Petrography and Lithogeochemistry of the Uppermost Middle Jurassic–Lowermost Lower Cretaceous Strata in the Lower Reaches of the Anabar River (East Siberia, Arctic)

In recent time, the Laptev Sea shelf has been considered a promising object for the search for hydrocarbons in the Arctic water area of Russia (Kontorovich et al., 2010, 2014; Polyakova and Borukaev, 2017; Skvortsov et al., 2020). According to one of the widespread viewpoints, the sedimentary cover of the Laptev Sea is composed mainly of Cenozoic and, possibly, late Mesozoic continental sediments, which have a low petroleum potential (Vinogradov and Drachev, 2000; Franke et al., 2001; Shkarubo and Zavarzina, 2011). Other researchers (Kim et al., 2011; Kontorovich, 2020) believe that the deposits of the Siberian Platform cover extend into the water area of the Laptev Sea (its western and central parts). In this case, the submerged Paleozoic and Mesozoic strata are of significant interest in terms of petroleum potential. Russia’s strategy for the development of the North Siberian shelf depends on the solution of this problem. Therefore, it is highly important to reconstruct the geologic history of the Mesozoic framing of the Siberian Platform and the Laptev Sea coast, identify the sources and provenances of rocks, and elaborate sedimentation models.

The uppermost Middle Jurassic–lowermost Lower Cretaceous horizons are a system of reservoirs and seals and thus are a promising petroliferous object (Zakharov et al., 2013; Kashirtsev et al., 2018). Of special interest are stratigraphic analogues of the highly carbonaceous Bazhenov Formation in West Siberia. The most stratigraphically complete and well-studied Upper Jurassic–Lower Cretaceous silt–clay strata of marine genesis are present in the sections of the Nordvik Peninsula (southwestern coast of the Laptev Sea). South of them, where mostly silt–sandy sediments with numerous traces of erosion and sedimentation break accumulated in the coastal environments, the Upper Jurassic–Lower Cretaceous sediments were seldom the object of attention of specialists.

In the lower reaches of the Anabar River, there is a composite Jurassic–Cretaceous section, which is a key section of the upper Bathonian–Upper Jurassic strata in the west of the Lower Lena facies region of the Ob’–Lena facies area (Nikitenko et al., 2013). During a comprehensive study of this section, based on the materials of field work in 2013 (Nikitenko et al., 2022), we analyzed the stratigraphic, petrographic, and lithogeochemical characteristics of silt–clayey, silt–sandy, and terrigenous-carbonate rocks. The studied deposits form a series of outcrops along the right bank of the Anabar River between the mouths of the Srednyaya and Sodiemykha rivers (Fig. 1). The composite section comprises the upper part of the Yuryung-Tumus Formation (Bathonian; ca. 15 m thick), the Sodiemykha Formation (upper Bathonian–lower beds of the upper Oxfordian; 2–11 m thick), and the lower part of the Buolkalakh Formation (upper beds of the upper Oxfordian–Boreal Berriasian; >15 m thick) (Nikitenko et al., 2013, 2022). A specific feature of the studied rocks is the low degree of lithification.

Petrographic studies were performed for 46 thin sections of rocks of different types. The composition of clay material in silt–clayey rocks (eight bulk samples) was estimated using X-ray diffraction data (analyst N.A. Pal’chik, Institute of Geology and Mineralogy, Novosibirsk). Lithogeochemical studies were carried out for 44 bulk samples. The contents of major rock-forming oxides were determined on an ARL-9900-XP (Thermo Electron Corporation) X-ray fluorescence spectrometer, and the contents of trace and dispersed elements were measured on an ELEMENT (Finnigan MAT) high-resolution ICP mass spectrometer in the laboratories of the Institute of Geology and Mineralogy, Novosibirsk.

Based on the data obtained, we established the regularities of the change in the composition of the studied deposits under the influence of various paleogeographic factors. In this work, the main attention is focused on the influence of the composition and degree of chemical weathering of the provenance rocks. Earlier, the main ideas of the provenances of detrital material in the Arctic regions of East Siberia in the Jurassic and Cretaceous were based on the data of terrigenous-mineralogical analysis (Ronkina, 1965; Osipova, 1968; Kaplan et al., 1972; Kaplan, 1976). It was established that in the second half of the Bathonian (according to modern concepts), the intensity of erosion of the Siberian traps decreased. At the same time, the sedimentary basin began to receive more erosion products of amphibolite gneisses and the Permo–Triassic terrigenous complex of the Anabar Shield, as well as epidote-bearing and alkaline rocks, from local sources. In the Callovian, the erosion of ancient felsic igneous and metamorphic rocks of the Anabar Shield intensified, and the supply of material from intrusive and effusive traps again increased. In the Late Jurassic, the activity in the provenances changed little. The contribution of the Anabar crystalline massif slightly increased and continued in the Berriasian. The intensity of chemical weathering in the provenances increased in the Callovian and Kimmeridgian. Based on Sm–Nd isotope studies, Malyshev et al. (2016) concluded that in the Early Jurassic–Cretaceous, the main clastic material was supplied to the northeastern Siberian Platform through the erosion of the mature continental crust.

The petrographic features of Middle–Upper Jurassic and Lower Cretaceous rocks in the Arctic regions of East Siberia were earlier considered elsewhere (Ronkina, 1965; Kaplan, 1976; Kaplan et al., 1979; Levchuk, 1985). The area of the lower reaches of the Anabar River was regarded as the Late Jurassic–Early Cretaceous Khatanga–Anabar petrographic–mineralogical province (Kaplan, 1976). It was shown that the silt–sandy rocks of the province have a predominantly quartz–feldspar composition with varying amounts of mica–chlorite and calcite cement. The Callovian deposits include poorly sorted interbeds enriched in ferruginous-chlorite oolites, and the Kimmeridgian and Volgian deposits often have interbeds enriched in glauconite and phosphates. Carbonate nodule horizons are present in all deposits. Clay rocks have a predominantly smectite–hydromica composition and contain much silt impurity. The accessory minerals of the deposits of the Anabar River basin are dominated by sphene and epidote–zoisite group and black ore minerals.

In petrographic studies of the Middle Jurassic–Lower Cretaceous strata in the lower reaches of the Anabar River, we have established the regularities of changes in the composition of deposits in the studied sections and thus refined the strata lithology. Below we consider the composition and structure of the upper parts of the Yuryung-Tumus and Sodiemakha formations and the lower part of the Buolkalakh Formation.

The Yuryung-Tumus Formation. The studied upper part of the formation (Fig. 1) is composed of coarse-grained sandy siltstones and silts and fine- to medium-grained sandstones and sands (Fig. 2a, b). There are also subordinate silt–clayey rocks and terrigenous-carbonate nodule interbeds. The grain size of rocks increases upsection. The rock clasts are angular or, more seldom, semirounded, often elongate, which indicates their transfer from close distances and rapid burial. According to the composition of clasts, the silt–sandy rocks are graywacke arkoses, essentially feldspathic (field of nonterrigenous rocks) (Fig. 3). The clasts are feldspar (on average, 56%), quartz (on average, 28%), and lithoclasts (on average, 16%). The carbonate interbeds are characterized by a drastic increase in the portion of quartz as a result of the replacement of rock and feldspar clasts by calcite.

Quartz is usually single-crystal, often with fine dust and gas–liquid inclusions; some fragments show wavy and cloudy extinction. Feldspar is dominated by potassic varieties (on average, 42% of all clasts), mainly orthoclase (sometimes, microcline). Plagioclase makes up, on average, 14%. Feldspar is of two types: pure and subjected to sericitization and pelitization. Lithoclasts are mostly of frame type (on average, 13% of all clasts) and are effusive rocks of felsic and intermediate (seldom, mafic) compositions (8%). There are also fragments of siliceous rocks, quartzites, and fine-grained siltstones and single fragments of carbonate rocks. Plastic clasts (4%) are clayey and chloritized rocks and, seldom, mica schists. Mica is present in all rocks (few percent; in finer-grained varieties, its content reaches 8–10%). It is represented by biotite (often, chloritized or, sometimes, sideritized) and, less, muscovite; in places, detrital chlorite occurs. Accessory minerals are strongly dominated by minerals of moderate chemical and low hydromechanical stability: the epidote–zoisite group minerals, sphene, and, more seldom, hornblende (Fig. 2a, b); there are also minor zircon and garnet and single findings of apatite and tourmaline.

The rocks have clay and carbonate cements. The content of pore–film clay cement formed by chlorite–hydromicaceous material varies from fractions to 5%. Carbonate cement (20–50%) is unevenly distributed and occurs mainly in nodule beds, where it is formed by fine-grained porous and basal calcite (Fig. 2b). Intense calcitization results in terrigenous-carbonate rocks. In places, there is porous pelitomorphic siderite (fractions of percent; seldom, up to 1.5%), sometimes developed after clay rock fragments. Humic organic matter is present as inclusions of plant detritus (on average, few percent) and various macro- and microfauna remains. Pyrite is developed after plant fragments and as fine nodules (up to 1.5%). The rocks underwent minor secondary alteration.

The Sodiemykha Formation. It is composed mostly of fine-sandy and silty rocks and silts (Figs. 1 and 2c, f, g) with subordinate silt–clay interbeds. The lower part includes a lenticular marker bed (Fig. 2d, e; bed 6 in the A9 section, bed 2 in the OK1 section, and bed 20 in the A7 section) with inequigranular (mostly angular) silt–sandy material and a drastically increased portion of chlorite–hydromica clay (35–50%, up to its transition into clayey rocks). The bed is enriched in Fe-containing components (siderite nodules and goethite–chamosite oolites and peas) amounting to 15–20%; there are also phosphatized clasts. Such rocks are regarded as basal transgressive beds of condensed strata (Posamentier and Allen, 1999; Baraboshkin, 2009).

Analysis of the petrographic composition of silt–sandy rocks showed that they are more polymict than the underlying ones. The deposits of the basal beds of the Sodiemykha Formation, located under the marker bed (bed 5 in the section A9, bed 2 in the section OK2, and beds 18 and 19 in the section A7) and assigned to graywacke arkoses, are the most similar in composition to the rocks of the Yuryung-Tumus Formation. In most samples from the main part of the formation, the clasts are predominantly feldspar (on average, 42%); there are also lithoclasts (on average, 31%) and quartz (on average, 27%). This composition corresponds to quartz–feldspathic graywackes. There are also scarce interbeds of feldspar–quartz graywackes (up to 45% quartz). In the deposits overlying the marker bed, the rock fragments are predominantly lithoclasts.

The main clastic components are similar in characteristics to the rocks of the Yuryung-Tumus Formation. Quartz shows mainly wavy and cloudy extinction. K-feldspar (30%) dominates over plagioclase (10%). Lithoclasts comprise framework (12–20%) and plastic (10–23%) clasts in varying proportions. The former are effusive rocks of felsic and intermediate (less often, mafic) compositions (on average, 10% of all clasts), fine-grained siltstones (2%), siliceous rocks (1%), and quartzites (1%). Plastic clasts are mainly chloritized rocks (altered effusive rocks and mafic tuffs) (up to 15–20%). There are also fragments of clayey rocks (4%) and, more seldom, mica schists (2%). Some interbeds contain chlorite aggregates with a radiate texture. Mica is permanently present (few percent; sometimes, up to 8%). Most of the accessory minerals are inherited (Fig. 2c, f, g) and show a high chemical stability.

The studied rocks have a film-pore clay cement of chlorite–hydromica composition (5–25%). A high content of clay material (up to 50% and more) is found at some sites of the marker bed (Fig. 2d) and the overlying interbeds. Pelitomorphic siderite cement is permanent but unevenly distributed; it fills some pores and is developed after clay cement and some clasts (few to 10%; seldom, up to 20%). There are several calcitized beds with up to 30–50% calcite; in places, their rocks pass into terrigenous-carbonate rocks. Plant detritus (1–3%) is permanently present as inclusions. There are also sapropelic organic matter in the form of algae fragments as well as phosphatized clasts (probably, of fish bone tissue). Pyrite is extremely rare. Chloritization and sideritization of the deposits are more intense as compared with the studied part of the Yuryung-Tumus Formation.

The Buolkalakh Formation. The studied lower part of the formation has an essentially silt–clay composition, and the base is a regional glauconite marker bed with carbonate–phosphate nodules, a large amount of plant detritus of different sizes, and belemnite rostra (Figs. 1 and 2h; bed 8 in the section A9, bed 6 in the section OK2, bed 6a in the section OK1, and bed 22 in the section A7). The bed rocks are composed mainly of rounded and oval fine- to medium-sandy glauconite globules, sometimes with syneresis cracks filled with ferruginous material. Some grains have a thin concentric chamosite rim, and some are variably chloritized. There are also limonite–goethite and chamosite grains (up to 10%). The clasts are of fine- to medium-sandy size, amount to few percent, and are potassic and plagioclase feldspar, quartz, fragments of felsic and intermediate effusive and siliceous rocks, and biotite. The deposits contain accessory minerals of the epidote–zoisite group, apatite, and hornblende. Chlorite–hydromica cement is unevenly developed and amounts, on average, to 30% (in lenses its content reaches 50% and more). There are also plant detritus of different sizes (up to 1%) and scarce phosphatized clasts. Like the marker bed in the lower part of the Sodiemykha Formation, the described rocks can be regarded as basal transgressive beds of condensed strata.

The deposits overlying the marker glauconite bed and the above-lying clay deposits have interbeds of clay–carbonate rocks (Fig. 2i). The carbonate material is mainly micritic calcite with varying portions of pelitomorphic siderite and amounts, on average, to 40–50% (in places, up to 90%). It contains unevenly distributed cracked chamosite–glauconite globules and limonite–hematite oolites and their fragments. The clastic material of silt size amounts to 10%. There are also fine plant detritus and phosphatized clasts. The interbeds have sites with variably sideritized chlorite–hydromica material. According to Nikolaeva (1977), the structure of the glauconite globules in the basal bed indicates that they are mostly rewashed and are not significantly transferred. In the overlying interbeds, the globules are transported and are not paragenetic for the host sediment.

The clay beds overlying the glauconite bed (Fig. 2j; bed 10a in the section A9, bed 7 in the section OK1, and bed 24 in the section A7) contain a minor (mainly few percent) impurity of terrigenous material of fine-silt size. There are also beds of grains and clasts of chloritized glauconite, chamosite beans, and oolites, and scarce phosphatized clasts. Fine plant detritus is sometimes pyritized and makes up few percent. Pyrite is also found as fine framboids (Fig. 2j). There are permanent traces of the vital activity of organisms. Upsection, the portion of siltstone impurity gradually increases from 5–10 to 40–50% (Fig. 2k, l). It becomes more coarse-grained; coarse-silty and, seldom, fine-sandy fractions appear. Mica is represented by biotite and, less often, muscovite (up to 5–10%). Accessory minerals become more diverse (zircon, epidote–zoisite group minerals, sphene, and scarce apatite), and the amount of plant detritus increases (up to 5–6%). Fine- and fine-to coarse-silty interbeds appear; their portion drastically increases in the studied upper part of the formation (section OK4).

The clay material of the glauconite bed, concentrated mostly in globules, is represented mainly by highly disordered smectite (70–80%) as well as dioctahedral mica and Fe–Mg chlorite present in approximately equal portions. The overlying clay beds have a similar composition, with a somewhat lower portion of smectite (65–75%). Note that the mixed-layer component is widespread in many Upper Jurassic clayey sediments of the northern Siberian Platform (Kaplan et al., 1972; Kaplan and Nikolaeva, 1975) and is mainly disordered dioctahedral smectite of the hydromica–montmorillonite series. Upsection, the composition of clay material changes. Mica of the muscovite type 2M1 is predominant (50–60%), Fe–Mg chlorite is subordinate (25–40%), and the portion of disordered smectite decreases to 15–20%.

There are little lithogeochemical data on the Mesozoic strata of the Arctic regions of East Siberia (Galabala, 1971; Kaplan, 1976). These are mainly data on the contents of major rock-forming oxides in the large lithostratigraphic units (formations and strata) of some key sections. Levchuk (1985) reported the results of geochemical studies of the fraction smaller than 0.002 mm in some sections of the Yenisei–Khatanga trough. Comprehensive studies of the Jurassic–Cretaceous deposits of the Nordvik Peninsula provided information about the geochemical features of the near-boundary Volgian–Berriasian part of the section (Dypvik and Zakharov, 2010; Mizera et al., 2010).

Our lithogeochemical studies of the Middle Jurassic–Lower Cretaceous deposits in the lower reaches of the Anabar River yielded the first data on the chemical composition of various beds and members in the upper parts of the Yuryung-Tumus and Sodiemykha formations and in the lower part of the Buolkalakh Formation. Because of the specific composition and structure of the strata, we considered the lithogeochemical features of the former two essentially silt–sand formations and the predominantly silt–clay latter formation. The contents of rock-forming oxides and trace and dispersed elements are given in Tables 1–3.

The Yuryung-Tumus and Sodiemykha formations. The contents of major rock-forming oxides (Table 1) are generally similar in most of the studied silt–sandy rocks; however, we revealed certain lithochemical specifics marking variations in the petrographic composition. The nearroof beds of the Yuryung-Tumus Formation, with a high portion of quartz fragments, have an elevated (on the general background) content of SiO₂ and reduced contents of TiO₂, Al₂O₃, Fe₂O_{3 tot}, MgO, and S. The overlying basal beds of the Sodiemykha Formation (underlying a marker bed enriched in Fe-containing components in some sections) are, on the contrary, characterized by a low content of SiO₂ and high contents of CaO and P₂O₅. In general, the main part of the Sodiemykha Formation differs from the Yuryung-Tumus Formation in slightly higher contents of TiO₂ and Fe₂O_{3 tot} and slightly lower contents of Na₂O and K₂O, which reflects a decrease in the portion of feldspar and an increase in the portion of lithoclasts among the clasts. Specific rocks of the marker bed, rich in various authigenic components and slightly depleted in terrigenous component, have high contents of Fe₂O_{3 tot} and MgO and low contents of SiO₂, Na₂O, and K₂O. The studied calcareous interbed of the marker bed (bed 20 in the section A7) is characterized by a high content of CaO and an elevated content of P₂O₅ (an elevated content of phosphatized bone detritus).

In general, the low SiO₂/Al₂O₃ values of the studied silt–sandy rocks of the above formations (on average, 4.5) indicate their low sedimentation maturity. In the classification diagram from Pettijohn et al. (1972), almost all composition points of these rocks lie in the field of graywackes near the field of litharenites (Fig. 4a). In the classification diagram from Herron (1988), their composition points fall mostly in the field of wackes, and the points of some samples from the roof of the Yuryung-Tumus Formation fall additionally in the fields of litharenites and arkoses (Fig. 4b). The composition points of the marker bed lie at the boundary between ferruginous shales and ferruginous sandstones.

The main lithogeochemical features of the deposits are clearly seen in the major-oxide patterns normalized to the average Meso-Cenozoic graywacke (Condie, 1993) (Fig. 5a). The studied silt–sandy rocks differ from this reference gray-wacke in lower contents of MgO and higher contents of K₂O. In addition, the rocks of the Sodiemykha Formation have slightly higher contents of TiO₂ and Fe₂O_{3 tot} and higher contents of CaO. The specific rocks of the marker bed have high contents of TiO₂ and Fe₂O_{3 tot} and low contents of MgO, CaO, and Na₂O and show a wide spread in P₂O₅ contents.

Analysis of petrochemical indices (ratios of major rock-forming oxides) by the method elaborated by Yudovich and Ketris (2000) helped to make a chemical classification of rocks and to get an idea of the initial petrologic composition of the sediments (Table 1). Most of the studied silt–sandy rocks are characterized by persistent values of the hydrolysate index (HI), averaging 0.36. Taking into account their low contents of MgO (on average, 1%) and the iron index (II) (0.27–0.68), we assign the rocks to low- and normal-hydrolysate siallites. The rocks of the near-roof beds of the Yuryung-Tumus Formation show lower HI values (on average, 0.24), which, with regard to the low contents of MgO, permits them to be considered miosilites. These beds have the lowest contents of hydrolysis products. The low HI values indicate a moderate intensity of chemical weathering in the provenances. The rocks of the marker bed of the Sodiemakha Formation differ in lithogeochemical characteristics. Their elevated HI values (on average, 0.59), related to the significant amount of authigenic Fe-minerals in the rocks, permit them to be considered hypohydrolysates. The studied terrigenous–calcareous interbeds of the Yuryung-Tumus Formation and the marker bed are assigned to Ca-carbonatolites (CaO = 31.3 and 34.8%, respectively).

The average value of the femic index (FI) in most of the studied silt–sandy rocks is 0.13. Its low values (on average, 0.06) are typical of the near-roof beds of the Yuryung-Tumus Formation. The average II value is 0.33 in the studied part of this formation and 0.49 in the silt–sandy rocks of the Sodiemykha Formation, which permits these deposits to be classified as normal ferruginous. The rocks of the marker bed, rich in ferruginous components, are characterized by high FI and II values (on average, 0.35 and 0.97, respectively). The rocks of the upper part of the Yuryung-Tumus Formation and the basalt beds of the Sodiemykha Formation show persistent low TI values (on average, 0.04), which suggests the minor contribution of high-Ti basites to their formation. Higher TI values (on average, 0.08) are observed in the major part of the Sodiemykha Formation, including the marker bed, which probably indicates the higher contribution of basic igneous rocks to its accumulation.

The normalized-alkalinity (or sodium–potassic) index (SPI) varies slightly throughout the section. The higher SPI values of the Yuryung-Tumus Formation (on average, 0.50) relative to those of the Sodiemykha Formation (on average, 0.37) are probably due to the higher content of feldspar (predominantly potassic one) in the former. The hydrolysates of the marker bed are characterized by low SPI values (on average, 0.27). The alkali index (AI) varies negligibly throughout the section, averaging 0.87. In general, the estimated SPI and AI values permit the studied rocks to be considered normal alkaline.

The studied silt–sandy rocks show undisturbed positive TI–II and negative SPI–HI correlations, which suggests a significant content of components of the first sedimentation cycle (Yudovich and Ketris, 2000). Thus, taking into account the weak postsedimentary alterations of the rocks, we can state that their specific geochemical features are much due to the composition of the feeding provinces. In the genetic F1–F2 diagram (Bhatia, 1983) depicting the main composition of the parental strata, the composition points of the rocks of the Yuryung-Tumus Formation fall in the field of felsic igneous rocks, almost all points of the rocks of the Sodiemykha Formation are localized in the field of intermediate igneous rocks, and the points of the marker bed rocks lie in the field of mafic igneous rocks (Fig. 6a). In the Kossovskaya–Tuchkova (1988) classification diagrams of the main types of eroded rocks, most of the composition points fall in the fields of granodiorites and basaltic andesites (Fig. 6b, c). The points of the near-roof beds of the Yuryung-Tumus Formation, slightly enriched in quartz fragments, are shifted to the fields of felsic igneous rocks. Thus, felsic and intermediate rocks made the maximum contribution to the formation of the studied deposits. The contribution of felsic rocks slightly decreased, and the role of mafic rocks increased during the accumulation of the Sodiemykha Formation, especially its marker bed. Analysis of the arrangement of the composition points in the Nesbitt and Young (1982, 1989) weathering trend diagram showed a gradual intensification of chemical weathering in the feeding provinces during the accumulation of the upper parts of the Yuryung-Tumus and Sodiemykha formations (Fig. 7), especially the marker bed.

The specific distribution of trace and dispersed elements in the rocks of the formations throughout the section (Tables 2 and 3) is clearly seen in the element patterns normalized to the average Meso–Cenozoic graywacke (Condie, 1993) (Fig. 5b). The rocks of the Yuryung-Tumus Formation are characterized by similar or somewhat lower contents of many transition elements (Sc, V, Cr, Co, Ni) and LILE (Rb, Sr, Ba) elements and similar or elevated contents of HFSE (Y, Zr, Nb, Hf, Ta, Th, U, REE). The silt–sandy rocks of the Sodiemykha Formation show similar or slightly elevated contents of trace and dispersed elements and low contents of Co and Rb. Higher contents of Sr and Y are found in the basal bed in the section A9 of the formation, and higher contents of Ba are detected in some samples from the section OK1. The rocks of the marker bed have contents of transition elements (Sc, V, Cr, Co, Ni) several times exceeding the reference ones, which probably reflect the greater contribution of the destruction products of mafic igneous rocks (Wronkiewicz and Condie, 1987; Cullers, 2002). These contents are consistent with the chemical composition of the Siberian traps (Al’mukhamedov et al., 2004).

The REE patterns normalized to the average Meso–Cenozoic graywacke are similar for most of the studied silt–sandy rocks. The upper part of the Yuryung-Tumus Formation is enriched in LREE (on average, (La/Yb)_grayw = 1.4; (La/Sm)_grayw = 1.3; (Gd/Yb)_grayw = 1). This probably reflects a higher content of quartz in the sediments, which are subject to recycling. In the rocks of the Sodiemykha Formation, these ratios are close to unity. The average values of Eu anomaly (Eu/Eu* = 2Eu_N/(Sm_N + Gd_N)) in the studied silt–sandy rocks of the Yuryung-Tumus and Sodiemykha formations are 0.69 and 0.73, respectively (Table 3). The less pronounced Eu anomaly as compared with the average Meso–Cenozoic graywacke (0.6) (Condie, 1993) might be due to the smaller relative contribution of felsic igneous rocks to the composition of sediments. The high REE contents in the rocks of the marker bed are probably related to its nature (basal transgressive bed of the condensed strata) (Loutit et al., 1988).

The Ti/Zr ratio, used by many researchers to estimate the distance of the transfer of terrigenous material, divides the studied part of the section into two intervals. The average Ti/Zr value in the upper beds of the Yuryung-Tumus Formation and the lower beds of the Sodiemykha Formation is 14.2. In the main part of the Sodiemykha Formation, including the marker bed, Ti/Zr is, on average, 28. Taking into account the insignificant portion of Ti-rich basaltic rocks in the sediments, the above Ti/Zr values might indicate a lower maturity of the deposits in the main part of the Sodiemykha Formation because of the shorter distance of transfer of their components.

The La/Sc–Th/Co diagram (Taylor and McLennan, 1985) shows the arrangement of the composition points of the studied rocks relative to the average reference values for different types of rocks (Condie, 1993) (Fig. 8a). The points of the upper part of the Yuryung-Tumus Formation fall in the region between the reference points of Phanerozoic felsic volcanic rocks and granites, whereas the points of the Sodiemykha Formation lie in the region between the reference points of Phanerozoic felsic volcanic rocks and andesites (mainly near the reference point of the average Meso–Cenozoic graywacke). The points of the rocks of the marker bed and the overlying rocks are shifted to the field of average andesites because of the high Co contents. In the Hf–La/Th composition diagram (Floyd and Leveridge, 1987) for provenance rocks, almost all points of the studied rocks are localized in (or near) the field of deposits resulted from the destruction of felsic volcanic sources (Fig. 8b).

To assess the presence of exhalation material in the studied deposits, we analyzed the (Fe + Mn)/Ti ratio (Strakhov, 1976). It varies from 9 to 17 (on average, 12) in the silt–sandy rocks of the Yuryung-Tumus Formation and from 5 to 20 (on average, 10) in the Sodiemykha Formation and is 17–19 in the rocks of the marker bed. Its low values indicate no influence of exhalation. The same conclusions follow from the Ce/La values (Strekopytov et al., 1999), which are persistent for all the studied rocks (on average, 2.1) and are significantly higher than the normative boundary values (≤0.12–0.40).

The Buolkalakh Formation. The lithogeochemical characteristics of the studied deposits are intimately related to their compositional variations. The contents of major rock-forming oxides in the rocks (Table 1) are depicted in the PAAS-normalized composition diagrams (Taylor and McLennan, 1985) (Fig. 5c). The rocks of the marker glauconite bed and the overlying clay beds (0.5–1.0 m thick), similar in characteristics, differ from PAAS in the high average content of Fe₂O₃ (3.1 x PAAS and 2.7xPAAS, respectively), the low content of MnO (0.4 x PAAS), the elevated contents of TiO₂ and MgO, and the reduced content of Na₂O. In addition, the marker glauconite bed has a high average content of P₂O₅ (5.4 x PAAS), and some interbeds are enriched in CaO. The clay beds overlying the marker bed have similar contents of major rock-forming oxides but somewhat lower contents of Fe₂O₃ and much higher contents of P₂O₅. Upsection, these deposits are gradually (but rapidly) replaced by silt–clayey rocks with persistent contents of major rock-forming oxides. The content of Fe₂O₃ is similar to that in PAAS, and the content of MgO is lower (on average, 0.6 x PAAS). The gradual increase in the content of Na₂O in the main part of the section (up to 2.5 x PAAS) is probably due to the increase in the content of a terrigenous component, especially plagioclase, in the rocks. The presence of authigenic calcite in the upper part of the formation explains the slight increase in the content of CaO.

The location of the composition points of Fe₂O₃-enriched rocks of the formation bottom and overlying silt–clay deposits on Herron’s (1988) classification diagram permits them to be assigned to ferruginous and common schists, respectively (Fig. 4).

The contents of major rock-forming oxides in the silt–clayey deposits of the Buolkalakh Formation are consistent with those in the boundary Volgian–Berriasian bed of the Nordvik section (Dypvik and Zakharov, 2010). The studied rocks have higher contents of SiO₂, Na₂O, and K₂O (because of the higher portion of clastic material) as well as CaO (authigenic calcite) and P₂O₅ (bone detritus).

Analysis of the lithochemical features of the deposits (Yudovich and Ketris, 2000) (Table 1) showed that according to their hydrolysate index (0.65–0.83), the rocks of the marker glauconite bed and overlying clay beds can be assigned to hypohydrolysates and, partly, pseudohydrolysates because of their MgO content of ca. 3%. The overlying silt–clayey rocks show persistent HI values (on average, 0.45), which permits them to be classified (with regard to the average II = 0.42) as siallites of normal-hydyrolysate type. The femic index has high average values in the glauconite bed and overlying clay beds (0.49 and 0.44, respectively), which reflects a high content of authigenic Fe-minerals. The overlying siallites show persistent II values (on average, 0.18) and thus are normal ferruginous. The TI values are high in the glauconite bed and overlying clay members (on average, 0.08) and are persistent in silt–clayey rocks (0.05). The high TI values in the lower part of the formation might be due to the greater contribution of high-Ti igneous mafic rocks from the provenances.

The average SPI value in the marker glauconite bed is 0.34 and slightly decreases in the overlying clay beds (0.24). The above-lying silt–clayey rocks show the average SPI value of 0.37. The average AI value gradually changes throughout the section: from 0.25 in the glauconite bed to 0.37 in the overlying clays and 0.91 in the main the silt–clay part of the formation. This behavior of geochemical indices probably marks an increase in the content of terrigenous impurity (including plagioclase) in the rocks upsection.

As in the case of the silt–sandy rocks, the distinct positive TI–II and negative SPI–HI correlations in the deposits suggest a high content of components that formed during the first sedimentation cycle (Yudovich and Ketris, 2000). In the genetic diagram (Bhatia, 1983) (Fig. 6a), the composition points of the glauconite bed and overlying beds fall in the field of mafic igneous rocks, which might be due to the significant contribution of such provenances (e.g., the Siberian traps). The points of overlying silt–clayey rocks are located near the boundary of the fields of quartz-rich sediments and felsic and intermediate igneous rocks, which suggests the diverse composition of the provenance deposits, including the presence of earlier accumulated sedimentary rocks.

Comparison of the contents of trace and dispersed elements in the rocks of the Buolkalakh Formation and in the PAAS (Tables 2 and 3; Fig. 5d) shows that the rocks of the marker glauconite bed and overlying clays (0.5–1.0 m thick) with similar chemical characteristics are significantly enriched in many transition elements and U, whose contents might be several times higher than those in the PAAS. As in the case of the marker bed of the Sodiemykha Formation, these elevated contents might be due to the strong effect of the destruction products of mafic rocks (Siberian traps) on the sediment formation. The extremely high contents of Cr and Ni (on average, 474 and 133 ppm, respectively), along with the low contents of Cu and elevated contents of V, suggest that ultramafic igneous rocks were also destroyed (Voitkevich et al., 1977; Garver et al., 1996). In addition, these deposits are almost twice depleted in Cs and, to a lesser extent, in Ba. The contents of trace and dispersed elements in the overlying silt–clay bed of the Buolkalakh Formation are closer to the reference ones, being enriched in Sr (by 2.5 times relative to the PAAS) and, to a lesser extent, in Ba and significantly depleted in Cs (0.15 x PAAS).

The PAAS-normalized REE patterns (Table 3) show slightly elevated average REE contents (except for HREE) in the glauconite bed and overlying clays (average (La/Yb) _PASS = 1.3; (La/Sm)_PASS = 0.8; (Gd/Yb)_PASS = 1.7). The overlying silt–clayey rocks have REE contents similar to those in the PAAS, being slightly depleted in HREE (average (La/Yb)_PASS = 1.5; (La/Sm)_PASS = 1; (Gd/Yb)_PASS = 1.5). Similar tendencies are observed in the chondrite-normalized (Taylor and McLennan, 1985) REE patterns. The marker glauconite bed and overlying clays are characterized by average La/Yb)_N = 11.8, (La/Sm)_N = 3.4, and (Gd/Yb)_N = 2.3, and the overlying silt–clayey rocks show average (La/Yb)_N = 13.7, (La/Sm)_N = 4.3, and (Gd/Yb)_N = 18.6. The REE patterns of the studied rocks show that mafic igneous rocks made a minor contribution to their formation, whereas intermediate and, to a lesser extent, felsic rocks played a predominant role (Taylor and McLennan, 1985; McLennan et al., 1990). During the formation of the lower beds of the strata, the influence of mafic rocks slightly increased. The average value of Eu anomaly is 0.75 in the lower beds of the Buolkalakh Formation and 0.70 in the overlying silt–clayey rocks, which is close to the values in the average Paleozoic granodiorites and average Meso–Cenozoic andesites, 0.73 and 0.77, respectively (Condie, 1993).

Compared with the Volgian–Berriasian boundary bed of the Nordvik section (Dypvik and Zakharov, 2010), the studied silt–clayey deposits of the same age of the Buolkalakh Formation have lower or similar contents of trace and dispersed elements. The highest contents were established for Sr, Ba, Nb, and Th.

To estimate the composition of the studied rocks, we used some indicator ratios. In the La/Sc–Th/Co diagram (Taylor and McLennan, 1985) (Fig. 8a), which gives an insight into the contribution of different types of rocks to the sediment formation, the composition points of the glauconite bed and overlying clays fall in the field between the points of the average Phanerozoic andesites and graywackes (Condie, 1993), and the points of the overlying silt–clayey deposits lie between the points of the average Phanerozoic graywackes and felsic volcanic rocks. As noted above, the lower beds of the formation include igneous ultramafic rocks. The steady low contents of Cr and Ni (on average, 84 and 50 ppm, respectively) in the above-lying silt–clayey deposits in the main part of the formation indicate a minor amount of the destruction products of ultramafic rocks (Garver et al., 1996). The presence of exhalation material in the studied rocks was assessed. The (Fe + Mn)/Ti values (Strakhov, 1976) in the glauconite bed and overlying clays are 18–33 (on average, 24), decreasing to 13–15 in the overlying deposits. The Ce/La values (Strekopytov et al., 1999) are 1.9–2.0 in the glauconite bed and the main silt–clayey part of the formation and 2.2–2.7 in the clays overlying the glauconite bed. The negative Eu anomaly and persistent Ce anomaly values close to unity (0.9–1.1) indicate no influence of exhalation processes on the formation of sediments.

The performed studies of the Middle Jurassic–Lower Cretaceous deposits in the lower reaches of the Anabar River revealed the regularities of changes in the petrographic and chemical compositions of the upper part of the Yurung-Tumus and Sodiemykha formations and the lower part of the Buolkalakh Formation.

The chemical composition of silt–sandy rocks (gray-wacke arkoses and essentially feldspathic varieties) in the upper part of the Yuryung-Tumus Formation, the assemblage of accessory minerals, and the low degree of mechanical reworking of clasts point to a short distance of the rocks from the provenances. The lithogeochemical features of the deposits also mark their low sedimentary maturity and the petrogenic nature of clastic material and permit them to be classified as siallites and miosillites (near-roof beds of the formation) of normal alkalinity, which indicates a moderate intensity of chemical weathering in the provenances. We have established that mafic and intermediate igneous rocks prevailed during the sediment formation and sedimentary and metamorphic rocks were subordinate.

The Sodiemykha Formation is characterized by successive changes in the petrographic and chemical compositions of the deposits. The basal beds formed by graywacke arkoses and classified as siallites show a slight decrease in the content of felsic rocks and an increase in the content of intermediate rocks in the provenances. The marker bed is in sharp contrast to the rest of the formation: It is enriched in authigenic Fe-minerals permitting the rocks to be classified as hypohydrolysates and indicating the intensification of chemical weathering in the provenances and the possible contribution of the destruction products of lateritic weathering crusts. The trace and dispersed-element patterns of the rocks show a significant increase in the content of mafic igneous rocks in the provenances.

The above-lying part of the Sodiemykha Formation is composed of quartz–feldspathic and, more seldom, feldspar–quartz graywackes classified as siallites of normal alkalinity, which resulted from chemical weathering of moderate intensity in the provenances. The deposits are characterized by a low degree of sedimentary maturity, slightly increasing upsection. Their petrographic and chemical compositions point to a predominance of intermediate igneous rocks, subordinate felsic and mafic rocks, and minor sedimentary and metamorphic rocks in the provenances.

The clayey material of the marker glauconite bed and overlying beds (the basal part of the Buolkalakh Formation) is predominantly highly disordered smectite. The lithochemical features of the deposits permit them to be classified as hypohydrolysates and pseudohydrolysates that formed during the first sedimentation cycle. They are also evident of the intensification of chemical weathering in the provenances. The chemical composition of the rocks points to their formation as a result of the destruction of intermediate and mafic igneous rocks and, to a lesser extent, ultramafic rocks. The overlying silt–sandy deposits of the Buolkalakh Formation show a gradual increase in the content of silty material upsection. The feldspar clasts have a significant content of plagioclase. According to the lithochemical features, the deposits are siallites of normal alkalinity formed during the first sedimentation cycle with a gradual decrease in the intensity of chemical weathering in the provenances. Igneous rocks of intermediate and, to a lesser extent, felsic compositions were the main supplier of material.

Thus, based on the data obtained, we have established several stratigraphic boundaries in the section of the Bathonian–Berriasian strata in the lower reaches of the Anabar River, which mark changes in the petrographic and chemical compositions of the deposits: the boundary between the Yuryung-Tumus and Sodiemykha formations, the bases and roofs of the marker horizons, and the boundary between the clays overlying the glauconite bed and the above-lying silt–clayey rocks of the Buolkalakh Formation. We have revealed successive changes in the composition of the provenances, which are one of the main factors influencing the composition of sediments, and estimated the intensity of chemical weathering of rocks in the provenances. It is shown that intermediate and, to a lesser degree, felsic igneous rocks (Anabar Shield, local structures) were the main sources supplying clastic material. Periodically, the contribution of mafic (and, sometimes, ultramafic) rocks (Siberian traps) increased, especially during the formation of marker beds in the lower beds of the Sodiemykha and Buolkalakh formations. Metamorphic and ancient terrigenous rocks also participated in the sediment formation. The results obtained are consistent with the earlier viewpoints of the evolution of the regional provenances and refine the known data for the studied part of the sedimentary basin. The changes in the parental strata influenced the composition of sediments within large areas. Therefore, the revealed lithogeochemical changes are likely in the Upper Jurassic–Lower Cretaceous deposits of the entire considered petrographic and mineralogical province and can be used for deposit correlation.

We thank the reviewers for valuable comments and suggestions on the paper.

This work was supported by projects 18-17-00038 and 19-17-00091 from the Russian Science Foundation and by project 0331-2019-0021 from the Foundation for Basic Research.