (Patiño Douce 1999; Fig. 8a, b). However, the involvement of pre-existing ancient continental crustal material in the formation of these quartz diorites is indicated by the presence of inherited zircons with U–Pb ages of 1582–552 Ma that are almost identical to the age of the metamorphic basement in this area (YNBGMR 1990; Zhong 1998; Song et al. 2010), which infer the significant involvement of metasedimentary materials. Furthermore, the characters of both low Al2O3/TiO2 and high CaO/Na2O ratios and low Rb/Ba and moderate Rb/Sr of these samples indicate that mixture sources between the predominantly basalt- and partly pelite-derived melts (Sylvester 1998; Janoušek et al. 2004; Fig. 7c, d). The low Rb/Sr and Rb/Ba ratios (Fig. 8d) and insignificant δEu anomalies (0.70–0.75) imply the insignificant fractional crystallization of plagioclase (Patiño Douce 1999). The decreasing Sr/Y ratios (10.3–16.3) with increasing Y concentrations (21.4–30.8 ppm) and fractionated REE pattern (LREE/HREE = 2.91–4.79) may indicate amphibole and/or garnet in the residue (Defant and Drummond 1990; Petrford and Atherton 1996). These samples have higher total REE concentrations (141.2–252.9 ppm) than typical lower crust and have REE patterns that are similar to island arc volcanic rocks, suggesting that the magmas were formed by involvement of sub-arc crustal material (Feeley and Hacker 1995). Furthermore, the lithospheric mantle-like isotopic composition of these intrusions (I Sr = 0.706467, εNd(t) composition = −5.7) suggests that the magmas were generated as a result of mixing between crust- and mantle-derived components (Fig. 9a–c), before the mixed magmas assimilated enriched mantle (EMII) material, as evidenced by the initial Pb isotopic composition of these samples (Fig. 10a, b). In addition, the quartz diorites contain late-stage, interstitial hydrous minerals such as hornblende and biotite, euhedral–subhedral zoned plagioclase, and polycrystalline clots of amphibole, all of which provide evidence for low concentrations of water within the magmatic system. The previous experimental and petrological analyses suggest that the generation of water-poor melts (~2 wt% H2O) from basaltic material within sub-arc lower crust that incorporated the assimilation of pelitic sediments requires temperatures of >1000 °C (Castro et al. 2010). In addition, Annel et al. (2006) suggested that the low water contents (~2 wt%) of tonalite magmas could be produced by 50% melting of an amphibole-bearing protolith within lower arc crust. These suggest that the quartz diorite was generated by a mixed source of metamafic rocks with significant metapelitic sedimentary material at relatively high temperatures but under water-undersaturated conditions within lower arc crust.

Fig. 8
figure 8

a Plots of SiO2 vs. Mg number. Marked fields outline experimentally obtained compositions of partial melts by dehydration melting of different source rocks under crustal P–T conditions (0.5–1.5 GPa, 800–1000 °C, based on Patiño Douce (1999) and Wolf and Wyllie (1994); b compositional field of experimental melts derived from melting of felsic pelites (muscovite schists), metagreywackes and amphibolites (Patiño Douce 1999); c CaO/Na2O vs. Al2O3/TiO2 diagram referenced by Sylvester (1998); d Rb/Ba vs. Rb/Sr diagram reference by Janoušek et al. (2004). Symbols as Fig. 6

Full size image
Fig. 9
figure 9

εNd(t) vs. T2DM (a), initial 87Sr/86Sr (b) and SiO2 (c). Permian mafic rocks are from Zhai et al. (2013); Late Triassic mafics are from Zhang et al. (2011); The curve representing the mixing proportion between two components, mantle-derived magma, and middle/upper crustal melts corresponding to the mafic rocks are from Zhai et al. (2013). Plots εNd(t) vs. (206Pb/204Pb)t, illustrating the input of subducted-related sediments into an India MORB source to form the Early Triassic volcanic rocks in the Lancangjiang zone. The compositions of the Indian MORB component are represented by 206Pb/204Pb = 17.31, Nd = 10, Pb = 0.03 ppm and Nd = 0.8 ppm (Saunders et al. 1988) and two distinct end-members of the sediments are selected: one is 206Pb/204Pb = 18.56, εNd(t) = −9, Pb = 9 ppm and Nd = 30 ppm, and the other is 206Pb/204Pb = 19.00, εNd(t) = −9, Pb = 60 ppm and Nd = 30 ppm (Ben Othman et al. 1989). Symbols as Fig. 6

Full size image
Fig. 10
figure 10

Initial 207Pb/204Pb (a), 208Pb/204Pb (b) vs. 206Pb/204Pb diagrams for the Early Triassic volcanic rocks in the Lancangjiang zone. The field of EM1 and EM2 is from Zindle and Hart (1986). The Northern Hemisphere Reference Line (NHRL) is from Hart (1984). The fields of central Indian Ocean MORB and Pacific MORB are from Price et al. (1986) and Ferguson and Klein (1993). The data of the Tethyan Basalts are from Mahoney et al. (1998) and the fields of global pelagic sediments are from Stolz et al. (1988, 1990). The early Triassic arc-like andesites are from Peng et al. (2008) and latest Triassic alkaline basalts and basaltic andesites are from Wang et al. (2010). Symbols as Fig. 6

Full size image

Late Triassic Lianghe two-mica granite

The two-mica granites in the study area are K-rich (5.10–5.7 wt%) with high K2O/Na2O ratios (1.86–2.19) that contrast with the Na-rich, low-K2O, and low K2O/Na2O ratio of the other intrusions in this area, indicating that a model for the formation of these granites via the fractionation of an aluminous-poor magma and the dehydration melting of tholeiitic amphibolite material is not possible. These granites also contain inherited zircons with U–Pb ages of 2393–507 Ma, suggesting the magmas that formed these granites involved ancient crustal material, most likely during the partial melting of crustal source rocks.

The two-mica granites have low CaO/Na2O ratios (0.32–0.41; Fig. 8c), and high Al2O3/TiO2 (73.95–97.60) and A/CNK values (1.18–1.23), Rb/Ba (0.87–1.29), and Rb/Sr (2.55–4.08) ratios (Fig. 8d), suggesting that they formed from magmas generally derived from a metapelite source (Sylvester 1998; Janoušek et al. 2004). These granites have high (Na2O + K2O)/(FeO + MgO + TiO2) values and low Na2O + K2O + FeO + MgO + TiO2 concentrations that imply the magma sources including both felsic pelites and metagreywackes (Patiño Douce 1999; Fig. 8b), a model that is consistent with the distribution of these samples on a Mg# vs. SiO2 diagram (Fig. 8a). In addition, these late Triassic granites have negative εNd(t) (−11.4) and high I Sr (0.729082) values that are typical of magmas derived from crustal sources (Fig. 9), also supported by Pb isotopic compositions that are indicative of derivation from the middle–upper crust. The low Fe2O3 + MgO + TiO2 (1.48–1.73 wt%) and CaO (0.82–1.00 wt%) concentrations within these samples are indicative of anatectic melts that incorporated negligible amounts of peritectic or restitic minerals from the source region (Patiño Douce 1999). In summary, we suggest that these strongly peraluminous two-mica granites formed from magmas generated by the partial melting of felsic pelite material that also contained some metagreywacke or psammitic material.

Latest Triassic Yingjiang metagabbros

The Yingjiang metagabbros contain low concentrations of SiO2 (50.17–50.96 wt%) but high concentrations of TiO2 (1.62–1.69 wt%) and are sodium-rich (Na2O = 3.89–4.13 wt%), suggesting that these units are high-Fe series tholeiites (Fig. 6a). The elevated Nb/Ta ratios (22.2–23.5) and low Zr/Sm ratios (16.3–19.4) of these rocks are indicative of formation from melts generated by the partial melting of rutile-bearing eclogitic material (Foley et al. 2002); however, the low Sr/Y ratios (20.2–21.7) preclude melting of an eclogitic source which should generate the high-Mg and adakitic magma rather than mafic magma (Martin et al. 2005). The rocks have low Mg# values (41.6–42.3) and contain low concentrations of Ni (1.08–3.72 ppm), Cr (2.24–6.94 ppm), and Co (20.7–23.8 ppm), suggesting that they are derived from parental magmas that underwent shallow-level fractional crystallization before emplacement (Wang et al. 2015). These metagabbros have SiO2 concentrations that negatively correlate with FeOT, MgO (Fig. 11), and CaO/Al2O3 ratios as well as having negative correlations between MgO and compatible elements, such as Ni, Co, and Cr, suggesting that they formed from magmas that fractionated significant amounts of olivine and clinopyroxene, although the Eu anomalies (δEu = 0.74–0.81), elevated CaO concentrations, and relatively low Na2O + K2O contents of these metagabbros provide a little evidence for plagioclase fractionation. In addition, the occurrence of negative P and Ti, and positive Sr anomalies in primitive-mantle-normalized trace element diagrams (Fig. 7e, f) probably reflects the nature of the source of the parental magmas for these metagabbros, as suggested by the lack of a correlation between SiO2 concentrations, TiO2 and P2O5 contents, and Rb/Sr ratios. The distribution of the metagabbro samples on La vs. La/Sm and Yb vs. La/Yb diagrams (Fig. 11c, d) is consistent with the model of Wang et al. (2015), who suggested that the composition of the gabbroic magmas in this region was controlled by partial melting and source heterogeneity rather than fractional crystallization. The high Al2O3 concentrations (19.31–19.64 wt%), negative εNd(t) values (−3.41 to −4.04), along with Pb isotopic compositions that are similar to those of the global sediment average (Fig. 10) and the presence of negative Nb–Ta and slightly negative Zr–Hf anomalies (Fig. 7e), suggest the involvement of crustal materials in magma-genesis (Wang et al. 2015). However, the mantle-like initial 87Sr/86Sr values (0.705–0.708; Ratajeski et al. 2005; Fig. 9b, c) may also indicate the admixture of mantle-like melts (Fig. 9a–d). These features are very similar with the metagabbroic rocks in the Nabang (Wang et al. 2015), for which three models have been proposed: mantle wedge newly modified by slab-derived fluid/melt, addition of subducted sediments into the depleted mantle, and input of slab-derived fluid/melt into the enriched lithospheric mantle. Our negative εNd(t) values are contrary to the characters of positive εNd(t) values of first model. The low Ba/Th, U/Th, and Sr/La, and high Th/La, Th/Nb, and La/Sm do not agree with the input of the slab-derived fluid/melts but the sedimentary-derived materials (Plank and Langmuir 1998; Rudnick et al. 2000; Elliott 2003; Plank 2005; Stern et al. 2006). Therefore, the model of a depleted mantle modified by the sedimentary-derived component is more suited (Fig. 9a–d). In addition, the presence of plagioclase within the crystallization sequence recorded by these metagabbros, the irregular contacts between hornblende and plagioclase, and the presence of hornblende clots are all indicative of high-temperature conditions (~1000–1100 °C) and water-undersaturated melts (Castro et al. 2010; Castro 2013). These metagabbros, therefore, have similar petrogenetic histories to the coeval basaltic andesites and porphyritic intrusions in the Sanjiang belt (Wang et al. 2010; Kou et al. 2012).

Fig. 11
figure 11

SiO2 vs. FeOT (a) and MgO (b) diagram, La vs. La/Sm (c), and Yb vs. La/Yb (d) for the Yingjiang gabbroic rocks after Wang et al. (2015). The evolution trends in the c and d suggest that the formation of gabbroic magma is controlled by the partial melting and source heterogeneity rather than crystallization fractionation. Symbols as Fig. 6

Full size image

Early Jurassic biotite granites

The biotite granites in the study area have high K2O/Na2O ratios (2.23–2.55) and K2O + Na2O values (7.18–8.34 wt%), and are compositionally similar to the late Triassic two-mica granites, suggesting that both were derived from fractionated alumina-poor magmas that usually generate metaluminous Na-rich and low-K2O/Na2O acid rocks during closed-system fractionation (Zen 1986; Gaudemer et al. 1988; Springer and Seck 1997; Sylvester 1998; Clemens 2003). This is in contrast to a previous model for these granites that describes their derivation from magmas generated by the dehydration melting of tholeiitic amphibolite, generating a granulite residue at pressures of 0.8–1.2 GPa and a garnet-bearing granulite to eclogite residue at pressures of 1.2–3.2 GPa (Rushmer 1991; Rapp and Watson 1995). The relatively high Al2O3/TiO2 ratios (82.7–99.2), A/CNK values (1.05–1.10), and slightly low CaO/Na2O ratios (0.67–0.79) of these granites are indicative of derivation from a parental magma that was probably generated by the partial melting of a metasedimentary source (Sylvester 1998; Fig. 8c), as supported by the presence of inherited zircons. In addition, these granites have lower Rb/Ba (0.37–0.54) and Rb/Sr (1.06–1.69) ratios than the Late Triassic two-mica granites, suggesting that the former were derived from a sedimentary source that contained more psammite than the latter (Janoušek et al. 2004; Fig. 8d). The high (Na2O + K2O)/(FeO + MgO + TiO2) and low Na2O + K2O + FeO + MgO + TiO2 values indicate that they were derived from a source containing both felsic pelite and metagreywackes material (Patiño Douce 1999; Fig. 7b), as also indicated by the distribution of data in a Mg# vs. SiO2 diagram (Fig. 8a). The fact that these samples have higher (but negative) εNd(t) (−6.8) and lower initial 87Sr/86Sr (0.720066) values than the late Triassic granites in this region also supports a crustal-dominated source for the former (Fig. 9), and Pb isotopic compositions that are indicative of middle to lower crustal affinities (Fig. 10). As discussed above, these evidences suggest that the biotite granites were derived from melting of a metasedimentary source which mainly dominated by metagreywacke and/or psammite.

Late Paleozoic-to-early Mesozoic evolution of the deep crust in the southern Sanjiang orogenic belt

Here, we outline the evolution of the deep crustal sources for the magmatism within the southern Sanjiang orogenic belt during subduction- and collision-related orogenesis by plotting the I Sr and εNd(t) data obtained for the late Paleozoic-to-early Mesozoic (ca. 300–190 Ma) intrusions with data for the surrounding tectonic zones on I Sr and εNd(t) vs. age diagrams (Fig. 12). These diagrams indicate the following points. (1) The I Sr and εNd(t) values of samples from the neighboring Lancanjiang and Jinshajiang–Ailaoshan tectonic zones record three different phases of magmatism that reflect changes in source regions and tectonism, from subduction (>ca. 250 Ma) through collision-related orogenesis (ca. 250 to 218 Ma) to post-orogeny extension (after ca. 218 Ma). (2). The first phase of magmatism generated intrusions with positive εNd(t) (generally >0 barring one sample with a value of around −3) and low I Sr (<0.7075) values, whereas the second phase generated units with generally negative εNd(t) values (generally <0) and high but variable I Sr values (>0.7075), and the third phase produced magmas with similar (but less variable) isotopic compositions to the first phase. (3) Our samples have similar compositions to the trends defined above, although there are some differences in the composition of magmas generated at around 195 Ma.

Fig. 12
figure 12

(87Sr/86Sr)i vs. age (Ma) and εNd(t) vs. age (Ma) diagram. The blue plots and light grey area represent the data from Jinshajiang–Ailaoshan tectonic zone, and these data from Xiao et al. (2004), Jian et al. (2004, 2008), Fan et al. (2010), Kou et al. (2012) and Zi et al. (2012a, b); The purplish red plots and dark grey area represent the data from Lancangjiang tectonic zone, these data from Peng et al. (2008), Hennig et al. (2009), Wang et al. (2010), and Peng et al. (2013); In addition, the magenta designs are plots of our samples

Full size image

These comparisons suggest that the study area records three different types of interaction between mantle-derived and pre-existing crustal sources that reflect the changing evolution of the southern Sangjiang orogenic belt from the late Paleozoic to the early Mesozoic. This indicates that mantle-derived magmas dominate the subduction-related magmatism in this area before 250 Ma, whereas magmas derived from ancient crustal sources were involved in the widespread magmatism associated with collision-related orogenesis in the Jinshajiang–Ailaoshan zone between ca. 254 and 234 Ma. The magmatism in this area after ca. 218 Ma involved magmas derived from deeper crustal sources that also mixed with mantle-derived magmas. Given that the metagabbros and granitoids in the study area have enriched I Sr and εNd(t) compositions that are not present in the two neighboring tectonic zones, we infer that the former involved more crustal-derived material than the magmatism in surrounding areas. In contrast, the Tengchong Block also records a similar evolutionary trend in terms of changes in magma sourcing from mantle- to crustal-derived sources during the Early to Late Triassic as evidenced by variations in a εHf(t) vs. zircon U–Pb age diagram (Fig. 13).

Fig. 13
figure 13

Zircon εHf(t) vs. zircon U–Pb age (Ma) diagram for the Triassic magmatisms in the Tengchong block, the data from Cong et al. (2010) and Huang et al. (2013)

Full size image

Geodynamic implications

The Tengchong Block was located within the southern Sanjiang Tethys belt (Yang et al. 2014), although the precise geodynamic setting and the relationship between this and other blocks (e.g., the Lhasa Block and others) during the evolution of the eastern Paleo-Tethys remain unclear. Clarifying these relationships and the setting of the block require the identification of the relationship between the Permian-Trassic magmatism in the Tengchong–Lhasa Blocks to the west and the Changning–Menglian and Jinshajiang–Ailaoshan and Jinghong belts to the east (Table 4; Fig. 1a, b). Many workers have confirmed the Changning–Menglian suture belt as the Paleo-Tethyan main ocean (Peng et al. 2008, 2013; Hennig et al. 2009; Jian et al. 2009a, b; Wang et al. 2010), to the east, the Jinshajiang–Ailaoshan and Jinghong suture belt was regarded as the remnants of a back-arc basin rather than the Paleo-Tethyan main ocean (Wang et al. 2000; Carter et al. 2001; Nam et al. 2001; Xiao et al. 2004; Jian et al. 2004, 2008, 2009a, b; Fan et al. 2010; Kou et al. 2012; Zi et al. 2012a, b, 2013; Wang et al. 2016). Interestingly, in the west, the Permain-Trassic magmatism in the Indosinian orogenic belt from Lhasa to Tengchong Blocks shares synchronous igneous rocks on both sides of the Paleo-Tethyan main ocean (Figs. 1a, b, 14; Table 4). To balance and interpret the interesting distribution, some typical features of both of them should be considered. In the east, there are the following characteristics (Fig. 1b; Table 4): (1) Voluminous Na-rich and subduction–arc affinity magmatism, including the formation of mafic–ultramafic complexes, blueschists, ophiolites, andesites, and granodiorites between 292 and 248.5 Ma within the Lancangjiang and Jinshajiang tectonic zone (Jian et al. 2009a, b; Wang et al. 2010), whereas the Jinshajiang–Ailaoshan tectonic zone was in a back-arc basin setting between 288 and 249 Ma (Fan et al. 2010; Zi et al. 2012a, b, 2013): (2) typical crustal source-derived rhyolitic and granitic magmatism in both zones between 247 and 214 Ma (Wang et al. 2010; Fan et al. 2010; Zi et al. 2012a, b, 2013; Peng et al. 2013); (3) post-orogenic alkaline basaltic to basaltic and andesitic magmatism in an extensional setting at 214–197 Ma (Wang et al. 2010; Kou et al. 2012). The late Paleozoic-to-early Mesozoic magmatism in the Tengchong Block even to the Indosinian orogenic belt from Tengchong to Lhasa Blocks occurred at a similar time to the magmatism in both the Lancangjiang and Jinshajiang–Ailaoshan tectonic zones, but the Indosinian orogenic belt to the west has a different lithological, sedimentological, and paleobiological characters from the other two tectonic zones to the east at this time.

Table 4 Concise comparison of magmatic evolution from Permian to early Jurassic in the southern of Sanjiang Tethys
Full size table
Fig. 14
figure 14

Distribution characters of the magmatisms from late Paleozoic to early Mesozoic in the Indosinian orogenic belt. These age data from Xu et al. (2007), Wang et al. (2008), Zhu et al. (2009), He et al. (2006), Liu et al. (2006), Li et al. (2003), Zhang et al. (2007), Zheng et al. (2003), Han (2007), Cong et al. (2010), Li et al. (2010), Zou et al. (2011), Huang et al. (2013), Yang et al. (2007, 2009), Li (2009), Li et al. (2011), Cheng et al. (2015) and Meng et al. (2016). MBT main boundary thrust, YLZBS Yarlung Zangbo suture, BNS Bangong-Nujiang suture, SL-CMS Shuanghu–Longmucuo Changning–Menglian suture, JS-ALS Jinshajiang–Ailaoshan suture

Full size image

In the west, the Indosinian orogenic belt lies to the northwest of the Sanjiang belt and extends from the Lhasa to Tengchong blocks (Fig. 14). Triassic granitoids (SHRIMP zircon U–Pb ages of 245–206 Ma) in the Tengchong Block (Cong et al. 2010; Li et al. 2011; Zou et al. 2011; Huang et al. 2013) are thought to represent either the southeastern extension of the Gangdese Indosinian magmatism in the Lhasa Block (Li et al. 2011) or magmatism related to collision/within-plate setting in an extensional environment as a result of the evolution of the Paleo-Tethyan Ocean in the Sanjiang orogenic belt (Cong et al. 2010; Zou et al. 2011). This can be further clarified by considering the following points. (1) Li et al. (2011) suggested that an Indosinian orogenic belt extends from the western Lhasa Block to the southeastern Tengchong Block, as evidenced by magmatism in this area between ca. 262 and 190 Ma (Fig. 14). This magmatism was associated with the formation of subduction-related eclogites, diorites, and granites between 262 and 245 Ma, the majority of which formed from magmas generated by the partial melting of mantle-derived rocks (Xu et al. 2007; Yang et al. 2007, 2009; Wang et al. 2008; Zhu et al. 2009). The later magmatism in this area is dominated by granitoids formed by the partial melting of crustal material between ca. 235 and 190 Ma (He et al. 2006; Liu et al. 2006; Li et al. 2003, 2011; Li 2009; Zheng et al. 2003; Zhang et al. 2007; Han 2007), all of which occurred during collisional orogenesis. (2) The lack of upper Permian to Lower Triassic sediments and the presence of an angular unconformity between the Middle–Upper Triassic and the Upper Carboniferous to Lower Permian sediments in the Tengchong Block (YNBGMR 1990; Yin and Harrison 2000) are consistent with evolving sedimentation within the Lhasa Block but not in the Lancangjiang and Jinshajiang–Ailaoshan tectonic zones. This suggests that both the Tengchong and Lhasa blocks were in a similar tectonic setting at this time. (3) Paleontological and paleomagnetic data for this area, combined with the presence of glacial–marine diamictites and cool/cold-water faunas in the lower Permian within both the Lhasa and Tengchong blocks, are indicative of their Gondwanan affinity at this time (Sengör 1984; Metcalfe 2013). In addition, paleomagnetic data for these areas suggest that they were located at similar paleolatitudes during the Permian (Van Der Voo 1993; Li et al. 2004; Chen et al. 2012). (4) The Tengchong Block was located at an equivalent geotectonic location as the Lhasa Block, with both blocks located between the Bangong–Nujiang and Yarlung–Zangbo–Myitkyina suture belts (Dewey et al. 1988; Yin and Harrison 2000; Kapp et al. 2005; Li et al. 2011; Xu et al. 2012; Huang et al. 2013; Wang et al. 2014, 2015). These features indicate that it is likely that the Indosinian orogenic belt extended from the Lhasa Block to the Tengchong Block during the late Paleozoic to the early Mesozoic.

As mentioned above, we may infer that it is not easy to conclude that the Indosinian belt as a branch of the Changning–Menglian Paleo-Tethyan main ocean looks the same as the Jinshajiang–Ailaoshan and Jinghong suture belt due to their definitely different data of lithological, sedimentological, paleomagnetic, and paleobiological characters. However, the synchronous magmatism on the both sides could be considered as the similar response to the evolution of the Eastern Paleo-Tethys.

The synchronous igneous rocks on the both sides of the Paleo-Tethyan main ocean and some differences have been considered during above discussion; however, the view of tectonic evolution should also be proposed: to the east, there was developed a Late Paleozoic-to-Early Mesozoic back-arc basin along the Ailaoshan and Jinghong–Nan tectonic zone in response to the northward subduction of the Paleo-Tethys main ocean and the final closure of the back-arc basin took place in the uppermost Triassic due to the diachronous amalgamation between the Yangtze and Simao-Indochina Blocks (Fan et al. 2010; Wang et al. 2016). At the same time, the large-scale Lincang–Sukhothai magmatic arc was located between the Changning–Menglian suture zone and the Jinghong–Nan back-arc basin (Wang et al. 2016). As a result, the Paleo-Tethyan pattern is spatially characterized by the Changning–Menglian–Inthanon suture zone, Lincang–Sukhokai arc, and Jinghong–Nan back-arc basin from west to east (Hennig et al. 2009; Fan et al. 2010; Peng et al. 2008, 2013; Wang et al. 2010, 2016 reference therein); to the west, the Lhasa–Tengchong Blocks experienced similar tectonomagmatic history since the Early Paleozoic (Xie et al. 2016): as part of the northern margin of the Australian Gondwana in the Early Paleozoic, with moveward to the Eura-Asian continent from Trassic and collision with Qiangtang–Baoshan during the Cretaceous. Then, this magmatism in the Indosinian orogenic belt from Lhasa to Tengchong Blocks in response to the evolution of the Paleo-Tethys has been divided into two stages: an early stage involving mantle-derived magmas generated by the late Permian-to-Early Triassic subduction of Paleo-Tethyan oceanic crust (ca. 263–245 Ma; Yang et al. 2009; Huang et al. 2013), and a later stage of crustal-derived granitoid magmatism generated by the Late Triassic (after ca. 235 Ma) collision between the Lhasa–Tengchong Blocks and the northern margin of the Australian continent (Li et al. 2011; Zhu et al. 2009; Zou et al. 2011). Therefore, in consideration of mentioned above, both of them share the synchronous igneous rocks, but there are quite different lithological, sedimentological, paleomagnetic, and paleobiological characters, especially, two distinct collision and final amalgamation between the Yangtze and Simao-Indochina Blocks and another between the Lhasa–Tengchong Blocks and the northern margin of the Australian continent. The Jinghong and Ailaoshan suture zone was regarded as the back-arc basin of the Paleo-tethys main ocean during the Permain-Trassic in the east, but the Lhasa–Tengchong Blocks were final amalgamated with Qiangtang–Baoshan Blocks during the Cretaceous in the west (Fig. 1a, b). That is why you could see the synchronous igneous rocks appear to the both sides of the Paleo-Tethyan main ocean.

According to the recent 1:250,000 regional geological map, their strata are characterized by transition from marine deposits, terrestrial deposits, to marine deposits during early Triassic to latest Triassic. These granitoid magmatism also records the evolution of this tectonism in this area from pre-collision to collision-related stages of orogenesis, as evidenced by the trends in an R1–R2 multi-cationic diagram (Batchelor and Bowden 1985; Fig. 15a) and by changes in the composition of the granitoids that formed at this time from island-arc-related, to syn- and post-collisional settings, as inferred from a Rb vs. Y + Nb diagram (Pearce et al. 1984; Fig. 15b). The metagabbros that formed during this magmatism also provide evidence of a within-plate and/or spreading center type tectonic setting, as evidenced by their distribution in Ti vs. Zr and FeOT vs. MgO vs. Al2O3 diagrams (Fig. 15c, d). All of these data indicate that this area underwent a systematic transition from subduction-related to collisional and finally within-plate tectonic settings (Fig. 16), which is consistent with the evolution of magmatism, sedimentology, and tectonic location in the Indosinian orogenic belt over time Zhu et al. (2011, 2013).

Fig. 15
figure 15

a R2 (6Ca + 2 Mg + Al) vs. R1 (4Si − 11(Na + K) − 2(Fe + Ti)) diagram after Batchelor and Bowden (1985); b Rb vs. Nb + Y diagram after Pearce et al. (1984); c plots of Ti vs. Zr diagram after Pearce (1982) and d FeOT–MgO–Al2O3 diagram. Symbols as Fig. 6

Full size image
Fig. 16
figure 16

Tectonic model of the late Paleozoic-to-early Mesozoic magmatisms in the Tengchong block, SW China

Full size image

Conclusions

  1. 1.

    New zircon U–Pb geochronological data provide evidence for four stages of magmatism in the Tengchong Block that span the time period between the late Permian and the Late Triassic. Combining the new petrological and geochemical data obtained during this study with the geology of this region suggests that these metagabbros and granitoids record the transition from the pre-collision/island arc stage of tectonism through collision-related orogenesis until finally reaching a within-plate setting, essentially reflecting the evolution of the eastern Paleo-Tethys within the southwestern segment of the Sanjiang orogenic belt.

  2. 2.

    The widespread magmatism within the Indosinian orogenic belt from the western Lhasa block to southeastern Tengchong block occurred in two distinct phases: an initial period of mantle-derived magmatism associated with Paleo-Tethyan subduction until the Early Triassic, and a later stage of crustal-derived granitoid magmatism associated with the collision between the Lhasa–Tengchong blocks and the northern margin of the Australian continent, which lasted until the Late Triassic.

  3. 3.

    Widespread late Permian-to-Late Triassic magmatism occurred in both the Sanjiang and the Indosinian orogenic belts, reflecting the subduction and collisional processes of the eastern East Paleo-Tethys. The different stages of magmatism in this area are all related to subduction (generating mantle-derived magmas) and later collisional (generating crustal-derived magmas) tectonism that affected the entire region, even though the study area has a somewhat different lithological, sedimentological, and paleobiological history to the other two tectonic zones in this area.