Evolution of Syenite Magmas: Insights from the Geology, Geochemistry and O-Nd Isotopic Characteristics of the Ordovician Saibar Intrusion, Altai-Sayan Area, Russia

: In this paper, we provide insight into the evolution of syenite magmas based on geological data and petrographic, geochemical, and O-Nd isotope parameters of rocks of the Saibar intrusion located within the Minusinsk Trough, Altay-Sayan area. The intrusive suite includes predominant syenites, few bodies of melanocratic and leucocratic nepheline syenites (foyaites), and granites. In addition, dykes of granites and maﬁc rocks are present. The U-Pb zircon age from the melanocratic foyaites was determined to be 457 ± 27 Ma. Examined rocks show fractionated light rare earth element patterns, normalized to chondrite, with (La/Sm) n varying from 4 to 9, and a weakly fractionated distribution of medium and heavy rare elements, with (Dy/Yb) n from 0.35 to 1.23 and (Sm/Yb) n from 0.63 to 2.62. The spidergram normalized to the primitive mantle shows negative Ba, Sr, Nb, Ta, Ti, and Eu anomalies (Eu* = 0.48–0.60) and positive Rb, Th, and U anomalies. The δ 18 O values vary within 6.3 to 10.2% , and ε Nd(t) from +4.1 to +5.0. We observe gradual transitions from syenites to foyaites. Assimilation by syenite magma of the host carbonate rocks was followed to transition from silica-saturated to silica-undersaturated conditions and removal of anorthite from the melt, which then led to nepheline. Granites of the main phase show depleted lithophile incompatible elements in comparison with syenites and foyaites. They originate via interaction of magmas at the marginal part (endocontact zone) of the intrusion, corresponding to north contact of the granites with the host felsic rocks. In comparison, the rock composition of granite dykes is enriched in lithophile incompatible elements, except for Zr, Hf, and Ti. These rocks are formed due to the differentiation of syenite magma without a signiﬁcant effect of host rock assimilation. Mantle magmas must be used as parent magmas for syenites based on analysis of the formation model of other alkaline intrusions, which are similar in age to the Saibar intrusion. In the line of syenite intrusions of the Altai-Sayan province, the Saibar intrusion is no exception, and its origin is related to the evolution of maﬁc magmas that arose during the melting of the mantle under the inﬂuence of a mantle plume.


Geological Background and Petrography
Intrusions, primarily composed of quartz syenites, syenites, and granites, occur in the mountainous surroundings of the Syda-Erba depression (Figure 2a), enclosed in the Minusinsk Trough [49]. They enclose the "Saibar group" consisting of the Saibar (Figure 2b), Burovskiy, Kosmatiy, and Vysokiy intrusions, which contain both alkali quartz and nepheline syenites. This paper presents the results of our study of Ordovician syenites, foyaites, and granites of the Saibar intrusion, which is typical of a large group of Early Paleozoic syenite intrusions within the Minusinsk Trough (Figure 1b), located in the eastern part of Altai-Sayan region. A comprehensive study of this intrusion contributes to understanding the petrological and geodynamic mechanisms responsible for alkaline magmatism in the Altai-Sayan region.

Geological Background and Petrography
Intrusions, primarily composed of quartz syenites, syenites, and granites, occur in the mountainous surroundings of the Syda-Erba depression (Figure 2a), enclosed in the Minusinsk Trough [49]. They enclose the "Saibar group" consisting of the Saibar ( Figure  2b), Burovskiy, Kosmatiy, and Vysokiy intrusions, which contain both alkali quartz and nepheline syenites. Pre-Ordovician felsic magmatic associations and Cambrian sedimentary deposits (siliceous shales, limestones, sandstones) are host rocks for this group of intrusions. In this group, the Saibar intrusion (Figure 3a) is considered a representative, since it contains the most diverse set of rocks: foyaites, syenites, and granites. A.G. Vologdin discovered the intrusion in 1924 near the Saibar Mount [50]. This author did not provide a detailed description of the rocks and their relations with each other. Later, other explorers [51][52][53][54][55][56] described in detail the intrusion with the separation of alkali nepheline syenites and quartz syenites. However, geological data on nepheline syenites were contradictory; G.M. Saranchina [52] classified them as dykes, and E.E. Fedorov [54] considered them xenoliths in quartz syenites. The structure of the Saibar intrusion is determined by syenites, connected in places by gradual transitions with foyaites and granites, as well as by dykes of granite and mafic rocks.
Syenites occupy about 90% of the rock mass (Figure 3a,e,f). These are mainly granular rocks. They are subdivided into quartz-free and quartz-bearing varieties. Quartz-free varieties (Figure 4a,b) are dominated by alkali feldspar with numerous crudely perthitic domains composed of twinned albite prevails. Aegirine and arfvedsonite form aggregate of needles. Aegirine is green and brownish-green. In some cases, arfvedsonite replaces aegirine, retaining the octahedral shape of its cross-section, and also yields initially magmatic crystals with hexagonal and rhombic sections. Microcrystals of hornblende, albite, and biotite are found only in quartz varieties. Accessory minerals are titanite, apatite, and fluorite. The structure of the Saibar intrusion is determined by syenites, connected in places by gradual transitions with foyaites and granites, as well as by dykes of granite and mafic rocks.
Nepheline syenites were initially called "saibarites" [55], but in petrographic properties, following the classification in [1], they correspond to foyaites. These rocks comprise separate bodies (from 20 × 50 to 50 × 200 m) connected by gradual transitions with the hosting syenites (Figure 3b). These bodies are found in the northern and central parts of the Saibar intrusion. They morphologically appear as small outcrops and ridges 0.5 to 3 m high (Figure 3c) and have a dyke-like appearance. However, did not discover any clear intrusive contacts and quenched zones typical of dykes. Foyaites are characterized by variability of petrographic features and gradually transform into syenites. Their texture varies from hypidiomorphic granular to porphyry. Foyaites have melanocratic and leucocratic varieties and banded structures. These are alternating bands with varying ratios of alkali feldspar, albite, aegirine, arfvedsonite, and nepheline. The aegirine-rich bands of 2 . Micro-photos of representative thin sections: (a,b) quartz-free syenites, alkali feldspar undergoes albitization, aegirine and arfvedsonite form aggregate of needles; (c,d) foyaites, nepheline forms large crystals and xenomorphic grains between aegirine and alkali feldspar; (e,f) granites are composed of lamellar alkali feldspar, quartz-alkali feldspar-albite aggregate, and rare amphibole crystals. Left column: crossed polars, right column: parallel polars. Ne = nepheline, Aeg = aegirine, Arf = arfvedsonite, Amf = amphibole, Afsp = alkali feldspar, Ab = albite, Q = quartz.
Nepheline syenites were initially called "saibarites" [55], but in petrographic properties, following the classification in [1], they correspond to foyaites. These rocks comprise separate bodies (from 20 × 50 to 50 × 200 m) connected by gradual transitions with the hosting syenites (Figure 3b). These bodies are found in the northern and central parts of the Saibar intrusion. They morphologically appear as small outcrops and ridges 0.5 to 3 m high ( Figure 3c) and have a dyke-like appearance. However, did not discover any clear intrusive contacts and quenched zones typical of dykes. Foyaites are characterized by variability of petrographic features and gradually transform into syenites. Their texture varies from hypidiomorphic granular to porphyry. Foyaites have melanocratic and leucocratic varieties and banded structures. These are alternating bands with varying ratios of alkali feldspar, albite, aegirine, arfvedsonite, and nepheline. The aegirine-rich bands of 2 to 50 cm thickness wedge out, break off or swell, and grade into irregularly shaped spots (Figure 3d). Aegirine crystals seem aligned in one direction. In places, the bands exhibit coarser fragments containing large (about 1 cm) aegirine and nepheline crystals. Leucocratic foyaites are rich in alkali feldspar and albite. In all foyaite varieties, the alkali feldspar contains perthite with albite exsolutions. Magmatic albite is found as infrequent elongate polysynthetic twins attached to nepheline (Figure 4c,d). Nepheline is hexagonal-and tablet-shaped and commonly replaced by zeolite or analcime, rarely by aggregate of colorless fine-flaked mica. Aegirine forms the sheaf-and felt-like minerals; its color varies from yellowish-green to dark green. Aegirine is always more idiomorphic than arfvedsonite, which is brown and green with a purple (indigo) tint. The bulk mass has a trachytic and hypidiomorphic granular textures defined by various relations of aegirine and albite. Accessory minerals are the same as in nepheline syenites: titanite, apatite, and fluorite.
Granites of the main phase are present as moderately alkaline and alkaline varieties according to [1]. They differ in geological conditions of occurrence and mineral composition. These rocks make up irregular spots varying in size from 10 to 100 m 2 in the endocontact zones in the north and south of the intrusion and are located near the host rocks. Towards the center of the intrusion, they gradually become a coarse-grained quartz variety of syenite. In the endocontact zones, the rocks (Figure 4e,f) have a porphyric structure. Together with plentiful tabular alkali feldspar, they contain quartz-alkali feldspar-albite aggregate and rare amphibole crystals. Granite dykes are 20 m long; their thickness does not exceed 5 m, and that of quenched zones is only a few centimeters. The dykes intrude the Saibar syenites and foyaites, as well as the host rocks [53,54]. Along with alkali feldspar, quartz, and albite, dyke granites contain infrequent grains of aegirine-augite and arfvedsonite along with accumulations of magnetite, accessory apatite, and fluorite on the boundaries. A small number of alkali feldspar crystals have turned gray due to postmagmatic alteration.
Mafic rocks make up infrequent dykes intruding syenites in the central part of the Saibar intrusion. The contact of these dykes with the host rocks is sharp and clear. The dykes are sub-vertical with a predominantly northern strike (350 • -30 • ). The thickness of these dykes varies from 0.2 to 1 m, and their length is 5-15 m. The rock-forming minerals are labradorite and augite. The rocks have a micro-grained ophitic texture. In these rocks, chaotically oriented crystals of labradorite are partially enveloped in augite, which has no rims of its own. Brown tabular hornblende, biotite, and ore minerals are less common. Apatite, titanite, and zircon are typical accessory minerals. The rocks underwent lowtemperature hydrothermal alteration, which caused the replacement of labradorite by sericite, mafic minerals by chlorite and epidote, as well as the formation of secondary calcite.

Zircon U-Pb Dating
U-Pb dating of zircon was executed at the Institute of Precambrian Geology and Geochronology of the Russian Academy of Sciences, St.-Petersburg, Russia. Zircons were extracted from crushed rock using heavy liquids and magnetic separation techniques. The chemical decomposition of zircons and U-Pb isolation were performed with the conventional technique of T.E. Krogh [57]. The isotope analysis was accomplished with a Finnigan MAT-261 multi-collector mass-spectrometer (Thermo Electron GmbH, Bremen, Germany). The coefficients of fractionation are Pb-0.10% and U-0.08% per atomic unit of mass (a.u.m.). Blank contamination was not more than 0.1 ng Pb and 0.01 ng U. The uncertainties of the measured U/Pb ratios and U and Pb concentrations were 0.8%. Experimental data were processed in the programs PbDAT-1.22 [58] and ISOPLOT-2.10 [59]. In estimating the ages, the generally accepted values of the decay constants of uranium were used [60]. Corrections for common lead were introduced according to model values [61]. All errors are listed at the 2σ level. Micrographs of zircon were taken on an Ultra 55 scanning electron microscope (accelerating voltage 20 kV) (Carl Zeiss, Oberkochen, Germany).

Whole-Rock Composition
Samples from the Saibar intrusion were selected for chemical analysis after excluding those that were weathered. Major and trace elements were analyzed at the Institute of Geochemistry (Shared-Use Analytical Center for Isotopic and Geochemistry Studies), Siberian Branch of the Russian Academy of Sciences, Irkutsk, Russia. Major elements were analyzed by XRF with a CPM-25 multichannel spectrometer (Nauchpribor, Orel, Russia), as described by Afonin et al. [62]. Measurements of rare elements were performed using the ICP-MS method on an ELEMENT-2 Finnigan MAT high-resolution mass spectrometer (Thermo Electron GmbH, Bremen, Germany). To address possible matrix effects and to consider instability in the spectra (XRF), the analysts used the internal standard Rh. To calibrate the calculations of elemental contents in ICP-MS measurements, multi-element certified solutions CLMS-1, -2, -4, and SPEX (USA) were applied. Rock samples of mafic composition were prepared by open acid decomposition, and samples of syenites, foyaites, and granites were prepared by fusion with Li-metaborate.

O-Nd Isotopic Analysis
To perform isotope analysis of oxygen, we collected feldspar samples without any traces of secondary alterations from the Saibar intrusion. These were the fragments weighing 1.5-2.5 mg. They were examined at the Geological Institute, Siberian Branch of Russian Academy of Sciences, Ulan-Ude, Russia, with a Finnigan MAT 253 gas source mass spectrometer (Thermo Electron GmbH, Bremen, Germany) using a double lap system in a classic version (reference sample-sample). To measure δ 18 O values, the samples were prepared using laser fluorination with BrF5 reagent employing the technique reported in [63]. The set included the MIR 10-30 facility, and the analyzed minerals were laser heated with CO2 to 1000 • C, and the vacuum main line was used to purify and finally concentrate oxygen. The δ 18 O estimates were acquired versus international reference materials NBS-28 (quartz) and NBS-30 (biotite). The internal control was governed by regular measurements of UWG-2 (garnet), inner standard GI-1 (quartz), and Polaris (quartz) from the laboratory at the Institute of Geology of Ore Deposits, Petrography, Mineralogy and Geochemistry, Russian Academy of Sciences, Moscow, Russia. The error of obtained δ 18 O values was at the level (1σ) ± 0.2% . Representative samples were selected for Sm-Nd analysis, which was performed at the Institute of Geochemistry (Shared-Use Analytical Center for Isotopic and Geochemistry Studies), Siberian Branch of Russian Academy of Sciences, Irkutsk, Russia. About 100 mg of the whole-rock powder was dissolved in the mixture of HF, HNO 3 , and HClO 4 . Before dissolution, 149 Sm-150 Nd spike solution was added to all samples. The REEs were separated on BioRad AGW50-X8 200-400 mesh resin using conventional cation exchange techniques [64][65][66]. Sm and Nd were separated by extraction chromatography with LN-Spec (100-150 mesh) resin [67]. During the measurement, total blanks were 0.1-0.2 ng for Sm and 0.2-0.5 ng for Nd.
Isotopic compositions of Sm and Nd were determined on a NEPTUNE Plus Multi-Collector ICP-MS (Thermo Electron GmbH, Bremen, Germany) in static mode. The precision (2σ) of Sm and Nd contents and 147 Sm/ 144 Nd ratios was ±0.4% and ±0.003%

Zircon U-Pb Geochronology
To identify the time of the Saibar intrusion formation in the central part of the intrusion, we collected a bulk sample (20 kg) of melanocratic foyaites (sample 73-99), from which four fractions of zircons varying in coloring were analyzed. From these samples, by means of optical microscopy in transmitted light with the use of immersion liquids, non-zoned translucent brown and colorless zircons were selected without inclusions, which could be considered "ancient cores". The crystals have dipyrimid shape with developed prism sides ( Figure 5). The translucent brown crystal is primarily short columnar or isometric with elongation of 1.0-1.3. Colorless zircons include some grains with elongation of 1.5-2.5. The morphological and crystallographic features of zircon indicate its magmatic nature. which could be considered "ancient cores". The crystals have dipyrimid shape with developed prism sides ( Figure 5). The translucent brown crystal is primarily short columnar or isometric with elongation of 1.0-1.3. Colorless zircons include some grains with elongation of 1.5-2.5. The morphological and crystallographic features of zircon indicate its magmatic nature. The results of U-Pb isotope analysis for melanocratic foyaites of the central site are shown in Table 1 and Figure 6. Fraction 1 zircon contains an insignificant amount of U (24.2 ppm) and has a slight inverse age-related discordance. The U content in grains of fractions 2, 3, and 4 exceeds 1000 ppm, and their ages have a direct age sequence t( 206 Pb/ 238 U) < t( 207 Pb/ 235 U) < t( 207 Pb/ 206 Pb). The isochron built from points 2, 3, and 4 comes through the beginning of coordinates (22 ± 6 Ma) and crosses the concordia in the point of age value 457 ± 10 Ma, and the mean squared weighted deviation (MSWD) is 0.14. Considering that melanocratic foyaites gradually transform into syenites and are therefore coeval, we infer that the obtained dating corresponds to the age of the Saibar intrusion. .0 Notes: а-isotope ratios corrected for fractionation and blank; b-isotope ratios corrected for fractionation, blank and The results of U-Pb isotope analysis for melanocratic foyaites of the central site are shown in Table 1 and Figure 6. Fraction 1 zircon contains an insignificant amount of U (24.2 ppm) and has a slight inverse age-related discordance. The U content in grains of fractions 2, 3, and 4 exceeds 1000 ppm, and their ages have a direct age sequence t( 206 Pb/ 238 U) < t( 207 Pb/ 235 U) < t( 207 Pb/ 206 Pb). The isochron built from points 2, 3, and 4 comes through the beginning of coordinates (22 ± 6 Ma) and crosses the concordia in the point of age value 457 ± 10 Ma, and the mean squared weighted deviation (MSWD) is 0.14. Considering that melanocratic foyaites gradually transform into syenites and are therefore coeval, we infer that the obtained dating corresponds to the age of the Saibar intrusion. through the beginning of coordinates (22 ± 6 Ma) and crosses the concordia in the point of age value 457 ± 10 Ma, and the mean squared weighted deviation (MSWD) is 0.14. Considering that melanocratic foyaites gradually transform into syenites and are therefore coeval, we infer that the obtained dating corresponds to the age of the Saibar intrusion.   Table 2 shows the chemical compositions of the Saibar intrusive rocks. Syenites are characterized by SiO2 values from 61.52 to 66.39 wt.%, total alkalinity (Na2O + K2O) from 8.92 to 11.63 wt.%, and agpaite coefficients less than 1. All foyaites differ from syenites in high alkali contents (Na2O + K2O = 12.27-13.95 wt.%), which is reflected in the appearance of modal nepheline. Variations in SiO2 are noted in the range from 53.91 to 61.64 wt.%, which corresponds to the compositions of foid syenites, on the classification TAS diagram (Figure 7a; [69]). The average SiO2 content in melanocratic foyaites is 57.41 wt.%; in leucocratic 59.14 wt.%, the agpaite coefficient varies from 0.97 to 1.30 and does not depend on the melanocratic degree of the foyaites. In the TAS diagram (Figure 8), the points of their composition are located in the syenite field and at the boundaries of the syenite-alkaline syenite and syenite-alkaline granite or granosyenite fields [69]. All granites are similar in SiO2 content (68.75-69.93 wt.%) but differ in total alkalinity. Dyke granites are characterized by Na2O + K2O in the range from 10.24 to 10.91 wt.%, with variations of the agpaite coefficient from 0.95 to 1.01 (A.C., Table 2). The granite of the endocontact zone is characterized by the sum Na2O + K2O of 9.09 and the agpaite coefficient of 0.82. SiO2 content of mafic dykes varies from 61.52 to 66.39 wt.%; the total Na2O + K2O is from 10.24 to 10.91 wt.%, which corresponds to the monzogabbrodiorite composition [69].   Table 2 shows the chemical compositions of the Saibar intrusive rocks. Syenites are characterized by SiO 2 values from 61.52 to 66.39 wt.%, total alkalinity (Na 2 O + K 2 O) from 8.92 to 11.63 wt.%, and agpaite coefficients less than 1. All foyaites differ from syenites in high alkali contents (Na 2 O + K 2 O = 12.27-13.95 wt.%), which is reflected in the appearance of modal nepheline. Variations in SiO2 are noted in the range from 53.91 to 61.64 wt.%, which corresponds to the compositions of foid syenites, on the classification TAS diagram (Figure 7a; [69]). The average SiO 2 content in melanocratic foyaites is 57.41 wt.%; in leucocratic 59.14 wt.%, the agpaite coefficient varies from 0.97 to 1.30 and does not depend on the melanocratic degree of the foyaites. In the TAS diagram (Figure 8), the points of their composition are located in the syenite field and at the boundaries of the syenitealkaline syenite and syenite-alkaline granite or granosyenite fields [69]. All granites are similar in SiO 2 content (68.75-69.93 wt.%) but differ in total alkalinity. Dyke granites are characterized by Na 2 O + K 2 O in the range from 10.24 to 10.91 wt.%, with variations of the agpaite coefficient from 0.95 to 1.01 (A.C., Table 2). The granite of the endocontact zone is characterized by the sum Na 2 O + K 2 O of 9.09 and the agpaite coefficient of 0.82. SiO 2 content of mafic dykes varies from 61.52 to 66.39 wt.%; the total Na 2 O + K 2 O is from 10.24 to 10.91 wt.%, which corresponds to the monzogabbrodiorite composition [69].      The rocks comprising the main volume of the intrusion exhibited high total Fe, calculated as Fe 2 O 3 , and low MgO contents; the Mg# value varies in foyaites from 4 to 34 and in syenites from 3.5 to 24, and in granite Mg# corresponds to 32. In dyke granites Mg# varies from 4 to 22 (Table 2). These variations indicate a high degree of fractionation of the initial magmas and practically do not affect the change in the concentrations of both major components (SiO 2 , P 2 O 5 , TiO 2 ) and trace elements (La, Th, Zr, and Y; Figure 8a-f). Compared to these granites, foyaites and syenites have the highest TiO 2 (0.6-1.3 wt.%) and Zr (700-1550 ppm) contents. In all granite compositions, Zr is about 300 ppm. Dyke granites are enriched in P 2 O 5 , La, Th, and Y compared to other granites. Mafic dykes are characterized by high Mg# value (71). These rocks are enriched in P 2 O 5 but exhibit depleted La, Th, and Zr compared to other rock types of the Saibar intrusion.

Major and Trace Element Geochemistry
A wide range of ΣREE is typical for all samples: syenites (210-755 ppm), foyaites (142-1087 ppm), and granites (246-1771 ppm). The REE spectra normalized to chondrite [72] and spidergram normalized to primitive mantle [72] show a great similarity in the rock compositions of the Saibar intrusion (Figure 9). These rocks are geochemically close to A-type granites [73], which are characterized by enrichment in high field strength elements (HFSEs) and heavy rare earth elements (HREEs) in combination with low Sr and Ti concentrations, which is reflected in a spidergram. All rocks comprising the main volume of the Saibar intrusion demonstrate enrichment in light rare earth element (LREE) concentrations ((La/Sm) n = 4-9), weakly fractionated REE patterns, almost horizontal distribution in the medium rare elements (MREEs) as well as HREEs ((Dy/Yb) n = 0.35-1.23; (Sm/Yb) n = 0.63-2.62), and moderate negative Eu anomalies (Eu* = 0.48-0.60). These rocks have generally positive Rb, Th, U and negative Ba, Sr, P, Ti anomalies. Dyke granites differ from syenites and foyaites in the presence of a negative Zr-Hf anomaly. The nature of these differences may be associated with the fractionation of zirconium during the formation of granite dykes.
Syenites and granites of the main phase have a similar distribution of REE patterns (Figure 9a,b). Dyke granites (samples SBR 4/1 and SBR 4/3) and granites of main phases (sample SBR 1/4) differ according to geochemical data. Dyke granites are characterized by negative Nb, Ta, Zr, Hf, and Ti anomalies (Figure 9c,d) with significantly higher ΣREE concentration than in syenites. Granites of the main phase are distinguished by a lower content of total alkalinity and a higher potassium composition (K 2 O/Na 2 O = 1.1) in comparison with dyke granites (Table 2; Figure 7a,b). The Nb, Ta, Zr, Hf anomalies are practically absent, the Ti anomaly is weakly expressed in granites of the main phase, and in terms of the content of incompatible elements, these granites are close to the most primitive syenite with low ΣREE concentrations. Syenites are similar in geochemical characteristics to foyaites. A-type (anorogenic) granites are from [73]. Continental crust is from [74]. Figure 9. (a-j) REE diagram, normalized to chondrite [72], and spidergram, normalized to primitive mantle [72], for the Saibar rocks. The figure illustrates that leucocratic foyaites are enriched in LREEs and MREEs in comparison with the melanocratic foyaites. Dyke granites are enriched in Th, U, and REEs in comparison with granites of the main phase. Syenites are similar in geochemical characteristics to foyaites. A-type (anorogenic) granites are from [73]. Continental crust is from [74].

O-Nd Isotopes
The oxygen isotopic composition was analyzed in alkali feldspars of syenite, foyaite, and granite (Table 3). The δ 18 O values vary over a wide range (6.3-10.2% ). In the composition range from foyaites to syenites and granites, as SiO 2 increases, enrichment in the heavy oxygen isotope occurs (Figure 10a).  The Nd isotopic composition was investigated in two samples of melanocratic foyaites, three samples of syenites, and two samples of granites (Table 3). εNd(t) values were calculated using the U-Pb age of zircon (457 Ma) obtained from the study of melanocratic foyaites. All samples are characterized by positive εNd(t) values from +4.1 to +5.0. The narrow range of εNd(t) values in the Saibar rocks characterizes a homogeneous magma source. The εNd(t) values in the analyzed samples do not correlate with the SiO2 content (Figure 10b). The Nd isotopic composition was investigated in two samples of melanocratic foyaites, three samples of syenites, and two samples of granites (Table 3). εNd(t) values were calculated using the U-Pb age of zircon (457 Ma) obtained from the study of melanocratic foyaites. All samples are characterized by positive εNd(t) values from +4.1 to +5.0. The narrow range of εNd(t) values in the Saibar rocks characterizes a homogeneous magma source. The εNd(t) values in the analyzed samples do not correlate with the SiO 2 content (Figure 10b).

Parent Magmas and Petrogenesis
The following important geological observations make it possible to assess the composition of parent magmas and the evolutionary patterns of rock compositions using geochemical and isotopic data. Note that the formation of the intrusion was accompanied by at least two phases of injection. The formation began with the injection of large volumes of magmas, which served as the parent magma for syenites of the first (main) phase. In some cases, these rocks pass into foyaites, forming a continuous row: quartz syenites-quartz-free syenites-nepheline syenites (foyaites). In the marginal parts of the main intrusive phase, corresponding to the endocontact zone, quartz syenites transformed into granites. Later (second phase), syenites of the intrusive were penetrated by rare dykes of granites and mafic rocks, which do not intersect with each other.
Within the Saibar intrusion, there are no mafic rocks with a SiO 2 content of less than 52 wt.% that preceded or accompanied the formation of syenites. This limits the possibility of reconstructing the composition of magmas that was parental for the syenitic rocks. At the same time, the other syenite-bearing intrusions of the Altai-Sayan fold belt contain mafic rocks formed synchronously with examined syenites, which indicates their correspondence to the composition of parent magmas [32][33][34][35][36][37][38][39][40][41][42][43][44][45]. Accordingly, the geological affinity of the Saibar intrusion to other similar intrusions suggests a genetic connection between its rocks and the mafic magmas, which occurred at levels of the magmatic system inaccessible to observation. Mafic dykes of the Saibar intrusion occurred after the formation of the main intrusion phase; therefore, we have no reason to classify them as rocks corresponding to the composition of the parent magmas.
Accordingly, we considered the part of the magmatic evolution of the initial magmas, which corresponds to syenites, foyaites, and all varieties of felsic rocks.
We found that the main intrusive phase contains a forbidden association of quartzbearing rocks and nepheline-bearing rocks. The points of the modal compositions of these rocks, calculated by the CIPW method, are subdivided in the Ne-Di-Q triangle diagram ( Figure 11).

Parent Magmas and Petrogenesis
The following important geological observations make it possible to assess position of parent magmas and the evolutionary patterns of rock compositions u ochemical and isotopic data. Note that the formation of the intrusion was accom by at least two phases of injection. The formation began with the injection of large of magmas, which served as the parent magma for syenites of the first (main) p some cases, these rocks pass into foyaites, forming a continuous row: quartz s quartz-free syenites-nepheline syenites (foyaites). In the marginal parts of the ma sive phase, corresponding to the endocontact zone, quartz syenites transformed in ites. Later (second phase), syenites of the intrusive were penetrated by rare dykes ites and mafic rocks, which do not intersect with each other.
Within the Saibar intrusion, there are no mafic rocks with a SiO2 content of 52 wt.% that preceded or accompanied the formation of syenites. This limits the po of reconstructing the composition of magmas that was parental for the syenitic r the same time, the other syenite-bearing intrusions of the Altai-Sayan fold belt mafic rocks formed synchronously with examined syenites, which indicates the spondence to the composition of parent magmas [32][33][34][35][36][37][38][39][40][41][42][43][44][45]. Accordingly, the geolo finity of the Saibar intrusion to other similar intrusions suggests a genetic conne tween its rocks and the mafic magmas, which occurred at levels of the magmati inaccessible to observation. Mafic dykes of the Saibar intrusion occurred after mation of the main intrusion phase; therefore, we have no reason to classify them corresponding to the composition of the parent magmas.
Accordingly, we considered the part of the magmatic evolution of the initial m which corresponds to syenites, foyaites, and all varieties of felsic rocks.
We found that the main intrusive phase contains a forbidden association o bearing rocks and nepheline-bearing rocks. The points of the modal compositions rocks, calculated by the CIPW method, are subdivided in the Ne-Di-Q triangle ( Figure 11).  Foyaite compositions are plotted in the Ne-Di-Ab field, while the compositions of all varieties of syenite, as well as granites, are located in the Di-Ab-Q field. At the same time, syenite compositions occupy an intermediate field between foyaites and granite compositions. Thus, gradual geological transitions from nepheline to quartz-bearing magmatic rocks cannot be explained by any regularity in crystallization differentiation of either the initial foyaite or initial syenite magma due to the existence of an insurmountable chemical barrier between nepheline-normative and quartz-normative melts. Moreover, among all rocks there are no correlations between Mg# ( Figure 8) and incompatible elements, which could be confidently explained by fractionation of mineral phases. The spidergrams among the compositions of all varieties of syenites also show compositions close to those of both melanocratic and leucocratic foyaites (Figure 9e,f). This similarity of compositions is especially clearly manifested in relation to Rb, Ba, Th, U, K, Nb, Ta, P, Zr, Hf, and Ti. At the same time, there are some features of REE differentiation for foyaites and for syenites. Foyaites differ from each other in LREE and MREE contents, while syenites differ from each other in LREE, MREE, and HREE contents. These data on the bulk composition of rocks contradict the formation model of syenites from the parent foyaite magma with relatively low SiO 2 concentrations. In this study, we considered additional mechanisms of formation of silica in under-and oversaturated magmas. Such a mechanism should be the contribution of the xenogenic component to the magmatic system corresponding to the composition of syenites, which would disrupt the course of differentiation. The assimilation by syenite magma of the limestone fragments of Cambrian host blocks confidently explains the appearance of nepheline magmas and rocks. Thus, the contribution of the calcium component to the parent syenite magma could cause a relative decrease in the SiO 2 content up to the unsaturation of the magmatic system. For example, if 3% CaO (7% CaCO 3 ) is added to the melt corresponding to the syenite (SBR 3/1), then its composition will shift towards the appearance of normative nepheline. During the differentiation of such magma, and primarily due to early crystallization and removal of anorthite plagioclase from the melt, a series of rocks richer in alkalis and, accordingly, nepheline can be formed. This is confirmed by the following reaction: CaCO 3 (calcite) + Na(Al,Si)Si 3 O 8 (albite) +1/2Al 2 O 3 (melt) = CaAl 2 Si 2 O 8 (anorthite) + NaAlSiO 4 (nepheline) + CO 2 . The contribution of carbonate component will practically not change the trace element contents of the magmas ( Table 2, hybrid melt); therefore, geochemically, foyaites remain similar to syenites. Generally, they are also more enriched in Sr, which is a typical trace element of carbonate rocks. The discovery of foyaite outcrops in different sites of the Saibar intrusion, regardless of their distance from the contact zone with the host limestones, indicates a random distribution of the supposed limestone fragments in the magma chamber.
In contrast to the foyaitic sites, granites of the main phase are confined to the endocontact zone, which directly indicates the interaction of syenite magma with host rocks. Accordingly, both syenite and foyaite samples are characterized by a general enrichment in REEs, which correlates with the appearance of its leucocratic types. On the contrary, granites of the endocontact have depleted geochemical characteristics in comparison with other rocks of the Saibar intrusion (Figure 9a,b). Their element concentration is close to the average composition of the continental crust, as well as to the composition of the host intrusive felsic rocks, and occupies the position of a hybrid melt formed during the interaction of syenite magma with such host rocks.
Evaluating the regularities of magmatic evolution in this time interval, we must note an increase in the total alkalinity of felsic rocks. This is evidenced by the composition of the latest granite dykes. The composition of dyke granites is close to syenites in terms of the contents of Rb, Ba, K, Nb, Ta, Sr, and P, but they are enriched in REE concentrations (Figure 9c,d). This indicates the fractionation of mafic alkaline minerals of syenite magma without significant interaction with host crustal rocks. The weak Zr-Hf and Ti negative anomaly in the dyke granites indicates the early crystallization of titanium-bearing minerals.
Thus, geochemical data indicate a disruption of the differentiation process during the evolution of syenite magma and a change in its composition during the assimilation of different host crustal rocks. Our conceptual scenario assumes sites with captured blocks of carbonate rocks were formed in different segments of the intrusion remote from the endocontact zone. Differentiation of the syenite magma and an increase in its total alkalinity and desilication occurred, leading to the formation of foyaites, including distinguished melanocratic and leucocratic types, differing in the pattern of distribution of incompatible elements.
The combination of fractionation and assimilation is indicated by the isotopic composition of oxygen, in particular the δ 18 O enrichment of alkali feldspars of syenites and granites of the endocontact zone.
Generally, the εNd(t) vs. age (Ma) diagram for the examined rocks shows that the Nd isotopic compositions are plotted in the field of Nd isotopic evolutions for Early Paleozoic granites of Central Asia, including the Early Paleozoic intrusions of the Kuznetsk Alatau ( Figure 11; [48,[75][76][77]). This assumes potential contaminants for Ordovician syenite magmas. At the same time, high εNd(t) values coincide with the isotopic neodymium characteristics for the Early Paleozoic mafic rocks and syenites related to the withinplate setting of the eastern Altai-Sayan region [79]. All these data point to the participation of magmas of a crustal and mantle nature during the formation of the Saibar intrusion.

Geodynamic Setting
Ordovician syenites are involved in the geological structure of volcano-plutonic areas, which are distributed in the mountainous frame of the Minusinsk Trough and more widely throughout the eastern part of the Altai-Sayan fold belt. As shown previously [33,80,81], these areas are composed of magmatic rocks with a wide range of SiO 2 content. Their geochemical and isotopic (Sr, Nd) characteristics reflect the interaction of mantle and crustal sources, which occurs both on active continental margins and in areas of withinplate magmatism with the participation of a mantle plume [82]. At the same time, geological data indicate a passive-margin geodynamic setting [83] and the withinplate position of magmatic areas. This geological restriction allows us to conclude that Ordovician magmatism was most likely determined by the effect of a mantle plume on the continental lithosphere in the northern part of the Altai-Sayan fold belt. This conclusion is consistent with the geochemical features of rocks of the Saibar intrusion, which are moderate and alkaline in nature; enrichment in most lithophile elements; and a clear geochemical difference from the composition of the crust and their affinity to the A-type granitoids.

Conclusions
The obtained geological and isotope-geochemical data for the Saibar intrusion rocks showed the following: 1. The Saibar intrusion is composed of dominant syenites, as well as few bodies of melanocratic and leucocratic nepheline syenites (foyaites) and endocontact granites. These rocks form the main magmatic phases. In addition, there are dykes of granites and mafic rocks. The U-Pb age of zircons in the melanocratic foyaites is estimated at 457 Ma.
2. The geological characteristics of mafic rocks indicate their corresponding formation in relation to the rocks of the main intrusion phase, which restricts the possibility of reconstructing of the parent magma composition. Accordingly, it is possible to discuss only that part of the magmatic evolution of parent magmas, which corresponds to syenites, foyaites, and all varieties of felsic rocks. The main phases of the Saibar intrusion were formed by a combination of fractionation and interaction of syenite magmas with host rocks. The foyaites appeared due to the fractionation and removal of Ca-rich plagioclase from the melt during the assimilation of host carbonate rocks. The composition of granite (main phase) was determined by the interaction with host intrusive felsic rocks in the endocontact zone. Dyke granite was formed by differentiation of syenite magma without a marked contribution of the host continental crust.
3. The geological affinity (age of formation, presence of nepheline rocks) of the Saibar intrusion with the Ordovician intrusions of the alkaline Altai-Sayan province makes it possible to associate its rocks with the evolution of mafic magmas, which occurred at levels of the magmatic system inaccessible to observation. The origin of such magmas was associated with the effect of a mantle plume on the continental lithosphere of the Altai-Sayan fold belt.