Tectonic Generation of Pseudotachylytes and Volcanic Rocks: Deep-Seated Magma Sources of Crust-Mantle Transition in the Baikal Rift System, Southern Siberia

Abstract

Volcanic rocks from deep-seated sources of the crust-mantle transition (CMT) are geochemically distinguished from those of ocean island basalts (OIB). Here, we report geochemical data on tectonic pseudotachylytes from the Main Sayan Fault (MSF) and volcanic rocks from the Kamar-Stanovoy Zone of Hot Transtension (KSZHT) that represent the deep-seated CMT magmatic sources in the central part of the Baikal Rift System (BRS). The tectonic generation of the KSZHT magmas between 18.1 and 11.7 Ma is compared with present-day seismogenic deformations in the middle-upper crust of the South Baikal Basin and adjacent Tunka Valley, where strong earthquakes are distributed along the Main Sayan and Primorye sutures of the Siberian paleocontinent. From a detail seismic tomography model and geological evidence, we infer that the KSZHT crust–mantle magmatic processes were due to delamination and lamination of a thickened root part of the South Baikal Orogen existed in the Late Cretaceous and Paleogene. In addition, we identify similar deep-seated CMT sources for melts erupted in the past 17 Ma from a delaminated root part of the East Hangay Orogen and adjacent Orkhon-Selenga Saddle in the southwestern BRS. We suggest that both in the central and in the southwestern BRS, the deep-seated CMT magma sources designate cooperative pull-to-axis and convergent effects created in the Japan-Baikal Geodynamic Corridor and in the Indo-Asian interactional region, respectively.

1. Introduction

In recent decades, intraplate volcanic eruptions were a priori interpreted as associated with deep mantle plumes, although alternative causes, such as abrupt lateral changes in stress at structural discontinuities of the lithosphere or the mantle global warming, were also advocated to explain the origin of some Large Igneous Provinces [1,2]. Less commonly hypothesized were tectonically generated melts as a result of decompression in the shallow lithospheric mantle or crust. Such interpretations were suggested for volcanic rocks from Cenozoic continental rifted areas of North America, Europe, Northeast Africa, and East Asia [3,4,5,6,7,8,9,10,11,12,13,14,15,16,17].

Volcanic rocks from the Baikal Rift System (BRS) have mainly OIB-like geochemical signatures, consistent with seismic tomography images that show low-velocity anomalies, possibly related to deep-seated magma sources occurring in the upper mantle and transition layer. These could provide migrated volcanic eruptions through the moving lithosphere [18]. Meanwhile, volcanic rocks from the East Hangay Orogen and adjacent Orkhon-Selenga Saddle of Central Mongolia (southwestern part of the BRS) show both mantle and crustal components geochemically different from OIB [19]. Volcanic rocks of some eruption phases in the Kamar-Stanovoy zone of Hot Transtension (KSZHT) (central part of the BRS) and in the Vitim Plateau (northeastern part of the BRS) also indicate their generation in shallow (garnet-free) sources. Such volcanic rocks of basaltic andesite compositions in the latter area have, however, ¹⁴³Nd/¹⁴⁴Nd isotope ratio similar to that in high-Mg basanites, indicating the secondary high-degrees shallow melting of an OIB-like material previously raised from the lower-middle part of the upper mantle. At the eastern end of the Tunka Valley, few high-Mg basanite flows also erupted, but basaltic ones, derived from shallow sources, distinctly show no OIB-type signatures [20,21]. In this case, it is reasonable to assume melting of a material in a region of the crust-mantle transition (CMT). In order to recognize crustal and shallow mantle components, it is necessary, first of all, to define the local geochemical signatures of the crust.

Tectonic motions of the crust are accompanied with heat release resulted from friction during incremental sliding along a fault zone, in which an adhesive wear mechanism dominates. Generated melts of silicate rocks are quenched in planes of strong fault displacements as pseudotachylytes [22,23,24,25,26]. Tectonic pseudotachylytes were described in mylonites of various metamorphic environments that developed in a regime of dominant plastic deformation of the lower crust [27,28,29,30] and in cataclasites typical of brittle deformations in the middle and upper crust [31,32]. Minerals of igneous rocks and glass, preserved in pseudotachylytes of the upper crust, are affected by metamorphic transformations in deeper conditions [33,34,35].

On the one hand, the similarity of bulk chemical compositions of pseudotachylyte veins and host rocks led to an assumption on melt generation by complete melting of host rocks [22,36,37]. On the other hand, the presence of relict mineral fragments of quartz and feldspar (but not amphibole and mica) in most pseudotachylyte veins from different parts of the world showed predominant melting of mafic minerals [38,39,40,41]. High-friction experiments and numerical simulations of pseudotachylyte generation demonstrated the formation of a gouge of finely crushed material later filled with a melt generated by friction. Such a melt could interact with small debris with further melting of the material and a partial change in the original size of frictionally molten rock fragments. Some pseudotachylytes could have purely crushed origin [31,42,43,44,45,46,47]. In pseudotachylyte veins, a heterogeneous groundmass composition on a scale of microdomains was revealed, usually more mafic relative to a host rock bulk composition [39,48]. A composition of the pseudotachylyte matrix, enriched in mafic components, was explained by predominant crushing of softer mafic minerals during host rock cataclasis, followed by massive melting of a cataclasite [40,49].

A coefficient of volumetric expansion of a water is greater than in a host rock, therefore frictional heating causes an increase in pore pressure in a shear zone and, as a consequence, reduces the effective stresses, preventing melting that can only occur in a case of low permeability of a host rock and effective retention of a fluid under excessive pressure [50,51]. It remains unclear whether pseudotachylytes could be a result of melt penetration from a deeper root part of an active fault into its upper part.

In Southern Siberia, pseudotachylytes have been described so far only in the Primorye Fault that bounds the basement of the Siberian paleocontinent from terranes accreted in the southeast. It was inferred that this type of rocks inherited a chemical composition of a host diorite with relative depletion in SiO₂ and enrichments in FeO_tot, MgO, and CaO [52]. A new pseudotachylyte site is found in the Main Sayan Fault (MSF) that separates the southwestern margin of the Siberian paleocontinent basement from accreted terranes and spatially corresponds to the South Baikal Basin and Tunka Valley of the central Baikal Rift System (BRS) [53,54]. In 2003–2019, seismicity focused within the South Baikal Basin without propagating into the Tunka Valley. Earthquakes distributions and temporal variations of ²³⁴U/²³⁸U activity ratio in groundwater indicated a future main seismic event of the 2020–2021 seismic reactivation in the South Baikal water area after an earthquake in the junction between the Tunka Valley and South Baikal Basin [55]. Respectively, the Bystraya earthquake (K = 14.5, M_w = 5.4) occurred on 21 September 2020 [56] and then the Selenga one followed on 9 December 2021 (K = 13.9).

The aim of this paper is to highlight significance of the deep-seated CMT magmatic sources for the structural development of the South Baikal Basin and Tunka Valley, presently affected by strong seismogenic deformations, through geochemical studies of the MSF pseudotachylytes and the KSZHT volcanic rocks. First, we consider the structural setting of magma generation and present results of the comparative analysis of pseudotachylyte and host mylonite compositions and then report results of the comparative study of the MSF pseudotachylytes and the KSZHT volcanic rocks to designate complementary relationships of their deep-seated sources.

2. Methods and Materials

In each volcano of the KSZHT, a sequence of lava layers was sampled from a base to a top. Data on compositions and ages of volcanic rocks, as well as compositions of their crystalline inclusions, were presented earlier [20,57,58,59,60]. For comparisons with the MSF pseudotachylytes, special attention was delivered to study compositions of lavas from the Kultuk and Meteo volcanoes. For understanding of volcanic activities in the entire hot transtension zone, rocks from the Karerny and Shirokiy volcanoes were additionally sampled. Pseudotachylytes were sampled in the MSF zone with their primary subdivision into rocks of basic, intermediate, and silicic geochemical/chemical compositions using a portable device to measure the magnetic susceptibility (produced by Geodevice Enterprise, Russia). In the KSZHT magma-controlling structure, syn-volcanic direction of extension was determined by the predominant orientation of dikes. Planes of main faults were reconstructed using the method of belts [61].

Major oxides of rock samples were determined by classical methods of analytical chemistry [62] and trace elements were analyzed by inductively-coupled plasma mass spectrometry (ICP–MS) using an Agilent 7500ce quadrupole mass spectrometer. Analytical results on major oxides and trace elements of the standard reference materials, obtained within the time interval of analyzing pseudotachylytes from the MSF and volcanic rocks from the KSZHT, are presented in Table S1 of the Supplementary Materials. Isotopic ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd ratios were measured on a Finnigan MAT 262 mass spectrometer. Standard reference materials JNd-1 (Japan) and NBS SRM-987 (USA) were used. Obtained during the period of measurements were ¹⁴³Nd/¹⁴⁴Nd value of 0.512104 ± 0.000004 (2σ) for JNd-1 and ⁸⁷Sr/⁸⁶Sr value of 0.710233 ± 0.000012 (2σ) for NBS SRM-987. Measurement results of samples were recalculated to the recommended values of the reference materials: 0.512103 and 0.71025, respectively [63]. A secondary alteration effect, causing distortion of ⁸⁷Sr/⁸⁶Sr values, was removed by treatment of samples by a 0.75 M hydrochloric acid solution for 24 h with its further removal through repeated washing by ultrapure water from an Elix-3 Millipore system (France). The methods applied were described previously by Rasskazov et al. [19] and Yasnygina et al. [64]. Analysis of Pb isotopes was performed by multiple-collector inductively coupled plasma mass spectrometry (MC-ICP-MS) using a Neptune Plus’ mass spectrometer and the procedure of sample preparation described by Rasskazov et al. [11]. During the period of measurements, obtained values for NBS-981 were: ²⁰⁶Pb/²⁰⁴Pb 16.9378 ± 0.0011, ²⁰⁷Pb/²⁰⁴Pb 15.4927 ± 0.0005, and ²⁰⁸Pb/²⁰⁴Pb 36.6972 ± 0.0013.

3. Structural Control of the South Baikal Basin and Tunka Valley, Structural Setting of Volcanism

The MSF pseudotachylyte site is located at a collisional border between the Khamardaban terrane and Siberian paleocontinent. The Slyudyanka paleocollision zone is designated by laterally changing degrees of metamorphism from granulite facies near the edge of the Siberian paleocontinent to greenschist ones in the inner part of the Khamardaban terrane. This zonal metamorphism was resulted from the Early Paleozoic compression of the terrane crust against the rigid basement of the paleocontinent [65]. One more convergent phase was expressed by uplift and thrust of the Archean basement of the paleocontinent onto Jurassic sediments of the Irkutsk basin [66]. Later, the orogenic regime was retained in the South Baikal area in the Late Cretaceous and Paleogene [67].

The Slyudyanka paleocollision zone is inherited by the Tunka Valley. The western part of the South Baikal Basin, adjacent to the Tunka Valley, is bounded by the Obruchev fault in the north and by the SW Boundary Fault in the south. Along the faults, a crystalline bed of the basin was subsided for at least 5 km. Miocene-Quaternary bottom sediments, lying on a seismically isotropic basement, are as thick as 4 km. A rapid Pleistocene tectonic subsidence of the basin bed, which was not compensated by accumulation of sediments, was accompanied by 1 km deep water infilling [68,69,70].

The subsided South Baikal bed changes abruptly into the uplifted part of the Tunka Valley, where sedimentary and lava layers of the Miocene-Lower Pliocene Tankhoi Formation occupy high elevations of the relief at altitudes from 750 to 2000 m. The Upper Pliocene-Lower Pleistocene and Middle-Late Pleistocene sediments are also uplifted and partially eroded, although occupying lower relief elevations [57]. The vertical shift of sedimentary strata evidences on contrasting tectonic motions of the South Baikal Basin bed relative to the Tunka Valley. In the eastern end of the latter, volcanoes show an echelon-like combination of the Kamar and Stanovoy volcanic zones and the Bystraya one shifted for 20 km to the west (Figure 1). The KSZHT structure, defined by northeastern strike of dikes and chains of volcanoes in an area of about 600 km², designates extension of the crust directed along the MSF and constrained within the time interval of volcanic activities from 18.1 to 11.7 Ma [20].

Oblique extension (transtension) of the upper crust in the Kultuk village and its vicinity is evidenced from structural observations. Extension is indicated by normal faults (occasionally with lateral displacements) that are the main kinematic units in the BRS basins [71,72]. Transition from extension to compression is shown by thrusts and reverse faults (sometimes with lateral displacements) that displace coarse-clastic Pleistocene and Holocene sediments. Extension of the crust resulted in subsidence of the South Baikal Basin; its compression, on the contrary, led to uplift of the eastern part of the Tunka Valley.

4. Pseudotachylytes and Their Structural Setting

Pseudotachylytes are observed as a series of black and pink stripes and veins in solid-gray mylonites of an outcrop located 1 km west of the Kultuk village (Figure 2a). The stripes are as thick as 10 cm (Figure 2b). Pink bands are cut by black veins (Figure 2c). Unidirectional black banded series are accompanied by wedge-shaped (upwardly tapering) bodies of dark-gray pseudotachylytes. This kind of bodies is up to 2.5 m wide at the base (Figure 2d). Different parts of the outcrop show horizontal fractures filled by pistachio-green epidote-chlorite material. As a result of the most recent deformations, these fractures cut through all rocks.

Pseudotachylytes of different bodies are grouped in terms of magnetic susceptibility. Black bands of basic compositions show high magnetic susceptibility (3.4 × 10^–3–6.17 × 10^–2 SI units). Dark-gray wedge-shaped bodies and veins of intermediate compositions indicate lower magnetic parameters (3.36 × 10^–4–5.71 × 10^–4 SI units). Pink bands of silicic compositions yield interval of 3.17 × 10^–5–4.80 × 10^–5 SI units. Mylonites show a range of values from 9.04 × 10^–5–2.70 × 10^–4 SI units falling mainly between those of intermediate and silicic pseudotachylytes, although some massive mylonites show also elevated values.

Fracturing in a site of mylonites with black bands was studied to define inclinations of major fault planes using the method of belts [61]. Reconstructed are dips of two faults: 200°, angle of 74° and 274°, angle of 56°. The former has an azimuth of 290° that corresponds to the MSF strike; the latter designates the north-south direction of the KSZHT (Figure 3). The MSF plane dips from the Siberian platform to the south-southwest, under the Khamardaban terrane and the KSZHT plane from the South Baikal Basin to the west, under the Tunka Valley. A series of injected bands, veins, and wedge-shaped bodies extends almost parallel to the MSF plane with a relative clockwise turn by six degrees. Pseudotachylyte melts injected into antithetical fissures of the MSF hanging wing.

5. Compositions of Pseudotachylytes and Host Mylonites

For analytical studies of the MSF rocks, samples of injected wedge-shaped bodies, veins, and bands, host mylonites (aluminosilicate and carbonate-aluminosilicate), and an epidote-chlorite rock were selected (

Table 1. Analytical results on representative samples of pseudotachylytes and mylonites from the MSF. MSF—Main Sayan Fault.

Sample	1	2	3	4	5	6	7	8	9
	KL-9	KL-23	KL-22	KL-1	KL-3	KL-11	KL-2	KL-5	KL-12
SiO2, wt%	47.87	49.60	45.47	51.86	61.69	71.06	72.30	28.76	55.71
TiO2	1.76	1.62	0.60	0.69	0.89	0.11	0.08	0.55	0.59
Al2O3	13.51	13.67	17.58	14.73	14.88	13.42	12.86	8.16	13.75
Fe2O3	5.91	5.67	2.72	2.42	1.98	0.30	1.12	1.13	3.06
FeO	7.06	8.74	5.20	7.61	4.94	0.99	2.50	2.53	3.72
MnO	0.21	0.18	0.14	0.19	0.09	0.05	0.05	0.06	0.13
MgO	7.30	5.33	7.50	5.69	2.98	0.44	1.38	2.38	3.93
CaO	9.33	5.40	9.37	6.37	3.98	3.13	1.91	29.18	5.49
Na2O	2.56	2.72	1.85	2.37	3.06	3.84	3.48	2.14	1.94
K2O	0.65	0.50	1.12	0.87	2.67	4.32	1.62	1.10	3.20
P2O5	0.36	0.15	0.12	0.08	0.21	0.04	0.02	0.20	0.18
H2O−	0.08	0.07	0.07	0.20	0.14	0.06	0.14	0.20	0.38
H2O+	3.21	4.21	4.76	4.36	1.85	1.35	1.70	1.93	5.76
CO2	0.37	2.20	3.39	2.40	0.41	1.11	0.61	21.80	1.83
Total	100.18	100.06	99.89	99.84	99.77	100.22	99.77	100.12	99.67
Sc, ppm	45.9	41.2	31.2	43.8	19.6	4.2	15.7	7.4	27.8
V	395	404	184	354	122	15	73	69	203
Cr	239	46	250	196	108	12.2	44	21	107.9
Co	47	42	29	28	20.1	2.9	6.7	9.3	21.6
Ni	116	32	115	52	50	6.3	16	24	39
Cu	52	30	60	59	76	10	33	<10	28
Zn	105	71	69	70	68	16	56	96	60
Ga	17.7	18.1	13.7	14.9	20.3	16.0	10.8	9.2	19.3
Rb	14	14.4	28.2	29	79	92	59	31	159
Sr	529	273	362	279	701	408	154	1812	146
Y	32.8	41.0	12.0	22.8	34.3	12.3	11.4	15.2	17.5
Zr	28	5.0	11.0	44	35	32	13	23	9.4
Nb	4.3	1.6	1.5	1.78	13.9	8.4	3.11	8.69	2.2
Cs	0.5	0.46	0.50	0.6	1.4	0.5	1.5	5.7	6.3
Ba	371	170	315	236	1457	961	624	525	426
La	18.7	2.44	4.66	3.45	74.5	15.3	3.13	22.7	5.9
Ce	44.0	8.25	11.0	8.60	156.2	29.7	7.91	44.9	14.4
Pr	6.00	1.63	1.59	1.41	17.19	3.34	1.19	5.54	1.99
Nd	26.0	9.23	7.27	6.68	64.1	12.4	5.37	21.7	9.43
Sm	6.10	3.55	2.05	2.25	11.07	2.37	1.55	4.19	2.62
Eu	1.91	1.31	0.71	0.80	2.58	0.52	0.30	1.10	0.93
Gd	6.22	4.99	2.06	3.20	8.55	2.04	1.74	3.68	3.09
Tb	1.01	0.96	0.35	0.56	1.11	0.31	0.31	0.50	0.52
Dy	6.01	6.72	2.22	3.75	6.54	2.02	1.94	2.84	3.25
Ho	1.26	1.45	0.46	0.80	1.27	0.43	0.41	0.53	0.61
Er	3.54	4.41	1.32	2.50	3.50	1.26	1.24	1.48	1.85
Tm	0.47	0.68	0.21	0.38	0.52	0.20	0.18	0.21	0.27
Yb	3.01	4.34	1.12	2.43	3.18	1.22	1.31	1.26	1.58
Lu	0.45	0.67	0.18	0.37	0.48	0.19	0.21	0.19	0.21
Hf	0.96	0.56	0.53	0.58	1.22	1.38	0.48	0.70	0.44
Ta	0.28	0.29	0.14	0.13	0.83	0.67	0.25	0.51	0.15
Pb	3.6	N.d.	3.4	N.d.	29.3	23.3	5.3	10.8	3.6
Th	2.67	0.11	0.63	0.55	14.7	2.81	0.49	4.19	0.92
U	0.57	0.04	0.15	0.16	1.55	0.55	0.17	2.88	0.56

Pseudotachylyte compositions: 1–2, basic, moderate-Ti basalt; 3–4, basic, low-Ti basalt; 5, intermediate, dacite; 6, silicic, rhyolite. Mylonite compositions: 7, aluminosilicate; 8, carbonate-aluminosilicate. 9, epidote-chlorite rock. N.d., not determined. All analytical results on pseudotachylytes and mylonites from the MSF are presented in Table S2 of Supplementary Materials.

). In a study the injected magmatic rocks, on the one hand, it is important to follow an experience of the pseudotachylyte classification as derivatives of seismogenic melts in faults. Therefore, the injected rock compositions are compared with those of two types of pseudotachylytes identified in the Sarwar-Junia Fault Zone, India: Pt I, from gneiss, and Pt II, from a contact between gneiss and metadolerite [32]. On the other hand, the injected rock compositions should be considered in terms of the nomenclature of non-crystallized volcanic rocks [75].

Samples of black stripes that correspond to basic compositions in terms of a SiO₂ content are subdivided into moderate- and low-Ti groups (TiO₂ contents 1.43–2.18 wt% and 0.6–0.69 wt%, respectively). Sample KL-17 with the lowest MgO content (4.17 wt%) of the moderate-Ti group has a reduced (but not the lowest) TiO₂ content (1.77 wt%). Specifically referred is sample KL-21, which shows TiO₂ = 1.28 wt% and MgO = 5.18 wt% between contents of these oxides in the moderate- and low-Ti groups. In terms of SiO₂, Al₂O₃, MgO, and TiO₂ abundances, these rocks are comparable to the type II pseudotachylyte from the Sarwar-Junia Fault Zone, but differ from it by elevated Na₂O and reduced FeO_tot and K₂O contents (Figure 4). In a thin section of the basic pseudotachylyte, shown in Figure 2b, a foliated groundmass composed of glaucophane, quartz, and magnetite is observed (Figure 5a). Obviously, this rock, originally solidified from an injected basaltic melt, was affected by metamorphism.

Samples, taken from individual wedge-shaped bodies, show an SiO₂ range of 61.69–66.43 wt% with total alkalis of 5.73–6.47 wt%. These rocks have intermediate (andesite and dacite) compositions. When recalculated to 100% without loss on ignition, all rocks are dacites. The MgO and FeO_tot contents in these rocks (1.85–3.18 wt% and 4.77–6.72 wt%, respectively) are lower than in basic pseudotachylytes. In terms of SiO₂, MgO, CaO, and FeO_tot contents, these rocks are comparable to the type I pseudotachylyte from the Sarwar-Junia Fault Zone, but differ from it by elevated Na₂O and reduced Al₂O₃ and K₂O contents (Figure 4). Thin sections of these rocks show large grains of amphibole, garnet, and microcline, surrounded by a fine-grained strongly deformed quartz-feldspar groundmass (Figure 5b–d).

Pink bands and veins have silicic (rhyolite) compositions with low MgO, FeO_tot, CaO and elevated Na₂O and K₂O contents. The banding style of these rocks indicates structural conditions of their generations similar to those of basic pseudotachylytes. The pink injected rocks are compositionally distinguished from those of the types I and II in the Sarwar-Junia Fault Zone and are referred to the type III.

Mylonites are composed of a mixed aluminosilicate and carbonate rubbed material. Carbonate is present both in bulk mylonites and in veins. The aluminosilicate mylonite group contains 72.3–73.4 wt% SiO₂. As SiO₂ of this group increases, MgO, FeO_tot, CaO, Al₂O₃, and K₂O decrease, while Na₂O increases. The carbonate-aluminosilicate mylonite group shows 18.0–61.8 wt% SiO₂. In this group, increase of SiO₂ content is also accompanied by decrease of MgO, FeO_tot, and CaO abundances, but the K₂O content consistently increases. A sample with the lowest SiO₂ content (KL-7) has a negligible Na₂O concentration (0.2 wt%) with Al₂O₃ content as low as 5.7 wt%. In other samples of the carbonate-aluminosilicate group, Na₂O abundance is also relatively low (1.08 wt% and 2.14 wt%). An epidote-chlorite rock shows TiO₂, FeO_tot, and Mgo contents that lower than values of basic pseudotachylytes but higher than those of mylonites. This could result from mechanical mixing between different materials in the fault zone. However, other oxides do not fit this simple mixing. For instance, K2O abundance of the epidote-chlorite rock, which is as high as 3.2 wt%, exceeds its values in both basic pseudotachylytes and mylonites (Figure 4).

The subdivision of the MSF samples into mylonites and pseudotachylytes in terms of major oxide abundances is consistent with their grouping in terms of trace-element signatures. On a MgO – Th/U diagram (Figure 6), illustrated are a high Th/U ratio (5–10) in intermediate pseudotachylytes, its value about 5 in a silicic pseudotachylyte (sample KL-11), and interval of 2.5–6.0 in basic pseudotachylytes. The Th/U ratio of mylonites and epidote-chlorite rock varies from 1.5 to 3.5.

The moderate- and low-Ti groups of basic pseudotachylytes indicate distinctly different distribution of rare earth elements (REE) (Figure 7a). Moderate-Ti samples KL-19,20 show chondrite-like patterns. Other moderate-Ti samples KL-9,17 and KL-23,24 demonstrate chondrite-normalized spectra enriched and depleted in light REE, respectively. Low-Ti pseudotachylytes have lower REE abundances. Their chondrite-normalized spectra show weak depletion and enrichment in heavy REE. The transitional sample KL-21 falls between the moderate- and low-Ti groups. All basic pseudotachylytes show no Eu anomaly. Pseudotachylytes of intermediate and silicic compositions have light-REE-rich patterns with distinct Eu troughs (Figure 7b).

Aluminosilicate mylonites have unenriched or slightly enriched REE spectra with and without Eu anomaly. A line of the epidote-chlorite rock extends above those of aluminosilicate mylonites (Figure 7b). Carbonate-aluminosilicate mylonites show a weak enrichment in light REE. The analyzed samples yield a range of normalized concentrations at the upper limit of the marbles from the Slyudyanka crystalline complex and slightly shift above them (Figure 7d).

Correlation between CaO/Al₂O₃ and CO₂ (Figure 8a) emphasizes a sufficient role of a carbonate mineral phase in the carbonate-aluminosilicate group and its decreasing in the aluminosilicate one. Common trends of mylonites on CaO/Al₂O₃-CO₂, CaO/CO₂-CO₂, and Ba/Sr-CO₂ diagrams (Figure 8a,b,d) indicate a mechanical mixing of an aluminosilicate material (with low CaO content and high Ba/Sr ratio) and a carbonate one (with high CaO content and low Ba/Sr ratio). A deviation from the mixing trend of a single data point of mylonite KL-10 indicates an admixture of magmatic material processed into mylonites. All pseudotachylytes definitely deviate from the mylonite trend of carbonate admixture and incorporate variable proportions of carbonate phase.

6. Compositions of Basic Pseudotachylytes and Volcanic Rocks

The defined three types of pseudotachylytes in one outcrop and the difference of their compositions from those of host mylonites may indicate their generation in different sources of the MSF root and ascending of melts to an upper common level. These kinds of deep-seated sources for basic pseudotachylytes might have links with those for the KSZHT volcanic rocks.

6.1. Major Oxides

Volcanic rocks from the KSZHT vary from silica saturated basalts to strongly undersaturated basanites. Basaltic lavas erupted on the Kultuk and Meteo volcanoes. The former occurred directly in the MSF zone and was active both at the beginning and at the end of the KSZHT volcanism (about 18 and 13 Ma), the latter was the farthest one from the MSF to the south and was active only at the beginning of the KSZHT volcanism (18.1–17.6 Ma).

Compared with volcanic rocks (

Table 2. Analytical results on representative samples of volcanic rocks from the KSZHT (Kamar-Stanovoy Zone of Hot Transtension).

Volcano	Meteo			Kultuk		Karerny	Shirokiy
Age, Ma	18.1	17.7	17.6	18	13	13	13
Sample	687/2	684/1	683/4	Klt-12	Klt-9	15-04	SL16-1
SiO2, wt%	49.62	47.99	47.78	48.26	48.26	44.77	44.38
TiO2	2.28	2.40	2.67	2.24	1.95	2.73	2.72
Al2O3	14.10	14.90	14.9	14.71	15.15	13.39	13.38
Fe2O3	2.52	2.37	3.2	2.96	3.59	2.23	2.92
FeO	8.85	9.24	8.72	7.46	7.95	9.42	9.29
MnO	0.16	0.15	0.15	0.13	0.14	0.17	0.18
MgO	7.85	7.88	8.1	8.20	8.06	10.44	9.72
CaO	8.13	7.98	7.98	7.62	8.00	9.86	9.53
Na2O	2.94	3.22	3.6	2.94	3.26	2.98	2.95
K2O	1.40	1.48	1.85	1.69	1.09	1.72	1.63
P2O5	0.43	0.50	0.56	0.58	0.34	0.66	0.59
H2O–	0.22	0.82	0.6	0.72	0.51	0.21	0.16
H2O+	1.87	1.40	0.9	2.49	1.67	1.80	2.71
Total	100.37	100.33	101.01	100.00	99.97	100.38	100.16
Sc, ppm	21.4	22.0	18.7	24.3	29.2	14.5	19.4
V	N.d.	N.d.	N.d.	173	197	218	251
Cr	N.d.	N.d.	N.d.	281	201	251	222
Co	N.d.	N.d.	N.d.	44	48	51	56
Ni	N.d.	N.d.	N.d.	127	125	196	172
Cu	51	39	55	41	58	45	59
Zn	126	147	126	119	111	126	133
Ga	N.d.	N.d.	N.d.	21.5	20.5	20.0	21.2
Rb	17	19	24	15	11	35	29
Sr	2015	620	745	1301	447	1129	764
Y	22.2	24.1	23.6	20.5	21.0	22.5	21.0
Zr	198	231	246	188	166	202	224
Nb	27.45	38.54	40.71	32.37	21.26	55.24	45.42
Cs	N.d.	N.d.	0.42	0.28	0.13	0.36	0.20
Ba	402	322	427	281	213	432	388
La	19.8	24.1	31.3	23.7	16.0	35.1	32.0
Ce	45.0	50.3	65.7	51.3	34.9	76.1	69.0
Pr	5.28	6.05	8.20	6.65	4.65	8.92	8.34
Nd	25.95	29.16	34.18	28.36	20.42	36.18	33.92
Sm	5.66	6.48	7.72	6.57	4.99	7.82	7.09
Eu	1.86	2.04	2.29	2.14	1.72	2.47	2.36
Gd	5.46	6.11	6.67	5.77	4.98	7.14	6.49
Tb	0.79	0.87	0.95	0.86	0.75	1.00	0.96
Dy	4.28	4.43	5.20	4.33	4.22	5.24	4.58
Ho	0.75	0.80	0.94	0.81	0.82	0.85	0.80
Er	2.07	2.06	2.23	2.03	2.10	2.17	1.91
Tm	N.d.	N.d.	0.33	0.27	0.29	0.28	0.25
Yb	1.59	1.66	1.78	1.50	1.73	1.49	1.46
Lu	0.25	0.25	0.26	0.20	0.23	0.21	0.20
Hf	4.33	4.93	5.90	4.73	4.36	5.06	5.12
Ta	1.59	2.34	2.81	1.93	1.36	3.89	2.92
Pb	2.5	2.7	4.6	5.4	3.6	2.1	1.9
Th	1.97	2.39	3.41	2.06	1.81	4.54	3.27
U	0.26	0.73	0.95	0.74	0.50	1.63	0.82
87Sr/86Sr	0.704676	0.704362	0.704279	0.704493	0.704523	N.d.	N.d.
±2σ	0.000009	0.000010	0.000013	0.000013	0.000016
143Nd/144Nd	0.512683	0.512705	0.512711	N.d.	N.d.	N.d.	N.d.
±2σ	0.000009	0.000017	0.000010
206Pb/204Pb	17.6474	17.8789	17.8163	17.7734	17.7096	18.1204	18.1591
±2σ	0.0020	0.0013	0.0022	0.0034	0.0020	0.0025	0.0030
207Pb/204Pb	15.4674	15.5014	15.4849	15.4736	15.4787	15.5111	15.5212
±2σ	0.0019	0.0012	0.0020	0.0030	0.0019	0.0022	0.0025
208Pb/204Pb	37.8221	38.0072	37.9359	37.8037	37.9245	38.1196	38.1885
±2σ	0.0046	0.0030	0.0048	0.0075	0.0042	0.0055	0.0065

N.d., not determined. All analytical results on volcanic rocks from the Kultuk, Meteo, Karerny, and Shirokiy volcanoes are presented in Tables S3 and S4 of Supplementary Materials.

), basic pseudotachylytes have a lower MgO content. Since abundance of this oxide in a melt is functionally dependent on temperature [78,79,80], the MSF pseudotachylyte melts had obviously lower temperatures than those of the KSZHT. On a MgO–SiO₂ diagram (Figure 9a), data points of pseudotachylyte samples KL-22 and KL-9 with the highest MgO content are plotted close to volcanic rocks with the lowest content of this oxide. Consequently, temperatures of melts quenched into pseudotachylytes KL-22 and KL-9 were comparable with minimal temperatures of erupted basaltic melts. Low-Ti pseudotachylyte KL-22 shows a lower SiO₂ content than moderate-Ti pseudotachylyte KL-9. Other low- and moderate-Ti pseudotachylytes show intersecting trends of increasing SiO₂ and decreasing MgO abundances

On a FeO_tot-SiO₂ diagram (Figure 9b), volcanic rocks and basic pseudotachylytes yield a prevalent FeO_tot interval of 11–12 wt% under SiO₂ content ranging from 44 to 53 wt%. Compositions of the Meteo Volcano are in a range of the prevalent FeO_tot interval. Initial (18.1 Ma) lavas yield data points compactly enclosed within a data field of those represented the second (17.7 Ma) eruption phase, while points of the third (17.6 Ma) eruption phase demonstrate a shift along the prevalent FeO_tot interval to the KSZHT compositions with a lower SiO₂ content. Data points of the initial (18 Ma) lavas from the Kultuk Volcano show FeO_tot abundance below the prevalent FeO_tot interval, while points of the final (13 Ma) lavas from this volcano overlap those of the 17.6 Ma phase of the Meteo Volcano.

Compared with the KSZHT volcanic rocks, basic pseudotachylytes show broader SiO₂ and FeO_tot variations. The low- and moderate-Ti groups indicate opposite trends of enrichment and depletion in SiO₂ and FeO_tot relative to the prevalent FeO_tot interval. Moderate-Ti sample KL-9 with elevated MgO content shows increase in FeO_tot and decrease in SiO₂ contents relative to the MT18.1 data field. Similar trend was obtained experimentally with increasing pressure for melts from orthopyroxene-enriched lherzolite HK-66 [81]. Data point of low-Ti sample KL-9 with elevated MgO content shifts relative to the MT18.1 data field with a relative decrease in both FeO_tot and SiO₂ contents. The KL18 data field falls on this trend. Moderate-Ti pseudotachylytes also show increasing FeO_tot content as SiO₂ content increases from sample KL-9. At the lowest MgO content, sample KL-17 shows the FeO_tot value in a range of the prevalent compositions.

6.2. La/Yb Ratio

In the Tunka Valley, basaltic and basanitic melts brought to the surface polycrystalline garnet-free cognate inclusions and xenoliths extracted from walls of magma conduits. The crystalline nodules from volcanic rocks incorporated olivine, ortho- and clinopyroxene, spinel, plagioclase, amphibole, and mica [57,60]. A garnet-bearing material, not represented by crystalline inclusions and shown only by erupted melts, are referred to as a deeper region [20,57]. Generally, basic pseudotachylytes and volcanic rocks from the Meteo and Kultuk volcanoes could be generated in the deep-seated CMT sources, while volcanic rocks from the Karerny and Shirokiy volcanoes with lower SiO₂ content than in rocks of the Meteo and Kultuk volcanoes could be produced in sources of the deeper lithospheric mantle.

Respectively, the transition from a garnet-bearing material to a garnet-free one can be considered as the crust–mantle boundary. A similar transition was suggested for Southeastern Australia, where the Moho boundary, determined from seismic data at a depth about 55 km, was compared with the same depth, estimated from a spinel–garnet transition in lherzolite xenoliths [82].

Garnet-free and garnet-bearing deep-seated sources of the KSZHT volcanic rocks were calculated through trace-element modeling of equilibrium melting [20]. It was shown that hypersthene-normative basaltic lavas, erupted on the Meteo Volcano 18.1–17.7 Ma, were generated in a garnet-free source with high degrees of melting (F = 0.15–0.23). Nepheline-normative basalts of this volcano, erupted 17.6 Ma, were modeled as derivatives of a garnet-bearing source with moderate degrees of melting (F ~ 0.08). For 18 Ma and 13 Ma basalts from the Kultuk Volcano, garnet-bearing and garnet-free deep-seated sources were defined, respectively Therefore, the beginning of hot transtension was perceived by activities of separate garnet-free and garnet-bearing sources of the Meteo and Kultuk volcanoes, while the end of hot transtension was spotted by activities of a garnet-free one in the Kultuk Volcano.

From results of modeling, melts of garnet-bearing and garnet-free deep-seated sources were discriminated by a (La/Yb)_N ratio of 9.5 (normalization to the pyrolite [77]). The values >9.5 and <9.5 were defined as indicators of garnet-bearing (mantle) and garnet-free (crustal) sources, respectively. The (La/Yb)_N ratio is used further for a gradation of the KSZHT melts from relatively low degrees of melting in a deeper lithospheric mantle source (F ~ 0.05, (La/Yb)_N = 13–18), through elevated degrees of melting in a sub-crustal level (F ~ 0.08, (La/Yb)_N = 10–12) to higher degrees of melting at the base of the crust (F = 0.15–0.23, (La/Yb)_N = 10–12) (Figure 10a).

Melting of the MSF basic pseudotachylytes also fits into the same gradation (without modeling, (La/Yb)_N = <4). Comparison of basic pseudotachylytes from the MSF with the one from the contact between gneiss and metadolerite in the Sarvar-Juniya fault zone [32] indicates their comparable SiO₂, Al₂O₃, MgO, and CaO, but different FeO_tot, K₂O, and Na₂O contents (Figure 4). The similarity of the former four components may indicate their distribution that obeys intensive (PT) parameters of tectonic generations of silicate melts, while the discrepancy between the latter three components may indicate their additive distribution, i.e., displaying of extensive (concentration) parameters of tectonic generation of silicate melts. Consequently, the FeO_tot, K₂O, and Na₂O contents in tectonically generated melts could reflect complementary redistribution of these components in the lithospheric mantle and crust.

On a (La/Yb)_N-FeO_tot diagram (Figure 10b), data fields of eruptive phases MT18.1, MT17.7, KL18, and KL13 of the Meteo and Kultuk volcanoes are displaced relative to the one of the KSZHT 17.6–11.7 Ma age interval with a decrease in a (La/Yb)_N ratio and with an increase in a FeO_tot content, denoting a composite trend that extends from the KL18 initial lava generation through phases MT18.1 and MT17.7 to the KL13 final lava generation and moderate-Ti basic pseudotachylytes. Consequently, in terms of a FeO_tot content, the garnet-free source of the MSF moderate-Ti pseudotachylytes was complementary to the garnet-bearing source that generated lavas of the Kultuk Volcano. Taking into account local occurrence of these complementary deep-seated sources in the MSF, one can assume that a deep-seated lava source of the Kultuk Volcano was depleted in an iron as a result of its removal from the garnet-bearing sub-crustal lithospheric mantle and its transfer to the crustal level. These additional portions were reflected in generation of moderate-Ti pseudotachylytes. Any complementary links between deep-seated sources of low-Ti pseudotachylytes and those of volcanic rocks are doubtful.

The KSZHT lavas erupted on the area between the Kultuk and Meteo volcanoes show a parallel complementary trend of rocks from garnet-bearing and garnet-free deep-seated sources, shifted to the right and above the one of the Kultuk and Meteo volcanoes. The mantle (garnet-bearing) and crustal (garnet-free) sources of this trend yield higher FeO_tot content than KL13, which corresponds to the prevailing FeO_tot level. The KSZHT volcanic rocks with an elevated (La/Yb)_N ratio and relatively narrow FeO_tot range are also distinguished as low-degrees derivatives of the garnet-bearing mantle source.

On (La/Yb)_N-K₂O and (La/Yb)_N-K₂O/Na₂O diagrams (Figure 10c,d), volcanic rocks of the eruptive phases MT18.1, MT17.7, KL18, and KL13 from the Meteo and Kultuk volcanoes form trends almost coincided with those of rocks from other KSZHT volcanoes. These diagrams demonstrate an increasing role of a K₂O in melts from garnet-bearing deep-seated sources with its complementary decreasing in those from garnet-free ones.

6.3. Ce/Pb Ratio

Sample KL-9 has a Ce/Pb ratio close to that in the undifferentiated mantle. In other samples of the moderate-Ti pseudotachylyte group, a Ce/Pb ratio decreases to values of the continental crust. The low-Ti pseudotachylyte group shows a Ce/Pb ratio in a range of typical lower crustal compositions.

On TiO₂-Ce/Pb and MgO-Ce/Pb diagrams (Figure 11a,b), sample KL-9 fits to lava trends of the first two phases in the Meteo Volcano (MT18.1 and MT17.7). Rocks of the initial eruption phase in the Kultuk Volcano (KL18) also partially fall into this trend. On the MgO-Ce/Pb diagram, rocks from this volcano are mostly shifted to the left with decreasing Ce/Pb ratio. In lavas of other KSZHT volcanoes, a Ce/Pb ratio increases to the ocean basalt value and partially exceeds it.

6.4. Th/Yb and Ta/Yb Ratios

On the Th/Yb versus Ta/Yb diagram, rocks of the continental crust are plotted above the mantle array displayed by OIB and MORB [87]. In a garnet-bearing source, both ratios increase; while in absence of garnet, both ratios decrease. Points of initial lavas from a garnet-free source of the Meteo Volcano (MT18.1) are shifted below the ocean basalt array with a relative increase in a Ta/Yb ratio (Figure 12a). Points of lavas of the next eruption phase from a garnet-free source of the same volcano (MT17.7) are slightly shifted relative the data field of initial lavas toward the lower crust composition. A more advanced shift is shown by final basalts from the garnet-free source of the Kultuk Volcano (KL13), while lavas of the initial eruptions of this volcano (KL18) belong to a trend of garnet-bearing deep-seated sources that extends along the OIB and MORB array. Melts from the KSZHT deep-seated sources are complementary enriched in tantalum (with increasing Ta/Yb and decreasing Th/Ta) relative to OIB melts due to melting both garnet-free lower crust and underlying garnet-bearing mantle. Melts of basic pseudotachylytes from the garnet-free crust, on the contrary, are complementary depleted in tantalum (with a relative decrease in Ta/Yb) (Figure 12b). The KSZHT magmas erupted from sources of modified restite material that had lost the basic pseudotachylyte component.

In the MSF, pseudotachylytes of intermediate and silicic compositions are also associated with sources of the collisional crust. On the Th/Yb–Ta/Yb diagram (Figure 12a), points of these pseudotachylytes correspond to the end of a trend exhibited by crustal xenolith. The trend extends from the ocean basalt array to the middle crust composition. A common nature of intermediate and silicic pseudotachylyte compositions is emphasized by similar REE patterns (Figure 7a). At the same time, the middle crust differs from its lower part by oxidation conditions and occurrence of water-bearing mineral phases. This is characteristic of the source for pseudotachylytes of intermediate composition that lost uranium with notably increased Th/U ratio (Figure 6).

6.5. Sr and Nd Isotope Ratios

On a ¹⁴³Nd/¹⁴⁴Nd–⁸⁷Sr/⁸⁶Sr diagram (Figure 13), a particular role of a garnet-free source for initial basalts from the Meteo Volcano is highlighted. Compared with the entire set of the KSZHT volcanic rocks, lavas of the MT18.1 eruption phase yield a separate group with the lowest ¹⁴³Nd/¹⁴⁴Nd and highest ⁸⁷Sr/⁸⁶Sr isotope ratios. The 17.7 Ma lavas from the Meteo Volcano show notable decrease in ⁸⁷Sr/⁸⁶Sr ratio at ¹⁴³Nd/¹⁴⁴Nd ratio comparable to that of the initial (18.1 Ma) basalts and partly exceeding it. Lavas of the 17.6 Ma phase show limited variations of Nd and Sr isotope ratios within their ranges corresponded to those of the previous eruption phase. At the same time, data points of the 17.6 Ma lavas start the 16.0–11.7 Ma KSZHT array of increasing ¹⁴³Nd/¹⁴⁴Nd and decreasing ⁸⁷Sr/⁸⁶Sr ratios.

6.6. Pb Isotope, Nb/U, and Th/U Ratios

A diagram of ²⁰⁷Pb/²⁰⁴Pb plotted versus ²⁰⁶Pb/²⁰⁴Pb incorporates only data on uranogenic Pb isotopes and yields timing of convective homogenization and modification of a source material. A diagram of ²⁰⁸Pb/²⁰⁴Pb plotted versus ²⁰⁶Pb/²⁰⁴Pb shows relationship between thorogenic ²⁰⁸Pb and uranogenic ²⁰⁶Pb. The former is a final decay product of the ²³²Th series, the latter is a final one of a ²³⁸U series that makes up 99.2743% of a total mass of uranium [90]. Consequently, distribution of points on the ²⁰⁸Pb/²⁰⁴Pb-²⁰⁶Pb/²⁰⁴Pb diagram designates variations of a time-integrated ²³²Th/²³⁸U ratio as a factor of the element concentrations. In ocean basalts, Th and U belong to a group of geochemically similar incompatible elements: Th > U~Nb [88]. Melts derived from mantle and crustal sources reveal markedly different Nb/U ratios: 47 ± 10 in OIB and ~10 in crust-derived volcanic rocks [85,91]. The Th/U value ~4 in mantle rocks increases to 6 in lower crustal ones [89]. This ratio may vary because of different garnet–melt and/or clinopyroxene–melt partitioning of U and Th [92,93]. Anomalously high or low Th/U values can also be related to input or removal of U, migrated under oxidizing conditions due to the formation of water-soluble uranyl $U{O}_2^{2+}$ compounds with hexavalent U.

On diagrams of Pb isotope ratios (Figure 14a,b), the KSZHT volcanic rocks are plotted both as slightly scattered fields and as discrete points. In Figure 14a, a bending array extends from the point of ²⁰⁷Pb/²⁰⁴Pb = 15.40 and ²⁰⁶Pb/²⁰⁴Pb = 17.28 to the one of ²⁰⁷Pb/²⁰⁴Pb = 15.52 and ²⁰⁶Pb/²⁰⁴Pb = 18.2. The lower and middle parts of the array show older source linearity directed at the Inner Asia common mantle component referred to as episodically raised uniform deep mantle material [94]. Points of a lava subgroup from the Kultuk Volcano, designated as KL18′, yield a line with a slope corresponding to an age of about 2.22 Ga. The line relation to the Inner Asia common component assumes the initial source generation at that time. Points, shifted to the right and to the left in the middle and upper parts of the array, indicate its distortion. A lava subgroup KL18″ designates a source processing, approximated with a line with a slope corresponding to an age of about 0.9 Ga. Separate data fields of the MT18.1 and MT17.7 eruptive phases assume a slope of points close to the 2.22 Ga array (or steeper than this), indicating coeval or older differentiation of a source material, while a shifted data field of the 17.6 Ma eruptive phase clearly designates a separate source material.

On a diagram of thorogenic and uranogenic Pb isotope ratios (Figure 14b), four points of the KL18′ subgroup (source age of 2.2 Ga) show well correlated ²³⁸U/²⁰⁴Pb (μ) and ²³²Th/²⁰⁴Pb (ω) ratios. The only point of this subgroup, shifted to the left, indicates decreasing μ without changing ω. The KL18″ subgroup (source age of 0.9 Ga) shows one point fitted to the KL18′ correlation line and three points shifted to the left. On the one hand, increasing μ and ω values result in transition from the DUPAL to none-DUPAL signature, on the other hand, ω values do not vary under decreasing μ. Three phases of the Meteo Volcano show data points generally shifted to the left of the 2.2 Ma KL18′ array of consistently varied μ and ω (but one point) and extend along the KL18″ array of a source region affected by decreasing μ without changing ω.

On Figure 14b, the most radiogenic leads of the KSZHT volcanic rocks correspond to the Inner Asia common component. From this relationship, Pb isotope data can be used for estimation of a Th/U ratio in a source region. Since the common component represents the uniform deep-mantle material [94], the array of Figure 14b characterizes a differentiation event of 2.22 Ga. A ²³²Th has a longer half-life than a ²³⁸U (14.010 × 10⁹ years versus 4.468 × 10⁹ years) [90]. Hence, a ²³⁸U produces more of ²⁰⁶Pb than a ²³²Th generates of ²⁰⁸Pb in the same period of time. The Inner Asia common component yields the ²⁰⁸Pb/²⁰⁶Pb value of 2.10 that can correspond to the initial ²³²Th/²³⁸U ratio as high as 6.58. Respectively, radioactive decays of ²³²Th and ²³⁸U can result in increasing ²³²Th/²³⁸U ratio from 6.58 up to 20.6. This value is the upper limit of the present-day Th/U ratio.

The CMT deep-seated sources of the KSZHT volcanic rocks show significant deviations of Nb/U and Th/U ratios from the OIB values. On diagrams of Figure 15a,b, rare points are notably shifted to the right. Due to increased partial melting, the (La/Yb)_N values in the left clusters of points are slightly decreased. Predominating data points with low Nb/U and Th/U ratios may indicate intensive magma generation processes in U-rich sources that are characteristic of both garnet-bearing and garnet-free layers of the CMT region.

A Nb/U ratio reveals a distinct contrast between melt generations in initial deep-seated sources of the Kultuk and Meteo volcanoes; in the KL18 lavas from a garnet-bearing source of the former, the Nb/U ratio decreases (to 25) and in the MT18.1 lavas from the garnet-free source of the latter, on the contrary, increases (to 110). A similar contrast between the initial KL18 and MT18.1 lavas is displayed in the Th and U distributions; in the KL18 lavas from the garnet-bearing source of the Kultuk Volcano, the Th/U ratio decreases (to 1.8), while in the MT18.1 lavas from the garnet-free source of the Meteo Volcano, on the contrary, it increases (to 8). During the initial KL18 phase of the Kultuk Volcano, only the melts from a garnet-bearing source, enriched in uranium, took place. Displayed at the end of the KSZHT activities was a material of a garnet-free source of the Kultuk Volcano that, in terms of Th/U and Nb/U ratios, was partly similar to the one of the MT17.7 phase of the Meteo Volcano.

The KSZHT volcanic rocks from the deeper lithospheric mantle source show mainly Th/U ratio less than 4 with rare samples exceeded this value (Figure 15b), although data points of this group demonstrate almost symmetrical distribution relative to the OIB average Nb/U ratio (~47) (Figure 13). This difference indicates better sensitivity of a Th/U ratio than a Nb/U one to U-redistribution in a source region due to a Th abundance in volcanic rocks 1–2 orders of magnitude lower than a Nb one.

In contrast to the KSZHT volcanic rocks, the MSF basic pseudotachylytes show inconsistent variations of Nb/U and Th/U ratios. The Nb/U one in depleted rocks ((La/Yb)_N < 1) approaches the ocean basalt value, decreasing to crustal one in rocks with chondritic slightly enriched REE distributions ((La/Yb)_N = 1.2–4.4). The Th/U ratio in the depleted rocks differs significantly from that of ocean basalts, overlapping the entire Th/U range of basic pseudotachylytes from 2.5 (MORB) to 6.0 (lower crust). Taking into account the better sensitivity of the Th/U ratio to uranium transfer, its low values in basic pseudotachylytes may be explained by local fluid enrichment in uranium of a source, similar to the ratio in the CMT deep-seated sources of volcanic rocks.

7. Discussion

7.1. Long-Term CMT Evolution

On the one hand, volcanic rocks of the initial and final eruption phases in the Meteo and Kultuk volcanoes yield common trends with basic pseudotachylytes (Figure 8, Figure 9 and Figure 10) that assumes searching for genetic relationships between the rock sources with a perspective of creating a comprehensive model for the development of seismotectonic processes in the crust and sub-crustal mantle of the Slyudyanka paleocollision zone, inherited by the South Baikal Basin and Tunka Valley. On the other hand, clearly expressed is the iron redistribution with its removal from the CMT garnet-bearing source of the KL18 lavas in the Kultuk Volcano to the crustal source of the MSF basic pseudotachylytes. From complimentarily redistributed Fe, U, and alkalis in deep-seated sources of the KSZHT volcanic rocks, the CMT processes are related to episodes from about 2.22 Ga and 0.9 Ga (Figure 14a).

The basement of the southern edge of the Siberian paleocontinent, accessible for observations, has a protolith age of 3.82 Ga estimated from the conjugated Holmes-Houtermans and Wasserburg model for J-type ore leads with μ as high as 20.1 [10]. The obtained Pb age estimate of the protolith is comparable with the Nd model age limit of 3.9 Ga for the Archean metamorphic rocks from the Irkutsk block of the paleocontinent [96].

The Baikal granulite–gneiss terrane of the Irkutsk block was affected by collisional metamorphic events, dated by the U-Pb zircon method from about 2.8 Ga, 2.65–2.56 Ga, and 1.866–1.853 Ga. The Archean metamorphic episodes were designated in rocks of granulite facies, and the Early Proterozoic one in those of high-temperature amphibolite and partially granulite facies [97]. In addition, metamorphic events were recorded using the Rb-Sr isotope system. In the Onot greenstone belt, migmatites and granites of the basement are constrained between 2.64 and 2.24 Ga [98]. This interval also is shown by an age of 2.45 ± 0.38 Ma obtained for phlogopitization of the Krutaya Guba komatiite-like bodies [99]. These events were concurrent with the initial CMT complementary Fe-U-alkaline processes related to the Inner Asia common component.

The model calculations show that ore leads were extracted from crystalline rocks in time intervals from 1.8–1.5 Ga, 1.2–0.9 Ga, 0.6–0.5 Ga, and about 0.25 Ga [10]. The first episode of the Pb extraction in ore deposits began after the Siberian paleocontinent assembly, accompanied by A-type granite magmatism of the Primorsky complex from about 1.86 Ga [100,101,102,103]. The end of the second episode coincided with the CMT complementary Fe-U-alkaline processes. Two other episodes are also connected with tectonic and magmatic events spotted in the accessible crust.

The remarkable one is displayed along the southern edge of the Siberian paleocontinent in the Late Precambrian alkaline-ultrabasic magmatism with carbonatites [104,105]. Syn-collisional granulite metamorphism of the Slyudyanka Group is constrained by U-Pb zircon ages of 488.5 ± 0.6 Ma and 488.0 ± 0.5 Ma obtained for hypersthene-bearing biotite and two-pyroxene trondhjemites. Post-collisional gabbro-syenite magmatism is dated at 471 ± 2 Ma [106]. The latter age is consistent with the Rb-Sr one of 470 ± 25 Ma, obtained for syenites and related rocks of the Bystraya massif with an initial ⁸⁷Sr/⁸⁶Sr ratio of 0.7047 ± 0.0007 [107]. Numerous small bodies and relatively large gabbro-syenite massifs (Snezhnaya, Bystraya, and Bezymyany), cut through metamorphic rocks of the Slyudyanka paleocollision zone, are compositionally similar to basic pseudotachylytes from the MSF. A TiO₂ content of these rocks varies from 0.5 to 3.0 wt% [108].

In the Mesozoic, sufficient iron redistribution provided formation of phoscorite ore deposits in the Angara Province associated with alkaline carbonatite magmatism in the southern edge of the Siberian paleocontinent [109] that affected also accreted terranes [110,111].

7.2. U in CMT Processes

In terms of Pb isotope ratios, the KSZHT deep-seated magma sources show high Th/U ratio of the Inner Asia common material raised from the deep mantle in the Paleoproterozoic (Section 6.5). This material yielded the time-integrated DUPAL anomaly. Some KSZHT volcanic rocks show the Th/U ratio as high as 10, but basically, it ranges from 2 to 4. From the discrepancy between μ and ω signatures on the Pb isotope diagrams and the Th/U ratio of the KSZHT volcanic rocks, one can assume recent fluid redistribution of U relative to Th with a preferential concentration of U in melts of the CMT deep-seated sources.

Similar to the KSZHT volcanic rocks, strongly depleted and enriched by uranium with variable Th/U and Nb/U ratios with respect to the OIB values are Miocene lavas from the Primorye and Lesser Khingan and also Cretaceous-Paleogene lavas from the Tien Shan (Figure 16). The Th/U ratio as high as 20 is characteristic of basalts and trachybasalts derived from the CMT garnet-free source in the Nakhodka volcanic field of the Primorye, Far East of the Russia. Associated with these rocks basaltic andesites, generated in a crustal source, show also elevated Th/U ratio (up to 10). The lower limit of the Th/U ratio in volcanic rocks from this area is 4 [9].

Uranium depletion with a Th/U ratio as high as 15.5 is characteristic of basanites and associated rocks from a garnet-free CMT source of the South Tien Shan (Tuyun Basin), in contrast to basalts and trachybasalts from a shallow (crustal) source of the Middle-North Tien Shan, which are anomalously enriched by U with a relative decrease in Th/U to 0.17. The deeper (garnet-bearing) mantle sources in the latter area do not show any signs of uranium redistribution, however. The crustal source could be enriched in uranium with its complementary depletion of the shallow garnet-free peridotite source due to the rise of oxidized fluids through the crust–mantle boundary during the Late Paleozoic closure of the Turkestan paleoocean. The Cretaceous–Paleogene disruption of the crust–mantle boundary under the South Tien Shan prevented an ascent of deeper mantle magmas and provided a long-term (122–46 Ma) generation of magmas in a shallow garnet-free peridotite source. A shielding effect of this rupture did not prevent, however, a short-term (61–53 Ma) upraise of magmas from the deeper mantle source accompanied by crustal melting in the Middle-North Tien Shan [112].

In the basalt-andesite assemblage from the Lesser Khingan, three groups of rocks are identified: (1) Moderate-K from a garnet-free source ((La/Yb)_N = 5.3–9.8, Th/U = 3.7–15.9, Nb/U = 41–176); (2) moderate-K from a source of garnet–spinel transition ((La/Yb)_N = 9.7–13.4, Th/U = 7.9–13.2, Nb/U = 84–179); and (3) high-K from a garnet-bearing source ((La/Yb)_N = 15.7–33.5, Th/U = 4.6–5.9, Nb/U = 22–56). Anomalous distribution of U, relative to Nb and Th, in the two former rock groups is associated with processes of uranium redistribution by oxidized fluids in the CMT, in contrast to the OIB-like relationship between these elements in the third group. This is explained by the absence of this effect in reduced conditions of a deeper mantle source. Rocks of the two former groups are found in the vast area of the Lesser Khingan, whereas those of the third one are limited only in the central part of the Udurchukan volcanic field.

From data comparisons of the selected regions (Figure 16), one can see peculiarities of the KSZHT deep-seated magma sources that show a wide range of Th/U ratios covered the entire depth interval from the crust to the deeper lithospheric mantle. In other Asian regions, wide ranges of Th/U ratio are recorded in volcanic rocks from sources of a single level (South and Middle-North Tien Shan) or two levels (Nakhodka and Udurchukan volcanic fields). A deeper lithospheric mantle material of the KSZHT volcanic rocks differs from OIB. In those from other areas, it is either absent (Nakhodka volcanic field, South Tien Shan) or indicates OIB-like signatures (Middle-North Tien Shan, Udurchukan).

7.3. CMT between South Baikal Basin and Tunka Valley

In the southern edge of the Siberian paleocontinent and adjacent Transbaikal areas, northwesterly oriented orogenic structures change laterally to the South Baikal Rift (Figure 17a). From deep seismic sounding, the crust of the Angara-Lena tectonic step as thick as 40 km decreases to 36 km beneath the Selenga saddle that corresponds to an antecedent breakthrough of the Selenga River between the Khamardaban and Ulan-Burgasy ranges of the Sayan-Baikal Upland. Along the southwestern edge of the Siberian paleocontinent, the crust thickness increases to 54 km and reaches 60 km under the Hangay [113].

Deep seismic sounding shows decrease of P-wave velocities in the upper (sub-crustal) part of the mantle under large basins of the BRS without any decrease under inter-basin uplifts [117]. Under the South Baikal Basin, a lens of low P-wave velocities (7.7–7.8 km/s) is constrained at the mantle depths of 36–93 km. Under the central part of the Tunka Valley, a similar low-velocity lens is traced, unlike its eastern part, where a normal velocity of 8.1 km/s is obtained, comparable to that under inter-basin uplifts. A similar relative decrease of velocities is shown in S-wave seismic tomography models [118,119].

A structure of the crust with a change in its thickness from 54 to 36 km is deciphered in the west–east V_S-section compiled along the Tunka Valley and South Baikal Basin with continuation to the southern coast of the lake (Figure 17b). From the central part of the Tunka Valley (KRN station) to the South Baikal, a dipping low-velocity zone is defined, ending at the junction between the South Baikal Basin and Tunka Valley with an abrupt lateral change in velocities under TAL station. The low-velocity crustal heterogeneity on the earth’s surface corresponds to the Paleozoic granites. Under the South Baikal, a low-velocity mantle block is bounded by the V_S isoline 4.4 km/s. Low speeds are traced to a depth of at least 80 km (stations TAL–LST). Below the central Tunka Valley (KRN station), a block of reduced velocities is outlined by the S-wave velocity isoline of 4.4 km/s at a depth of more than 80 km. Between the low-velocity blocks of the central Tunka Valley and South Baikal Basins, the S-velocities increase slightly.

Late Cretaceous and Paleogene sediments and plant communities of the Prebaikal and Selenga-Vitim foredeeps were separated from each other by the South Baikal Orogen that was extended along the southeastern edge of the Siberian paleocontinent [67]. The South Baikal Basin was formed in the Neogene and Quaternary as the orogen inversion structure. Decreasing crustal thickness from the Sayan-Khamardaban mountainous region to the Selenga saddle and South Baikal Basin (Figure 17a), accompanied by abrupt lateral change of velocities at the CMT (Figure 17b), indicates recently-created geodynamic instability of the crust–mantle boundary.

The thickened Cretaceous-Paleogene crust of the South Baikal Orogen was delaminated in the Neogene and Quaternary beneath the South Baikal Basin with underplating of mantle melts. No such kind of processes occurred beneath the eastern end of the Tunka Valley. To register relative motions of deep layers at the junction between these structures, the V_S change is indicative under the TORI and TAL stations, respectively (Figure 17c). The low-velocity block is rejected under the TAL station from the base of the crust to a depth of 55–65 km with its relative thinning up to 40 km.

Between stations TORI and TAL, the structure of the middle and upper parts of the crust is generally consistent with deformations displayed on the earth’s surface. Under TORI station, alternating layers with lower and higher S-wave velocities occur approximately 5 km higher than under TAL one. A vertical displacement of layers below these stations corresponds to a subsidence of the South Baikal Basin bottom relative to the uplifted end of the Tunka Valley.

From sediment distributions in the southern coast of Lake Baikal, initial subsidence of the Central Baikal area is dated back to the Eocene [120,121]. Over time, subsidence accompanied by sedimentation propagated westwards. In the Early Miocene (about 18 Ma), tectonic reactivation reached the eastern end of the Tunka Valley, providing initiation of the KSZHT volcanism, and in the Middle Miocene (about 16–15 Ma), it caused volcanism in the central part of the valley. A sharp structural reorganization at the beginning of the Pliocene led to the development of a deep basin filled by water of Lake Baikal.

Under the South Baikal basin and Tunka Valley, immediately above the region of instability at the base of the crust and in underlying mantle, a velocity of about 4.0 km/s, retained at a depth level of 37–38 km, may indicate a leveling layer of crustal and sub-crustal inhomogeneities by a through flow of low-viscosity material. This leveling layer could be generated simultaneously with differential vertical motions of the earth’s surface and crustal layers beneath the South Baikal Basin and Tunka Valley. The vertical motions of the crust, associated with those in the low-viscosity layer of 37–38 km, were initiated simultaneously with magma generation in the KSZHT crust–mantle sources since ca. 18 Ma.

While assuming the main role of contrast vertical tectonic motions at the junction between the South Baikal Basin and Tunka Valley in reactivation and termination of crust–mantle magmatism in the KSZHT, it is noteworthy that the former originated due to crustal deformations in the Primorye suture. In addition, a fragment of the paleocontinent basement was split off along the Obruchev fault and subsided beneath the Baikal. The northeastern ruptures were stretched along the Main Sayan suture of the Siberian paleocontinent, where the KSZHT volcanism displayed during the paleocontinental block isolation along the Obruchev fault, but became extinct after a structural disjunction between the South Baikal Basin and Tunka Valley with the subsidence of the former and uplift of the latter (Figure 18).

Numerical simulations [122] indicate a dependence of magma intrusion on tectonic forces applied to the lithosphere and/or heat flow recorded on the earth’s surface. Extensional forces are usually concentrated along a rheological boundary, so a lateral propagation of a fracture leads to intrusion of magmatic melts along it. Such control takes place in case of a significant application of tectonic stress to the lithosphere (>10¹¹ N/m). At a relatively low heat flux (<80 mW/m²), the Moho boundary acts as a trap for rising mantle melts, while in case of a high heat flux (>80 mW/m²), a stress in the mantle weakens, so the melts rise unhindered. The reported in our study control of U enrichment by oxidized fluids with progressed magma generation in the CMT layered structure suggests a propagation of low-viscosity layer at the depth of 37–38 km from the South Baikal Basin to the adjacent Tunka Valley that caused the cessation of ascends and eruptions of the KSZHT magmatic melts about 12 Ma. The volcanism extinction signified a temporal transition from effective hot transtension of the crust and underlying mantle to strict control of magmas by the CMT layered structure.

7.4. Structural Control of Strong Earthquakes

The South Baikal Basin is infilled with sedimentary layers as thick as 4 km, lying on a seismically isotropic basement [70]. At depths of 6–14 km, the velocity section shows, however, a layer-like structure that was presumably interpreted by Suvorov and Mishen’kina [123] as Mesozoic and Paleozoic sedimentary units. Such interpretation contradicts a composition of the South Baikal oil, derived from a source with land angiosperm plants that cannot be older than Cretaceous [124]. In addition, the granulite metamorphic zone, recorded along the southern coast of Lake Baikal [73], suggests the MSF extension under the South Baikal from the Kultuk village to the Tankhoi tectonic step. Therefore, the layer-like structure of the middle and upper crust could result from seismogenic sub-horizontal tectonic move apart displacements at the suture zone of the Siberian paleocontinent.

In an area between Lake Baikal and Oka River (Eastern Sayans), the MSF plane dips to the southwest (angle 60–70°) with sub-vertical fragments [53,125,126]. Similar inclination of the MSF plane is reconstructed in the eastern part of the Tunka Valley from fractures in mylonites (Figure 3) and is displayed in a nodal plane of the 2020 Bystraya earthquake mechanism solution. The latter coincides with the dip of a nodal plane obtained for the 2008 Kultuk earthquake, the epicenter of which was in the Baikal area (i.e., in the east-southeastern continuation of the MSF, about 40 km from the Kultuk village) (Figure 19). Consequently, the Main Sayan suture zone played a major role in generating pseudotachylytes, the KSZHT volcanic rocks, and modern seismogenic deformations both in the South Baikal basin and in the adjacent Tunka Valley.

In the present-day structural development of the central BRS, elastic stresses accumulate with strong seismic shocks along the Primorye and Main Sayan suture boundaries of the Siberian paleocontinent. Along strike changes are displayed in both directions of the earth’s surface motions, recorded by GPS geodesy, and earthquake mechanism solutions [56,130]. Deformations of the Primorye suture were implemented in the South Baikal earthquake on February 25, 1999 (Mw = 6.0). In a focal zone, a transverse rupture was reactivated, parallel to the Obruchev fault, with sub-vertical and sub-horizontal positions of the nodal planes. This indicates a control of extension by a sub-horizontal low-viscosity layer. The weaker aftershock on May 31, 2000 (Mw = 5.3) had a normal fault mechanism of slightly tilted nodal planes. In a seismic shock of the Kultuk earthquake on 27 August 2008 (Mw = 6.3) of the subsided (Baikal) part of the Main Sayan suture, a strike-slip mechanism of displacement of the transverse to it (northeastern) rupture, parallel to the Primorye suture, was implemented. In the 2020 Bystraya earthquake, lateral displacements occurred in the uplifted (Tunka) part of the Main Sayan suture. The general northeastern extension of the crust was preserved both in the South Baikal Basin and in the adjacent eastern part of the Tunka Valley.

During seismic reactivations of 2003–2017, epicentral fields of earthquakes in the South Baikal Basin were spatially separated from those in the Tunka Valley by an aseismic area. Similarly distributed, but small, seismic events occurred in the 1995–1999 time interval before the strong South Baikal earthquake on February 25, 1999 (M_w = 6.0) [55]. The 2020 Bystraya earthquake was spatially related to the MSF reactivated in the eastern part of the Tunka Valley and spatially coincided with the Anchuk volcano of the Bystraya volcanic zone in the KSZHT (Figure 1).

Seismogenic crustal deformations in the central part of the Tunka Valley differ from those in its eastern part. The 1995 Zaktui and 2003 Khoytogol earthquakes took place to the east and to the west of the Tunka basin, respectively, with focal mechanism solutions showing opposite propagations of compressional and extensional waves.

The 2008 Kultuk earthquake focus as deep as 16 km corresponded to a low-velocity layer (Figure 17c). The seismic activity reflected its low viscosity. The deeper crustal leveling layer of 37–38 km is aseismic, but the one of 25–27 km with reduced velocities corresponded to the 2020 Bystraya earthquake. This low-viscosity seismically active layer is located under the Tunka Valley deeper than the one of the South Baikal Basin. Seismogenic deformations of this earthquake could be transferred to the underlying leveling layer of 37–38 km and trigger motions along it in the entire central part of the BRS. This sequence explains why the main earthquakes of seismic reactivations in the South Baikal water area are preceded by seismic events of the Kultuk area.

7.5. Geodynamic Control of CMT Magmatism

On a Th/Yb versus Ta/Yb plot, points of andesites and trachyandesites from the East Hangay Orogen correspond to a lower crustal source above the OIB and MORB array and basalts show a trend extended bellow this array [19]. A trend of volcanic rocks from the Orkhon-Selenga saddle is shifted below the OIB and MORB array, similar to the one of the KSZHT volcanic rocks shown in Figure 12. Both in the Tunka Valley and in Central Mongolia, the CMT signatures of volcanic rocks change to the OIB-type ones westward (Figure 20a).

From spatial-temporal distribution of volcanism and velocity structure of the mantle, it was inferred that the lithosphere of the Baikal and Hangay-Belaya mobile zones was affected by primary melting anomalies of the transition layer (West Transbaikal and Gobi, respectively) and by secondary melting anomalies of the upper mantle (Hangay, Vitim, Udokan, and others) [18,21,132]. The Vitim and Udokan melting anomalies and those of the Sea of Japan mobile system designated quasi-periodical responses of the Japan-Baikal geodynamic corridor to 2.5 Ma great eccentricity cycles of the Earth’s rotation. Similar responses recorded melting anomalies of the vast area in Eastern Sayans and Central Mongolia, synchronized with those of the Tibetan-Himalayan Orogen, as constituents of the Indo-Asian convergent region. The Tunka Valley belongs to the western “hot” (Hangay-Belaya) part of the BRS, affected by Late Cenozoic volcanism (Figure 19a), whereas the South Baikal Basin belongs to its central-northeastern “cold” (Baikal) part, which reveals melting anomalies only at the ends of the Vitim-Udokan transtension zone (Figure 20b).

From various earthquake mechanism solutions and a crustal velocity structure, we propose that the crust and sub-crustal mantle of the South Baikal Basin was attenuated due to divergent forces of the Japan-Baikal geodynamic corridor, while the crust and sub-crustal mantle of the adjacent eastern part of the Tunka Valley, on the contrary, was flattened against the rigid edge of the Siberian paleocontinent basement in connection with the development of convergent processes in the Central Asian Orogenic System. From GPS data, it was inferred that blocks of the Siberian platform and Transbaikal diverge in the South Baikal basin at the NW–SE direction (130°) at a rate of 3.4 ± 0.7 mm/year [133]. This direction of motion corresponds to the axis of the Japan-Baikal geodynamic corridor. A maximal extension of the crust results in its thinning under the Middle Baikal up to 36 km (Figure 17a). Therefore, the crust of this part of the basin is about 2 km thinner than the leveling low-crustal layer. Respectively, the low-viscosity leveling layer likely provides a dynamic link between processes of the most extended and thinned crust of the Middle Baikal and the compressed crust of the eastern end of the Tunka Valley.

Lithospheric and sub-lithospheric materials moved along the axial part of the Japan-Baikal geodynamic corridor with a maximum rate, while the rate decreased in its NNE and SSW flanks. This change created pull-to-axis forces, caused alternating volcanic impulses in the Vitim and Udokan volcanic fields of the Vitim-Udokan transtension zone [132]. Pull-to-axis forces resulted also in the almost symmetrical development of the Vitim and Selenga basins. In addition, the crust of the latter basin was affected by the Indo-Asian convergence with suppressed pull-to-axis effect of the Japan-Baikal geodynamic corridor in its upper part, but with magma generation in the deeper crust and underlying mantle.

Hot processes in the eastern end of the Tunka Valley were contemporaneous with those in the Sea of Japan mobile system. The 22–18 Ma volcanic rocks from Northeast Japan showed no systematic variations in a ⁸⁷Sr/⁸⁶Sr at the ocean–continent direction. Along the continental margin (from the north to the south), this isotopic ratio of volcanic rocks ranged from 0.703890–0.704195 in the Okushiri Island to 0.704874–0.705165 in the Honyo area, indicating varying proportions of asthenospheric and lithospheric components [134]. In that time, the Sea of Japan was not opened yet. Since about 18 Ma, deep-seated sources of erupted lavas indicated zoning of a dynamically linked back and frontal sides of the Northeast Japan volcanic arc. The KSZHT volcanic interval corresponded to the one of the initial subduction of the Pacific plate beneath Northeast Japan, accompanied by back-arc spreading of the crust with the Sea of Japan opening in the time interval of 18–13 Ma [135,136]. The volcanic extinction in the KSZHT coincided with the structural reorganization in the Sea of Japan mobile system caused by the collision of the Izu-Bonin and Kuril island arcs with the Northeast Japan one in the time interval of 14–12 Ma [137,138].

We infer that both in the central and in the southwestern parts of the BRS, the CMT magmatism was controlled by pull-to-axis effects in the Japan-Baikal geodynamic corridor enhanced by forces of the Indo-Asian convergence. In addition to the CMT geochemical signatures for deep-seated sources of volcanic rocks in Central Mongolia, it is noteworthy that the seismotomographic profile MOBAL-2003 [118] clearly denotes a common leveling layer of the crust foot at a depth of about 40 km that extends from the Tunka Valley through the Orkhon-Selenga saddle to the East Hangay.

8. Conclusions

To recognize genetic links between pseudotachylytes and volcanic rocks from the central BRS, we present data on major oxides and trace elements of pseudotachylytes and host mylonites from the MSF and emphasize a magmatic origin of the former. We also characterize compositions of the MSF basic pseudotachylytes and the KSZHT volcanic rocks to show their geochemical differences from OIB that assume tectonically-derived melt generation in deep-seated sources of the Slyudyanka paleocollision zone.

Pseudotachylyte melts of basic, intermediate, and silicic compositions, injected into the Main Sayan Fault, were controlled by its junction with the Kamar-Stanovoy Zone of Hot Transtension that was active in the time interval of 18.1–11.7 Ma. From mylonite fracturing, we determine dips of these conjugated magma-controlling fault zones. We infer that the KSZHT volcanism accompanied a rift-related split off a rigid block of the Siberian paleocontinent basement along the Obruchev fault, but it became extinct when the South Baikal Basin and Tunka Valley turned out to be structurally separated from each other due to subsidence of the former and uplift of the latter.

The KSZHT volcanic rocks and the MSF basic pseudotachylytes show complementary variations of Fe, U, and alkalis in their sources occurred in the CMT and crust. From trace-element abundances of the KSZHT volcanic rocks, we distinguish garnet-bearing and garnet-free source materials and link the latter with garnet-free crystalline inclusions in lavas from the Tunka Valley. Meanwhile, lavas of the Kultuk and Meteo volcanoes erupted from both garnet-free and garnet-bearing deep-seated sources; the latter were not represented by a crystalline material. Respectively, we relate these sources to the sub-crustal garnet-bearing mantle of the CMT region. In addition, we show that lavas of the Karerny and Shirokiy volcanoes could generate in the deeper lithospheric mantle.

From uranogenic Pb isotope ratios of volcanic rocks, we define an initial generation of the CMT deep-seated sources of about 2.22 Ga from the Inner Asia common (deep mantle) material and its subsequent long-term evolution with distinct processing about 0.9 Ga. Comparisons between data points, distributed on the ²⁰⁸Pb/²⁰⁴Pb–²⁰⁶Pb/²⁰⁴Pb diagram, and those, showing variations of a Th/U ratio in the KSZHT volcanic rocks, indicate a high ²³²Th/²³⁸U ratio in sources and their recent uranium enrichment, associated with increasing degrees of melting. No mixing of components from garnet-free and garnet-bearing sources emphasizes separate melt generations in different CMT levels during volcanic activities. We speculate that the progressively laminated CMT region could cause stricter control of melting by layers and could prevent the rising of melts, resulting in the extinction of volcanism.

To decipher a character of tectonics, responsible for the KSZHT magmatism, we use a high-resolution S-wave seismic tomography profile and interpret the CMT laminated structure of the central BRS as formed in the Neogene and Quaternary, when rift basins subsided after the Late Cretaceous and Paleogene development of the South Baikal Orogen. We suggest that vertical displacements both in the overlying rifted crust and in the underlying delaminated orogen root were compensated by lateral motions in the low-viscosity layer at the depths of 37–38 km. We explain the generation of the CMT magmatism both in the central and in the southwestern BRS as a result of cooperative pull-to-axis and convergent effects created in the Japan-Baikal Geodynamic Corridor and in the Indo-Asian interactional region, respectively. Present-day crustal deformations result in strong earthquakes distributed along the Main Sayan and Primorye sutures of the Siberian paleocontinent. Earthquake sources correspond to crustal low-velocity layers, located under the South Baikal Basin at shallower depths than under the adjacent Tunka Valley.