Late Ordovician Maﬁc Magmatic Event, Southeast Siberia: Tectonic Implications, LIP Interpretation, and Potential Link with a Mass Extinction

: A geochronological, isotopic, and geochemical study of the Suordakh event of maﬁc magmatic intrusions on the southeast Siberian margin was undertaken. U-Pb baddeleyite dating of a maﬁc sill intruding lower Cambrian rocks, yielded a 458 ± 13 Ma emplacement age. The chemical composition and stratigraphic setting of this dated sill di ﬀ ered from that previously attributed to the Suordakh event, implying that additional intrusions, previously mapped as Devonian, potentially belonged to the Suordakh event. No correlation between L.O.I. and concentration of highly mobile major and trace elements was documented, showing small or no inﬂuence of hydrothermal alteration on the chemical composition of the intrusions. A new tectonic reconstruction located an island arc and active margin relatively close to the study area. However, all samples had chemical compositions close to that of OIB and did not display Ta-Nb and Ti-negative anomalies, nor other features typical for subduction-related magmatism. The major and trace element distribution was most characteristic of within-plate basalts with the mantle source composition being transitional from spinel to garnet lherzolite. Combining four U-Pb baddeleyite dates of maﬁc sills and dykes from southeast Siberia, the age of the Suordakh event was estimated at 454 ± 10 Ma. The area of the Suordakh event was at least 35,000–40,000 km 2 (an estimate including sills previously interpreted as Devonian), and could be increased with additional dating in Southeastern Siberia. Similar ages for within-plate intrusions were reported from South Korea, West Mongolia, South Argentina, North Iran and Northwest Canada, and these ca. 450 Ma ages were collectively close in timing with the latest Ordovician (Hirnantian) mass extinction. More high-precision dating is necessary to fully test a link between the Suordakh event (and the other age-correlative events) and the end-Ordovician mass extinction.


Introduction
The southeastern margin of the Siberian Craton is commonly known as a passive margin with well-recognized early Neoproterozoic (ca. 1000-950 Ma) and Middle-Late Devonian rifting events associated with mafic magmatism, normal faulting, and deposition of clastic rocks and evaporites [1][2][3][4]. In contrast to these events, most of the Ordovician and Silurian are reported as a period of tectonic quiescence, with deposition of thick units of carbonates with numerous fossils and reef-like build-ups [3][4][5][6]. However, U-Pb baddeleyite dating of a small mafic sill cutting an Ediacaran (Vendian) sedimentary succession for the first time, identified a Late Ordovician mafic magmatic event commonly known as the Suordakh [7,8]. Subsequent studies showed that the dykes and sills of the Ordovician age had specific chemical fingerprints [9,10], but the distribution and duration of the magmatic event were still not clear. Moreover, recent tectonic reconstructions [11] located an island arc of similar age nearby, thus leading to uncertainty regarding the tectonic setting of the Suordakh event-extension or subduction related.
During the past decades mafic intrusive and volcanic events of similar age were also documented in South Korea [12], Mongolia [13], Northern Iran [14], South America [15,16], and Canadian Cordillera [17], pointing to the world-wide distribution of the Late Ordovician mafic magmatism. Global environmental changes and mass extinctions triggered by gigantic magma eruptions occupy a majority of the geological literature ( [18][19][20] and references therein). The main objective of this paper is to discuss new data on the age and distribution of Suordakh mafic magmatic event, modified interpretation of the tectonic evolution of the eastern margin of the Siberian Craton and its possible relationship with the Ordovician-Silurian boundary mass extinction. To support our interpretation, we present new geochronological, isotopic, and chemical data.

Geological Setting and Petrography
The eastern margin of the Siberian Craton is marked by a set of thrusts separating it from the Verkhoyansk Fold and Thrust Belt (Verkhoyansk FTB). The outer part of the latter consists of several segments where the southernmost part is commonly known as the South Verkhoyansk segment [21,22]. The study area is located in the western part of this segment, within the Sette-Daban Range ( Figure 1).
The Sette-Daban Range contains the most complete Mesoproterozoic to Carboniferous sedimentary succession within the eastern margin of the Siberian Craton. Three major subdivisions in the sedimentary succession, mark the occurrence of separate sedimentary basins with eastward migrating depocenters, and reflect the major stages of tectonic evolution of the eastern margin of the Siberian Craton [2][3][4].
The lower Mesoproterozoic succession is exposed in the core of the Gornostakh Anticline and in the hanging wall of the northern branch of the Kyllakh Thrust, where it consists of clastic and carbonate rocks formed in variable depositional environments [27]. These units are unconformably overlain by a Mesoproterozoic succession that consists of several unconformity-bounded, kilometer-scale, clastic-carbonate cycles deposited in shallow-marine environments. The uppermost unit of the Mesoproterozoic succession (Uy Group) consists of clastic rocks, part of which show evidence for slope deposition. In the easternmost exposures, mafic volcanic rocks are also documented. The total thickness of Mesoproterozoic rocks is estimated at 12-14 km [1,23,27].
The Ediacaran (Vendian) Lower Devonian succession unconformably overlies the Mesoproterozoic units, locally truncating the pre-Ediacaran folds and thrusts [1,4]. Carbonate rocks with shale units and interbeds predominate. To the west from the Gornostakh anticline, only shallow-marine Ediacaran to Ordovician rocks are exposed, whereas to the east, slope to basinal environments predominated in the Cambrian and Early Ordovician. Middle Ordovician to Lower Devonian rocks consist mainly of massive shallow-marine carbonates. The total thickness of Ediacaran-Lower Devonian rocks is estimated at approximately 11 km [1,23].
The Middle Devonian to Jurassic succession consists of carbonate and clastic rocks with mafic volcanic and evaporite units in its lower (Devonian) part. Overlying upper Lower Carboniferous to Jurassic rocks are represented only by the siliciclastic units collectively known as the Verkhoyansk Complex, and deposited in several Mississippi-size submarine fan-delta systems [2,23]. The total thickness of the Verkhoyansk Complex is estimated at 14-16 km [3,23].
According to common interpretations, deposition occurred on a wide passive margin, although most Mesoproterozoic rocks were likely deposited in an intracratonic basin [2,3,27]. Three rifting events of early Neoproterozoic, early Cambrian, and Middle-Late Devonian ages are recognized. The early Neoproterozoic and Middle-Late Devonian rifting events are associated with major mafic magmatic events, the 1000-950 Ma Sette-Daban event and 370-360 Ma Yakutsk-Vilyui LIP [1][2][3][4][28][29][30][31]. Mafic intrusions associated with the early Cambrian rifting are widely distributed on the northeast margin of the Siberian Craton and are collectively known as the Kharaulakh event [10]. Mafic sills and dykes that locally cut Carboniferous and Permian sedimentary rocks of the Verkhoyansk Complex are associated with the ca. 250 Ma Siberian Traps LIP [2,3].
A wide range of K-Ar whole-rock ages from the mafic sills and dykes could suggest an early Paleozoic mafic magmatic event(s). However, almost all mafic intrusions that cut Mesoproterozoic and lower Paleozoic rocks were mapped as Neoproterozoic or Devonian [1,5,26,29]. More recently, conventional and SIMS U-Pb baddeleyite dating, yielded the ages of 450 ± 24 Ma (sample #190) and 444 ± 22 Ma (sample # X04-20) for mafic sills, and an age of 457 ± 34 Ma (sample # X04-29) for a N-S trending mafic dyke (errors at 2σ level [8][9][10]) and support recognition of an Ordovician mafic magmatic event termed the Suordakh magmatic event [7,10]. These intrusions are located within the Gornostakh Anticline and cut Mesoproterozoic to Cambrian rocks ( Figure 1). Study of chemical and isotopic compositions showed that Ordovician intrusions are characteristically enriched in TiO 2 (3.5-5.8%), K 2 O (0.6-1.8%), P 2 O 5 (0.3-1.0%), have distribution of trace elements close to that of ocean island basalts (OIB) with some enrichment in heavy rare earth elements (HREE) and ε Nd (t), ranging from +3.6 to +6.5 [10]. A negative Sr anomaly was interpreted as a result of hydrothermal alteration. Most intrusions with such chemical and Nd isotopic composition are mafic sills that cut Ediacaran rocks. Mafic rocks intruding Ediacaran and lower Paleozoic rocks with different chemical composition were interpreted to be Middle to Late Devonian in age [5,10,26,29].
New samples of the Suordakh mafic intrusions were mostly collected from mafic sills that cut Ediacaran rocks on the east limb of the Gornostakh Anticline ( Figure 2). Two samples (L15-14 and L15-15, Figure 2b) are from an E-W trending dyke, which cuts the Ediacaran rocks and has a chemical composition close to that of typical Ordovician intrusions. One sample (L15-19) is from the same N-S-trending dyke as X04-29, which is located on the west limb of the Gonostakh Anticline ( Figure 1). Two more samples from mafic sills intruding the Cambrian rocks (X04-14-1, AS18-17) were added after new U-Pb baddeleyite dating of the sample X04-14-1 discussed in this paper. To make discussion of the chemical and isotopic composition of the Ordovician intrusions more complete, we also included information from the previously studied intrusions [10]. Coordinates of all samples included in this study are listed in the Supplementary Materials (Table S1).
Ordovician mafic sills are most widespread on the east limb of the Gornostakh Anticline (Figures 1 and 2). All studied mafic sills are composed of fine to medium-grained dolerite ( Figure 3) and vary in thickness from a few meters to approximately 50 m. Host rock alteration zone is up to 2-3 m thick and is represented by hornfels and marble. The least-altered samples had ophitic and poikilophitic textures and consisted mainly of plagioclase (40-60 modal%), clinopyroxene (25-45%), Fe-Ti oxides (up to 10-15%), and locally, olivine (0-10%) and minor biotite, apatite and opaque minerals. Hydrothermal alteration is typically represented by replacement of plagioclase with albite, and replacement of albite by sericite, chlorite, and carbonate [10]. Hydrothermal quartz, hornblende, epidote, and chalcedony are also locally developed. The abundance of hydrothermally-derived minerals varies from 15 to 70%.  Figure 2 are from sills. Samples with red numbers are from [10], samples with black numbers are from this study.
Ordovician mafic sills are most widespread on the east limb of the Gornostakh Anticline ( Figures  1 and 2). All studied mafic sills are composed of fine to medium-grained dolerite ( Figure 3) and vary in thickness from a few meters to approximately 50 m. Host rock alteration zone is up to 2-3 m thick and is represented by hornfels and marble. The least-altered samples had ophitic and poikilophitic textures and consisted mainly of plagioclase (40-60 modal%), clinopyroxene (25-45%), Fe-Ti oxides (up to 10-15%), and locally, olivine (0-10%) and minor biotite, apatite and opaque minerals. Hydrothermal alteration is typically represented by replacement of plagioclase with albite, and replacement of albite by sericite, chlorite, and carbonate [10]. Hydrothermal quartz, hornblende, epidote, and chalcedony are also locally developed. The abundance of hydrothermally-derived minerals varies from 15 to 70%.

In Situ SIMS U-Pb Baddeleyite Dating
Baddeleyite grains were located in polished thin sections through automated x-ray mapping, using wave dispersive spectroscopy (WDS) on the JEOL electron microprobe at the University of Wyoming (UW). Manual analysis by energy dispersive spectroscopy (EDS) and back-scattered electron imaging (BSE), using the scanning electron microscope (SEM) at UW, determined the sizes and shapes of the baddeleyite grains and whether there were any complex replacement textures with zircon.
Target-rich subsections, approximately 8 × 8 mm, were cut out with a thin kerf diamond saw and mounted in 25.4 mm epoxy disks, along with pre-polished epoxy blocks of reference baddeleyites.
Isotopic U-Pb measurements were performed during 3 sessions using two different secondary ionization mass spectrometers (SIMS) at UCLA, a CAMECA 1270ims and CAMECA1290-HR. The 1270 employs an Oprimary source with a spot size of approximately 20 microns. A field aperture was used to reduce the sampling region to the inner third of the pit. The 1290 has a Hyperion ion source that generates a primary O 2 ion beam that is approximately 1-2 microns in diameter.
The beam was rastered over 3 × 3 micron areas for analysis to lengthen the sputtering time and to average the domains over small regions. U-Pb calibration (rsf, relative sensitivity factor) used UO 2 /U from Phalaborwa and Duluth gabbro baddeleyite standards (2060 Ma, 1099 Ma, respectively) and a linear model was calculated using the in-house ZIPS software. Concordia, and the weighted mean plots and calculations used IsoplotEx, based on the algorithms of Ludwig [35,36]. The results of the U-Pb isotopic study are presented in Supplementary Materials (Table S2).

Whole-Rock Major and Trace Element Measurements
Whole-rock geochemical analyses were carried out at the All Russian Geological Research Institute (VSEGEI) in St. Petersburg, Russia. Major element concentrations were analyzed by X-ray fluorescence (XRF), using an ARL 9800 spectrometer. Concentrations of trace and rare earth element (REE) were determined by Inductively Coupled Plasma-Mass Spectrometry (ICP-MS), using an ELAN 6100 DRC ICP-MS instrument. Measured concentrations were significantly above the detection limits and analytical uncertainties were <3% for major elements and 4-10% for trace elements and REE, except for Ni, which had an uncertainty of about 15%. The precision of geochemical analyses was estimated from the repeated analyses of OU-6 and SBC-1 international standards. Measured concentrations of the major elements, trace elements, and REE with the detection limits are listed in Supplementary Materials (Table S3).

Whole-Rock Sm-Nd and Rb-Sr Isotope Measurements
Nd isotope analyses for samples AS18-17, SM18-20, and SM18-24 were performed on a Finnigan MAT 262 mass spectrometer at the Geological Institute of the Kola Science Center RAS (Apatity, Russia), using the method described in [37]. During the course of this study, the mean value of the 143 Nd/ 144 Nd ratio in the JNdi-1 standard was 0.512075 ± 31 (N = 5). Isotope ratios were normalized to 146 Nd/ 144 Nd = 0.7219, and adjusted to the JNdi-1 standard with 143 Nd/ 144 Nd = 0.512115 [38]. Total blank measurements during the analyses were 0.3 ng in Nd and 0.06 ng in Sm. The analytical error (2σ) of Sm and Nd concentration measurements was ± 0.5%. For the 147 Sm/ 144 Nd and 143 Nd/ 144 Nd ratios, the analytical errors (2σ) were 0.3% and 0.005%, respectively. Sample X04-14-1 was analyzed on a TRITON TI mass spectrometer (at the Institute of Precambrian Geology and Geochronology of the Russian Academy of Sciences, St. Petersburg, Russia) using a method, similar to that of [37] and described in [39]. Total blank measurements were 0.03-0.2 ng for Sm and 0.1-0.5 ng for Nd. Analytical errors (2σ) for Sm and Nd concentrations and the 147 Sm/ 144 Nd ratio were 0.5% and 0.005% for the 143 Nd/ 144 Nd ratio. The isotopic 143 Nd/ 144 Nd ratio was normalized to a 146 Nd/ 144 Nd ratio of 0.7219 and adjusted to 143 Nd/ 144 Nd = 0.512017 for the JNdi-1 standard [38]. ε Nd (t) values for all samples were calculated using the present-day ratios of 143 Nd/ 144 Nd = 0.512638 and 147 Sm/ 144 Nd = 0.1967 in CHUR [40].
Sr isotopes in sample X04-14-1 were analyzed on MI-1201-T mass spectrometer at the Diamond and Precious Metal Geology Institute, Siberian Branch of the Russian Academy of Sciences, Yakutsk, Russia, using the method described in [41]. Total blank measurements of the laboratory were 0.006-0.01 ppm for Rb and 0.03-0.05 ppm for Sr. Analytical errors (2σ) of 87 Rb and 86 Sr concentrations and 87 Sr/ 86 Sr ratio were 0.5%, 0.4%, and 0.03%, respectively. Long-term repeated measurements of the 87 Sr/ 86 Sr ratios for the Carbonate-70 Sr standard averaged 0.7089 ± 0.0002.
Measured Nd and Sr isotopic compositions are listed in Supplementary Materials (Table S4).

U-Pb Geochronology of the X04-14-1 Sample
Forty-two spots were analyzed, but the data from many showed evidence for alteration, even though only pristine grains were targeted in BSE ( Figure 4). The data were filtered by 3 factors to identify those with the most robust geochronological information. The first filter used a Th/U cutoff < 0.3, as Th/U is a good indicator of minor alteration and intergrown younger zircon. Pristine baddeleyite tended to have Th/U < 0.1, and often < 0.01, whereas zircons often had Th/U > 1. The second filter limited the analyses to those that were close to the range of UO 2 /U that were measured for the standards (Supplementary Materials, Table S2) to minimize potential errors introduced by long extrapolations of model U vs. Pb relative sensitivity factors. The U-Pb calibration was critical for the Phanerozoic dates, as it directly affected the calculated date. The third filter was %radiogenic 206 Pb, which was also an indicator of alteration. For this, we used a cut-off of >80%. One spot with high UO 2 /U and another that had <80% radiogenic Pb were kept within the robust set, as these spots still yielded dates within error of the ones that passed all 3 filters.  Ultimately 19 out of 42 spots passed these filters, and 13 of these had overlapping concordant data (Supplementary Materials, Table S2, Figure 5), whereas 6 produced younger 206 Pb/ 238 U dates, which we interpreted to reflect the recent bulk Pb loss or alteration. The best estimate on the intrusive age of the X04-14-1 mafic sill was 458 ± 13 Ma ( Figure 6, 95% confidence, MSWD 1.16), based on the weighted mean 206 Pb/ 238 U date of 13 analyses. Pristine baddeleyite domains such as in (a-d) produced robust data. Alteration to zircon in 12bz@2 (e) and both spots in 25bz (f) yielded poor quality data and failed the data quality filters discussed in Section 4.1.
Ultimately 19 out of 42 spots passed these filters, and 13 of these had overlapping concordant data (Supplementary Materials, Table S2, Figure 5), whereas 6 produced younger 206 Pb/ 238 U dates, which we interpreted to reflect the recent bulk Pb loss or alteration. The best estimate on the intrusive age of the X04-14-1 mafic sill was 458 ± 13 Ma ( Figure 6, 95% confidence, MSWD 1.16), based on the weighted mean 206 Pb/ 238 U date of 13 analyses.
Minerals 2020, 10, x FOR PEER REVIEW 10 of 24 Figure 5. Concordia plot of in-situ SIMS U-Pb analyses from X04-14-1. Blue ellipses overlap Concordia and each other with a Concordia age [36] of 459 ± 18 Ma (small black ellipse). The red ellipses also overlap Concordia but had younger 206 Pb/ 238 U dates and were excluded from the date calculations. These were interpreted to reflect minor, recent, bulk Pb loss and/or alteration. Figure 6. Plot of 206 Pb/ 238 U dates from in-situ SIMS analyses from X04-14-1. Blue bars overlap and produce a weighted mean date of 458 ± 13 Ma (green box), which was interpreted as the best estimate for intrusive age. Red bars were excluded from the calculation, as their dates were younger than this mean. They were interpreted to represent Pb loss and/or alteration.

Geochemistry
A suite of 13 new and 6 already published [10] samples of the Suordakh mafic intrusions were selected for the major and trace element analysis, and the characteristic element ratios are presented in Supplementary Materials, Table S5. Mafic intrusions display variations in SiO2 concentrations from 42.1% to 49.4%, and Na2O + K2O from 2.26% to 5.96%, mainly occupying the field of basalt, with few samples in the picrobasalt and trachybasalt fields in the TAS diagram ( Figure 7a). Approximately  . Plot of 206 Pb/ 238 U dates from in-situ SIMS analyses from X04-14-1. Blue bars overlap and produce a weighted mean date of 458 ± 13 Ma (green box), which was interpreted as the best estimate for intrusive age. Red bars were excluded from the calculation, as their dates were younger than this mean. They were interpreted to represent Pb loss and/or alteration.

Geochemistry
A suite of 13 new and 6 already published [10] samples of the Suordakh mafic intrusions were selected for the major and trace element analysis, and the characteristic element ratios are presented in Supplementary Materials, Table S5. Mafic intrusions display variations in SiO2 concentrations from 42.1% to 49.4%, and Na2O + K2O from 2.26% to 5.96%, mainly occupying the field of basalt, with few samples in the picrobasalt and trachybasalt fields in the TAS diagram (Figure 7a). Approximately Figure 6. Plot of 206 Pb/ 238 U dates from in-situ SIMS analyses from X04-14-1. Blue bars overlap and produce a weighted mean date of 458 ± 13 Ma (green box), which was interpreted as the best estimate for intrusive age. Red bars were excluded from the calculation, as their dates were younger than this mean. They were interpreted to represent Pb loss and/or alteration.

Geochemistry
A suite of 13 new and 6 already published [10] samples of the Suordakh mafic intrusions were selected for the major and trace element analysis, and the characteristic element ratios are presented in Supplementary Materials, Table S5. Mafic intrusions display variations in SiO 2 concentrations from 42.1% to 49.4%, and Na 2 O + K 2 O from 2.26% to 5.96%, mainly occupying the field of basalt, with few samples in the picrobasalt and trachybasalt fields in the TAS diagram (Figure 7a). Approximately half of the samples were related to the alkali series and half to the sub-alkali series. On the Zr/Ti-Nb/Y diagram, the samples occupied the same fields (Figure 7b). Although the samples from this study (cutting Cambrian age rocks) had much higher variation in the TiO 2 and K 2 O concentrations than observed in the previous study of Suordakh intrusions cutting the Mesoproterozoic and Ediacaran rocks of the Gornostakh Anticline [10], in both classification diagrams, most samples occupied quite small fields. The data plot as mid-ocean ridge basalts (MORB) to OIB-type basalts on the V-Ti diagram and were in the tholeiitic series field on the alkali-FeO tot -MgO diagram (     Classification diagrams for the Ordovician Suordakh event mafic intrusions cutting the Gornostakh Anticline [10] and those cutting Cambrian units as well (this study). (a) Total alkalis vs. silica (TAS) [42,43], (b) Zr/TiO2 vs. Nb/Y [44]. Fields of the Devonian Sette-Daban dykes and Vilyui rift dykes (of the Yakutsk-Vilyui LIP) are shown for comparison, after [29,45]. See location of the Vilyui rift in Figure 1 (inset). On the TAS diagram, composition fields of the Ordovician and Devonian intrusions overlap each other, whereas on the Zr/TiO2 vs. Nb/Y diagram, the Ordovician and Devonian Vilyui rift intrusions occupy similar fields but the Sette-Daban Devonian dykes differ from them.  However, highly incompatible, large ion lithophile elements (LILE) and high field strength elements (HFSE), including rare earth elements (REE), are significantly less affected by fractional crystallization than major elements, and are primarily controlled by the degree of partial melting, preserving more information on the primary magma composition [48].
On  Table S5). The variation in (La/Lu)n from 3.82 to 7.53 and in (La/Sm)n from 1.85 to 2.26, averaging at 5.1 and 2.0, respectively, were close to those reported by [10]. They represent variable enrichment of light REE (LREE) over heavy REE (HREE) and a gentle slope of the LREE distribution. The europium anomaly (Eu/Eu*) varied from 0.85 to 1.11, being negative for samples with the lowest REE concentrations (Supplementary Materials, Table S5). The distribution of other trace elements on the multi-element diagram was mainly close to that of OIB, with some variations for Rb, K, Ba, Th; none or very small negative Ta-Nb anomaly; well-expressed negative Sr anomaly; and none or small positive Ti anomaly (Figure 9b).  Table S5). The variation in (La/Lu)n from 3.82 to 7.53 and in (La/Sm)n from 1.85 to 2.26, averaging at 5.1 and 2.0, respectively, were close to those reported by [10]. They represent variable enrichment of light REE (LREE) over heavy REE (HREE) and a gentle slope of the LREE distribution. The europium anomaly (Eu/Eu*) varied from 0.85 to 1.11, being negative for samples with the lowest REE concentrations (Supplementary Materials, Table S5). The distribution of other trace elements on the multi-element diagram was mainly close to that of OIB, with some variations for Rb, K, Ba, Th; none or very small negative Ta-Nb anomaly; well-expressed negative Sr anomaly; and none or small positive Ti anomaly (Figure 9b).  [29,45]. The Ordovician data of this study and Devonian Vilyui rift intrusions occupy the overlapping fields, whereas the Sette-Daban Devonian dykes differ from them. (b) Primitive mantle normalized diagrams for the Suordakh event mafic intrusions. Normalized using the Chondrite and Primitive Mantle composition of [49], respectively. N-MORB, E-MORB, and OIB compositions after [49].
The Suordakh (Ordovician) and Yakutsk-Vilyui (Devonian) events mafic intrusions occupy overlapping fields on the diagrams, based on the major elements, and cannot be separated from each other (Figures 7a and 8a,b). However, on the classification diagram based on the HFSE elements, the Devonian and Ordovician intrusions of the Sette-Daban area were clearly separated, although the Ordovician intrusions of the Sette-Daban area and Devonian dykes of the Vilyui rift occupied overlapping fields (Figure 7b). Similarly, on the REE diagrams, the Ordovician intrusions of the Sette-Daban area and the Devonian dykes of the Vilyui rift occupied overlapping fields, whereas the  [29,45]. The Ordovician data of this study and Devonian Vilyui rift intrusions occupy the overlapping fields, whereas the Sette-Daban Devonian dykes differ from them. (b) Primitive mantle normalized diagrams for the Suordakh event mafic intrusions. Normalized using the Chondrite and Primitive Mantle composition of [49], respectively. N-MORB, E-MORB, and OIB compositions after [49].
The Suordakh (Ordovician) and Yakutsk-Vilyui (Devonian) events mafic intrusions occupy overlapping fields on the diagrams, based on the major elements, and cannot be separated from each other (Figures 7a and 8a,b). However, on the classification diagram based on the HFSE elements, the Devonian and Ordovician intrusions of the Sette-Daban area were clearly separated, although the Ordovician intrusions of the Sette-Daban area and Devonian dykes of the Vilyui rift occupied overlapping fields (Figure 7b). Similarly, on the REE diagrams, the Ordovician intrusions of the Sette-Daban area and the Devonian dykes of the Vilyui rift occupied overlapping fields, whereas the Devonian dykes of the Sette-Daban area differed from them through flat distribution of REE (Figure 9a).
Hydrothermal alteration is optically discernible and mainly affects plagioclase, which is widely replaced by albite, sericite, chlorite, and carbonate ( Figure 3). However, in both classification diagrams (Figure 7), the studied samples occupied fields with rocks of similar composition, even though one diagram, TAS, was based on highly mobile alkalis, and the other diagram was based on immobile Ti, Zr, Y, and Nb. This relationship made significant modification of alkali concentrations unlikely. Moreover, all but 3 samples had an L.O.I. less than 2.5% and there was no correlation between L.O.I. and the concentration of mobile elements like Ca, Na, K, Rb, Sr, Ba, and U (Figure 10), showing no or very small influence of hydrothermal alteration on the chemical compositions of the intrusions.

Nd and Sr Isotopic Study
Four new Nd and one Sr isotope analyses of mafic intrusions were combined with four previously published Nd and Sr isotope analyses [10], to extend the available database. Some isotopic characteristics are presented in the Supplementary Materials (Table S5). εNd(t) values ranged from 1.4 to 6.9 indicating a mixture of depleted and enriched sources (Figure 11a). However, the Nd isotope compositions did not show any correlation with the SiO2 contents (Figure 11b), implying that direct crustal contamination, as suggested by the range of Nd isotope values, was very unlikely. Sr isotope

Nd and Sr Isotopic Study
Four new Nd and one Sr isotope analyses of mafic intrusions were combined with four previously published Nd and Sr isotope analyses [10], to extend the available database. Some isotopic characteristics are presented in the Supplementary Materials (Table S5). ε Nd (t) values ranged from 1.4 to 6.9 indicating a mixture of depleted and enriched sources (Figure 11a). However, the Nd isotope compositions did not show any correlation with the SiO 2 contents (Figure 11b), implying that direct crustal contamination, as suggested by the range of Nd isotope values, was very unlikely. Sr isotope compositions also showed significant variation with the 87 Sr/ 86 Sr initial ratios, ranging from 0.7066 to 0.7081, pointing to the enrichment in radiogenic 87 Sr (Figure 11c) (Supplementary Materials, Table S5) that might be a result of variable processes like crustal contamination, hydrothermal alteration, or sedimentary recycling [50,51]. However, the absence of a correlation between Nd isotopic compositions and SiO 2 contents did not support crustal contamination (Figure 11b). Previously, radiogenic 87 Sr enrichment in the studied intrusions was interpreted as a result of hydrothermal alteration processes [10,29], but relationship between highly mobile elements and L.O.I. likely pointed to no significant influence of hydrothermal alteration on their chemical compositions ( Figure 10). However, Jourdan et al. [51] reported alteration processes that affected Sr isotope compositions, but did not affect trace element compositions of mafic dykes. Available data on Sr isotopic compositions of the Suordakh mafic intrusions were not enough to identify reasons for radiogenic 87 Sr enrichment and we do not further address this question in the following discussion.  Table S5) that might be a result of variable processes like crustal contamination, hydrothermal alteration, or sedimentary recycling [50,51]. However, the absence of a correlation between Nd isotopic compositions and SiO2 contents did not support crustal contamination (Figure 11b). Previously, radiogenic 87 Sr enrichment in the studied intrusions was interpreted as a result of hydrothermal alteration processes [10,29], but relationship between highly mobile elements and L.O.I. likely pointed to no significant influence of hydrothermal alteration on their chemical compositions ( Figure  10). However, Jourdan et al. [51] reported alteration processes that affected Sr isotope compositions, but did not affect trace element compositions of mafic dykes. Available data on Sr isotopic compositions of the Suordakh mafic intrusions were not enough to identify reasons for radiogenic 87 Sr enrichment and we do not further address this question in the following discussion. εNd(t) vs. SiO2. No correlation between εNd(t) and SiO2 is shown. Primitive mantle composition (PM) is from [49], (c) ( 87 Sr/ 86 Sr)i versus εNd(t) diagram for the Suordakh event mafic intrusions (from [52,53]). EMI and EMII-enriched mantle sources; HIMU-high µ ( 238 U/ 204 Pb) mantle source.

Age and Distribution of the Suordakh Event Intrusions and Its Correlates and Relationship with the Late Ordovician Mass Extinction
The weighted mean of four U-Pb baddeleyite dates of the Suordakh mafic intrusions was 454 ± 10 Ma, which we interpreted as an estimate of the timing of the Suordakh mafic magmatic event ( Figure 12). Although the distribution of the Suordakh event mafic sills was previously thought to be located predominantly within the Mesoproterozoic and Ediacaran (Vendian) rock units [10], the new dating of the X04-14-1 sample showed that the Suordakh mafic sills also cut the Cambrian rocks, significantly increasing the areal extent of this magmatic event ( Figure 13). Furthermore, prior to this study, a typical feature of these ca. 450 Ma intrusions that cut the Mesoproterozoic and Ediacaran rocks [10] was high Ti (TiO2 > 3.5%) and P (P2O5 > 0.3%) concentrations. However, the X04-14-1 sample had lower concentrations of both Ti (TiO2 = 2.37%) and P (P2O5 = 0.23%) (Supplementary Materials, Table S3). These results showed that a significant number of mafic sills cutting the Cambrian rocks or showing much lower concentrations of Ti and P might also be Ordovician. Distribution of the Ordovician dykes is not clear. U-Pb baddeleyite dating and the specific chemical composition showed that there were at least few Ordovician N-S-trending dykes (samples X04-29, L15-19) and E-Wtrending dykes (samples L15-14, L15-15). However, all N-S-trending dykes that cut the Cambrian and Ordovician rocks to the east from the Gornostakh Anticline were supposed to be Devonian [26,29,32,34]. Trace elements composition of the Devonian dykes seemed to be different from that of the Suordakh event intrusions (Figures 7b and 9a). However, the trace elements composition of the Devonian dykes was reported from only 5 intrusions and was very different from that of the Figure 11. Nd isotope characteristics of the Suordakh event mafic intrusions cutting the Gornostakh Anticline [10] and those cutting the Cambrian units as well (this study) (a) ε Nd (t) vs. (Th/Nb) PM . (b) ε Nd (t) vs. SiO 2 . No correlation between ε Nd (t) and SiO 2 is shown. Primitive mantle composition ( PM ) is from [49], (c) ( 87 Sr/ 86 Sr)i versus ε Nd (t) diagram for the Suordakh event mafic intrusions (from [52,53]). EMI and EMII-enriched mantle sources; HIMU-high µ ( 238 U/ 204 Pb) mantle source.

Age and Distribution of the Suordakh Event Intrusions and Its Correlates and Relationship with the Late Ordovician Mass Extinction
The weighted mean of four U-Pb baddeleyite dates of the Suordakh mafic intrusions was 454 ± 10 Ma, which we interpreted as an estimate of the timing of the Suordakh mafic magmatic event ( Figure 12). Although the distribution of the Suordakh event mafic sills was previously thought to be located predominantly within the Mesoproterozoic and Ediacaran (Vendian) rock units [10], the new dating of the X04-14-1 sample showed that the Suordakh mafic sills also cut the Cambrian rocks, significantly increasing the areal extent of this magmatic event ( Figure 13). Furthermore, prior to this study, a typical feature of these ca. 450 Ma intrusions that cut the Mesoproterozoic and Ediacaran rocks [10] was high Ti (TiO 2 > 3.5%) and P (P 2 O 5 > 0.3%) concentrations. However, the X04-14-1 sample had lower concentrations of both Ti (TiO 2 = 2.37%) and P (P 2 O 5 = 0.23%) (Supplementary Materials, Table S3). These results showed that a significant number of mafic sills cutting the Cambrian rocks or showing much lower concentrations of Ti and P might also be Ordovician.
Distribution of the Ordovician dykes is not clear. U-Pb baddeleyite dating and the specific chemical composition showed that there were at least few Ordovician N-S-trending dykes (samples X04-29, L15-19) and E-W-trending dykes (samples L15-14, L15-15). However, all N-S-trending dykes that cut the Cambrian and Ordovician rocks to the east from the Gornostakh Anticline were supposed to be Devonian [26,29,32,34]. Trace elements composition of the Devonian dykes seemed to be different from that of the Suordakh event intrusions (Figures 7b and 9a). However, the trace elements composition of the Devonian dykes was reported from only 5 intrusions and was very different from that of the synchronous dykes of the Vilyui rift [29]. More isotopic dating is necessary to identify the areal extent of the Ordovician dykes.
Minerals 2020, 10, x FOR PEER REVIEW 15 of 24 synchronous dykes of the Vilyui rift [29]. More isotopic dating is necessary to identify the areal extent of the Ordovician dykes.   synchronous dykes of the Vilyui rift [29]. More isotopic dating is necessary to identify the areal extent of the Ordovician dykes.  . Areal distribution of the previously dated Suordakh event mafic intrusions [10], and their possible extent within Cambrian rocks with reference to the new U-Pb dating of the X04-14-1 mafic sill. Geology from [22,26,54]; simplified and modified. Figure 13. Areal distribution of the previously dated Suordakh event mafic intrusions [10], and their possible extent within Cambrian rocks with reference to the new U-Pb dating of the X04-14-1 mafic sill. Geology from [22,26,54]; simplified and modified.
The areal distribution of the Suordakh event mafic intrusions in the Sette-Daban area was less than that of the 100,000 km 2 minimum for classification as an LIP [7,55], including mafic sills that cut the Cambrian rocks in the Suordakh event intrusions that estimated its areal extent as 35,000-40,000 km 2 ( Figure 13). Based on apatite and zircon fission track data, the total erosion in this area likely reached 6-8 km [25] showing that a significant amount of mafic rocks, including, probably, mafic volcanic rocks, was removed after formation of the Verkhoyansk FTB in Mesozoic, thereby expanding the likely volume and distribution of the Suordakh rocks. The areal extent of the Suordakh event intrusions might also include regions to the east, as the Ediacaran and lower Paleozoic rocks are hidden below the younger Paleozoic rocks units ( Figure 13).
Mafic intrusions and volcanic rocks correlating with the Suordakh event are found in various parts on the Earth's surface. These might have formed near SE Siberia, and together with the Suordakh event, could define a single LIP. However, given reconstruction uncertainties, these coeval intraplate units could also represent separate but coeval events [7]. In South Korea, the bimodal volcanic rocks of the Ongnyeobong Formation formed in a within-plate tectonic setting and yielded SHRIMP U-Pb zircon ages at 445.0 ± 3.7 Ma and 452.5 ± 3.2 Ma [12,56]. CA-ID-TIMS U-Pb zircon dating of rift-related bimodal volcanic rocks of the Teel Formation in Western Mongolia, yielded a 446.03 ± 0.21 Ma age of their eruption [13]. Within error, both the South Korean and Mongolian volcanic rocks overlapped in age with mafic intrusion from Southeast Siberia. Stratigraphic relationships suggest a Late Ordovician age for continental rift-related basalts in Andean Argentine Precordillera [15] and Late Ordovician-Silurian age for continental rift-related basalts in Northern Iran [14]. Middle Ordovician-Silurian within-plate basalts were documented in Selwyn Basin, Northwest Canada [17,31]. Although the amount of magmatic rocks documented for Southeast Siberia, South Korea, Mongolia, and other areas, could be smaller than those in a typical LIP, all together they point to ca. 450-440 Ma timing, for a globally important pulse of intraplate magmatism.
A wide distribution of either a single mafic magmatic event or multiple coeval events might trigger climate and environmental changes, glaciation, and mass extinction, which was well-documented in the latest Ordovician age (e.g., [19,[57][58][59][60] and references therein). Correlation between mass extinctions and flood-basalt eruptions and related LIPs starting from the Late Permian is now well documented, e.g., the ages of flood-basalt episodes match in age with 7 out of 11 major extinction events at 66, 94, ca. 120, 183, 201, 252, and 260 Ma [7,19,61,62]. For Paleozoic time, the relationship between most mass extinctions and LIPs is recognized as well [7,63]. However, the major Ordovician-Silurian boundary extinction does not contain an obvious LIP correlate. The age of the Late Ordovician glaciation is estimated as the Hirnantian Stage, whereas mass extinction occurred mainly at the Katian/Hirnantian stages boundary (445.2 ± 1.4 Ma [64]) and, probably also, at the Ordovician/Silurian boundary (443.8 ± 1.5 Ma [64]). Within error, the age of the Suordakh mafic magmatic event overlap with both boundaries, making it a good candidate to trigger the environmental changes responsible for the Late Ordovician extinction [7,63]. However, more high-precision dating of magmatic rocks is necessary to test this possible link more robustly.

Tectonic Setting and Magma Source of the Suordakh Event Intrusions
Available paleomagnetic data as well as correlation of sedimentary successions and tectonic events show that during Mesoproterozoic-early Paleozoic-the Okhotsk and Omolon massifs with Archean to Proterozoic basement and adjoining passive margin of the Omulevka terrane composed of carbonate and clastic rocks, were likely parts of the same large continental block, along with the Siberian craton forming a Siberian paleocontinent ( Figure 14) [11,[65][66][67]. In the eastern direction, the passive margin Omulevka terrane was in contact with the Rassokha terrane, where the Ordovician rocks consisted of clastic rocks interbedded with tuffs of basalt to andesite composition, and rare basalt flows marking the approximate location of the Amandykan island arc [11,65,68,69]. Recent SHRIMP U-Pb zircon dating of a small granite massif with an island arc-type chemical composition, yielded a 440 ± 2 Ma age of intrusion [70]. A granite intrusion that was close in age (445.7 ± 1.5 Ma, SHRIMP U-Pb zircon dating) was documented on the northwest margin of the Okhotsk massif ( Figure 14) [71]. Combining all available data, we present a paleotectonic reconstruction where our study area with Late Ordovician mafic intrusions was surrounded by island arcs or active margins, raising questions on the tectonic setting of the southeast margin of the Siberian craton in Ordovician time ( Figure 14). Traditionally, it was interpreted as a passive margin (e.g., [2][3][4]72]), but in view of the island arc magmatic activity, it could be interpreted as a wide, back-arc basin. If so, the mafic intrusions of the Suordakh event (454 ± 10 Ma) and granite (445-440 Ma) might be related to the same subduction-related magmatism. However, it is also known that LIP events can have associated voluminous silicic magmatism and there can be ambiguity in distinguishing silicic magmatism associated with arcs from that associated with LIPs [7,73]. Chemical composition of the Suordakh event intrusions can test this arc-link hypothesis.
None of the studied samples show either Ta-Nb or Ti-negative anomalies, which are typical for the subduction-zone basalts. The Nb/La ratio was high, varying from 0.70 to 1.30, and averaging at 0.98 (Supplementary Materials, Table S5), which is typical for within-plate basalts not affected by contamination [52]. The distribution of HFSE and REE elements also supported the within-plate origin of the Suordakh event mafic intrusions. On the Th/Yb-Nb/Yb diagram, they plot within the mantle MORB-OIB array along with Devonian Sette-Daban and Vilyui rift dykes of the Yakutsk-Vilyui LIP [29,45], close to OIB and far from the field of arc basalts. They are also distinct from ca. 1000-950 Ma Neoproterozoic mafic intrusions from the same area, ca. 490-480 Ma post-orogenic intrusions from the southwest margin of the Siberian craton and 183 Ma flood basalts of the Karoo LIP (of South Africa), all of which show some influence of inherited subduction material or crustal contamination (Figure 15a) [30,51,74]. On a Zr/Nb-Nb/Th diagram (Figure 15b), the studied intrusions occupy the field of oceanic plateau basalts close to the margin of OIB and out of the field of island arc basalts. According to [75], the relationship between Zr, Nb, and Y calculated as ΔNb  Table S5). Modeling of the mantle source using REE, indicated melting in transition from spinel to garnet lherzolite mantle material, with 3-10% of partial melting ( Figure 16). Combining all available data, we present a paleotectonic reconstruction where our study area with Late Ordovician mafic intrusions was surrounded by island arcs or active margins, raising questions on the tectonic setting of the southeast margin of the Siberian craton in Ordovician time ( Figure 14). Traditionally, it was interpreted as a passive margin (e.g., [2][3][4]72]), but in view of the island arc magmatic activity, it could be interpreted as a wide, back-arc basin. If so, the mafic intrusions of the Suordakh event (454 ± 10 Ma) and granite (445-440 Ma) might be related to the same subduction-related magmatism. However, it is also known that LIP events can have associated voluminous silicic magmatism and there can be ambiguity in distinguishing silicic magmatism associated with arcs from that associated with LIPs [7,73]. Chemical composition of the Suordakh event intrusions can test this arc-link hypothesis.
None of the studied samples show either Ta-Nb or Ti-negative anomalies, which are typical for the subduction-zone basalts. The Nb/La ratio was high, varying from 0.70 to 1.30, and averaging at 0.98 (Supplementary Materials, Table S5), which is typical for within-plate basalts not affected by contamination [52]. The distribution of HFSE and REE elements also supported the within-plate origin of the Suordakh event mafic intrusions. On the Th/Yb-Nb/Yb diagram, they plot within the mantle MORB-OIB array along with Devonian Sette-Daban and Vilyui rift dykes of the Yakutsk-Vilyui LIP [29,45], close to OIB and far from the field of arc basalts. They are also distinct from ca. 1000-950 Ma Neoproterozoic mafic intrusions from the same area, ca. 490-480 Ma post-orogenic intrusions from the southwest margin of the Siberian craton and 183 Ma flood basalts of the Karoo LIP (of South Africa), all of which show some influence of inherited subduction material or crustal contamination (Figure 15a) [30,51,74]. On a Zr/Nb-Nb/Th diagram (Figure 15b), the studied intrusions occupy the field of oceanic plateau basalts close to the margin of OIB and out of the field of island arc basalts. According to [75], the relationship between Zr, Nb, and Y calculated as ∆Nb = 1.  Table S5). Modeling of the mantle source using REE, indicated melting in transition from spinel to garnet lherzolite mantle material, with 3-10% of partial melting ( Figure 16).  [76]. Data source for Devonian Sette-Daban dykes [29], Devonian Vilyui rift dykes [29,45], Karoo basalts [51], Neoproterozoic (1000-950 Ma) intrusions, SE Siberia [30], early Paleozoic intrusions, SW Siberia [74]. (b) Zr/Nb-Nb/Th plot [77].   [51]. The tick marks on the curves correspond to the melting degrees. The grey curve is the melting curve of a spinel-bearing lherzolite source, with the following modal composition: 55% olivine, 15% orthopyroxene, 28% clinopyroxene, and 2% spinel. Black curves are the melting curves of garnet-bearing lherzolite mantle sources containing 2% (dashed and dotted), 4% (dashed), and 7% (solid) modal garnet, respectively. Source modal composition is 69-64% olivine, 20% orthopyroxene, 9% clinopyroxene, and 2-7% garnet.
However, some geochemical and isotopic features might be interpreted to reflect an influence of the arc-related processes. The enrichment of LILE over the HFSE and REE elements, likely points to addition of aqueous fluids during magma evolution, and the Ba/Th and Sr/Th ratios can be used to estimate the contributions from aqueous fluids in the melting source [78,79]. On the Ba/Th and Sr/Th vs. Th/Nd diagrams (Figure 17) studied, the samples followed a trend pointing to possible influence  [51]. The tick marks on the curves correspond to the melting degrees. The grey curve is the melting curve of a spinel-bearing lherzolite source, with the following modal composition: 55% olivine, 15% orthopyroxene, 28% clinopyroxene, and 2% spinel. Black curves are the melting curves of garnet-bearing lherzolite mantle sources containing 2% (dashed and dotted), 4% (dashed), and 7% (solid) modal garnet, respectively. Source modal composition is 69-64% olivine, 20% orthopyroxene, 9% clinopyroxene, and 2-7% garnet.
However, some geochemical and isotopic features might be interpreted to reflect an influence of the arc-related processes. The enrichment of LILE over the HFSE and REE elements, likely points to addition of aqueous fluids during magma evolution, and the Ba/Th and Sr/Th ratios can be used to estimate the contributions from aqueous fluids in the melting source [78,79]. On the Ba/Th and Sr/Th vs. Th/Nd diagrams (Figure 17) studied, the samples followed a trend pointing to possible influence of aqueous fluids on the magma composition, which in turn, might reflect subduction-related enrichment of mantle sources [79]. Variation of the ε Nd (t) values from 1.4 to 6.9, likely point to mixing of depleted and enriched sources (Supplementary Materials, Table S5). The radiogenic 87 Sr enrichment might also be interpreted as an influence of the subduction processes, although magmatic rocks with similar Nd and Sr isotope characteristics were also reported from LIPs [29,51,52,79,80]. Moreover, continental LIPs quite often display some subduction-like signatures [7,51,52,81] and interpretation of the Suordakh event mafic intrusions, as within-plate basalts seems to be the most reasonable. of aqueous fluids on the magma composition, which in turn, might reflect subduction-related enrichment of mantle sources [79]. Variation of the εNd(t) values from 1.4 to 6.9, likely point to mixing of depleted and enriched sources (Supplementary Materials, Table S5). The radiogenic 87 Sr enrichment might also be interpreted as an influence of the subduction processes, although magmatic rocks with similar Nd and Sr isotope characteristics were also reported from LIPs [29,51,52,79,80]. Moreover, continental LIPs quite often display some subduction-like signatures [7,51,52,81] and interpretation of the Suordakh event mafic intrusions, as within-plate basalts seems to be the most reasonable.

Conclusions
New U-Pb baddeleyite geochronological, chemical, and isotopic data provide new constraints on the origin and tectonic setting of the Suordakh event mafic intrusions in the Sette-Daban region of the Southeastern Siberian craton:

•
Based on the weighted mean date of U-Pb baddeleyite dates of 4 intrusions, the age of the Suordakh event was estimated at 454 ± 10 Ma; • New U-Pb baddeleyite dating significantly increased the possible volume of the Suordakh event intrusions to include intrusions in the Sette Daban region that have much lower TiO2, K2O, and P2O5 concentrations, than it was previously supposed by [10] and to include intrusions into the Cambrian strata; • Within error, the ages of mafic intrusions of the Suordakh event from Southeast Siberia, overlapped with those of within-plate mafic magmatic rocks from South Korea, Western Mongolia, Southern Argentina, Northern Iran, and Northwest Canada, supporting a significant pulse of ca. 450-440 Ma intraplate magmatism, which is approximately the age of the latest Ordovician mass extinction event; • Despite the relative close location of the Amandykan island arc and an active margin during the Ordovician, the Suordakh event mafic intrusions have chemical compositions typical of withinplate basalts.
Supplementary Materials: The following are available online at www.mdpi.com/xxx/s1, Table S1. Coordinates of samples of the Suordakh event mafic intrusions; Table S2. SIMS data from in-situ analyses of baddeleyite, Suordakh mafic sill, X04-14-1 [82]; Table S3. Major (wt.%) and trace (ppm) elements concentrations of the Suordakh event mafic intrusions; Table S4. Nd and Sr isotope data of the Suordakh event mafic intrusions; Table  S5: Chemical and isotopic characteristics of the Suordakh event mafic intrusions.

Conclusions
New U-Pb baddeleyite geochronological, chemical, and isotopic data provide new constraints on the origin and tectonic setting of the Suordakh event mafic intrusions in the Sette-Daban region of the Southeastern Siberian craton:

•
Based on the weighted mean date of U-Pb baddeleyite dates of 4 intrusions, the age of the Suordakh event was estimated at 454 ± 10 Ma; • New U-Pb baddeleyite dating significantly increased the possible volume of the Suordakh event intrusions to include intrusions in the Sette Daban region that have much lower TiO 2 , K 2 O, and P 2 O 5 concentrations, than it was previously supposed by [10] and to include intrusions into the Cambrian strata; • Within error, the ages of mafic intrusions of the Suordakh event from Southeast Siberia, overlapped with those of within-plate mafic magmatic rocks from South Korea, Western Mongolia, Southern Argentina, Northern Iran, and Northwest Canada, supporting a significant pulse of ca. 450-440 Ma intraplate magmatism, which is approximately the age of the latest Ordovician mass extinction event; • Despite the relative close location of the Amandykan island arc and an active margin during the Ordovician, the Suordakh event mafic intrusions have chemical compositions typical of within-plate basalts.