BOREAL–TETHYAN CORRELATION OF THE UPPER BERRIASIAN–VALANGINIAN: CONTRIBUTION OF NEW δ¹³C AND ⁸⁷Sr/⁸⁶Sr CHEMOSTRATIGRAPHIC DATA FROM ARCTIC SIBERIA

The current state of the Lower Cretaceous Standard Mediterranean Ammonite Zonation (SMAZ) (Western Tethys) has been characterized only recently [Szives et al., 2024]. This scale serves a chronostratigraphic function. The Boreal–Tethyan correlation of the Berriasian, Valanginian, and Hauterivian stages, which together form the so-called Neocomian, is based almost exclusively on the analysis of the paleontologic characteristics of ecotone paleobasins, in which the Boreal and Tethyan biota occasionally mixed [Baraboshkin et al., 2007; Bragin et al., 2013; Reboulet et al., 2014]. At the same time, the interregional correlation of these deposits remains difficult, because the paleobiota is highly provincial. Based on historical precedents of ammonite dispersal, the global correlation of the Lower Cretaceous on a biostratigraphic basis is also complicated by the nonisochronous appearance of common taxa in different marine paleobasins, especially during meridional migrations [Guzhikov and Baraboshkin, 2006; Baraboshkin et al., 2007]. Thanks to the long-term efforts of many scientists from different countries, the global correlation of the Jurassic–Cretaceous boundary interval has been substantiated on the basis of a comprehensive analysis of the bio-, magneto-, and chemostratigraphic characteristics of the Tethyan and Boreal key sections of the upper Tithonian–lower Berriasian. The main results of these works are presented in two papers written by the Berriasian Working Group of the International Subcommission on Cretaceous Stratigraphy, headed by W. Wimbledon from 2007 to 2020 [Wimbledon et al., 2020a,b]. The study of the upper Berriasian, Valanginian, and Hauterivian performed using nonpaleontological stratigraphic methods still has many noticeable gaps.

The paleomagnetic method has been successfully applied in recent decades to study the upper Berriasian–Hauterivian, including the Boreal regions of Russia. In particular, according to generalized information [Baraboshkin and Guzhikov, 2018], magnetostratigraphic studies are conducted in Central Russia (Middle Volga region), northern European Russia (Izhma River), West Siberia near the Subpolar Urals (Yatriya River), and northern East Siberia (Boyarka River). However, the coverage of regional stratigraphic columns with paleomagnetic data is still far from complete and continuous not only because of numerous gaps in sedimentation or condensation of deposits (Middle Volga region), but also because of the lack of polarity data in many (small or large) sectional intervals (the Izhma, Yatriya, and Boyarka rivers). Comprehensive C, O, and Sr isotope characteristics in Boreal regions (not without gaps) are reported for the following sites: the uppermost Ryazanian–Hauterivian of England (Speeton) and northeastern Greenland (Wollaston Foreland) and the Valanginian–lowermost Hauterivian of the Lower Saxony Basin in northwestern Germany [Jones et al., 1994; Price et al., 2000; McArthur et al., 2004; Mutterlose et al., 2014; Meissner et al., 2015; Möller et al., 2015]. A similar set of studies was previously conducted in the Boreal regions of Russia for the Volgian–Ryazanian boundary interval in West Siberia near the Mauryn’ya River in the Northern Urals [Dzyuba et al., 2013; Kuznetsov et al., 2017] as well as for the Ryazanian–lowermost Valanginian in Central Russia: Kashpir settlement, Middle Volga region [Gröcke et al., 2003].

The first δ¹³C and δ¹⁸O data for the Neocomian formations of Siberia were obtained for belemnites from the Hectoroceras kochi–Speetoniceras versicolor ammonite zones on the Yatriya River in West Siberia [Price and Mutterlose, 2004]. The δ¹³C variation curves constructed in this work clearly show the Weissert event [Erba et al., 2004], the first of the largest δ¹³C positive excursions in the Cretaceous Period, most clearly expressed in the late Valanginian. Subsequently, the interpretation of the δ¹³C values and paleomagnetic characteristics of the Yatriya section formed the basis for the conclusion about the late Valanginian age of the Homolsomites bojarkensis ammonite Zone [Baraboshkin et al., 2006], recognized in Siberia and the Arctic and usually considered to be Hauterivian.

The next publication is focused on comparing the composite δ¹³C variation curve obtained using belemnites and fossil wood from the interval between Hectoroceras kochi and Homolsomites bojarkensis zones on the Boyarka River (northern East Siberia) as well as using belemnites from the interval between ammonite layers with Pseudocraspedites and Surites and the Dichotomites bidichotomus Zone on the Izhma River (northern European Russia) with the Tethyan δ¹³C curve for the upper Berriasian–Valanginian [Nunn et al., 2010]. If we take into account the strongly manifested middle Valanginian episode of rapid growth of δ¹³C values in both the Tethyan and Boreal sections, the boundary between the Valanginian substages is identified at the base of the Boreal Polyptychites polyptychus ammonite Zone, which corresponds to the Polyptychites beani Subzone of the Siberites ramulicosta Zone in northern Siberia [Baraboshkin, 2004]. At the same time, the base of the D. bidichotomus Zone, which is usually correlated with the base of the upper Valanginian [Bogomolov, 1989; Nikitenko et al., 2013], is compared with the base of the Pronecostatum Chronosubzone (the upper subzone of the upper Valanginian Verrucosum Chronozone). The H. bojarkensis Zone is placed at the top of the Valanginian as in the Yatriya section. Interestingly, this is the first time that there are available data for verifying these constructions using the Sr isotope stratigraphy (SIS) method. Due to the lack of clear δ¹³C benchmarks in the Ryazanian–lowermost Valanginian, the correlation of the Boreal and Tethyan ammonite scales in this interval was reproduced in [Nunn et al., 2010] according to one of the previously accepted versions [Baraboshkin, 2004]. Moreover, in view of the biostratigraphic correspondence of the Boreal Tollia tolli ammonite Zone to the Otopeta Chronosubzone, identified as a Berriasian terminal subzone [Reboulet et al., 2009], the base of the Valanginian in the Boreal sections is accepted as the base of the Neotollia klimovskiensis ammonite Zone rather than of the Tollia tolli Zone, as accepted previously [Baraboshkin, 2004]. The Ryazanian interval is fully compared with the upper Berriasian.

Later research in Siberia was closely focused on the Jurassic–Cretaceous boundary deposits, which also cover the boundary between Volgian and Ryazanian deposits, for which δ¹³C and δ¹⁸O characteristics (from belemnite rostra) were obtained for the first time in the Nordvik Peninsula (northern East Siberia) and Mauryn’ya River (northwestern part of West Siberia) sections [Žák et al., 2011; Dzyuba et al., 2013]. For the upper Volgian–Ryazanian, a composite δ¹³C curve (reference for Boreal regions) was proposed and compared with a succession of bio- and magnetozones based on data from Siberian sections [Dzyuba et al., 2013]. Finally, Sr isotope data were presented for the Mauryn’ya section for the first time [Kuznetsov et al., 2017]. These data filled the gap that existed at that time in the ⁸⁷Sr/⁸⁶Sr ratio curve in the ocean in the Jurassic–Cretaceous boundary interval corresponding to the former Jacobi Chronozone [Ogg et al., 2012; Reboulet et al., 2014], which is known as the Chaperi–Jacobi Chronozone interval [Szives et al., 2024], and supplemented the characteristics of the Occitanica Chronozone interval. It was concluded on the basis of the analysis of bio-, magneto-, and chemostratigraphic correlation criteria that the basal Ryazanian Chetaites sibiricus ammonite Zone approximately corresponds to the middle Berriasian Occitanica Chronozone [Bragin et al., 2013; Dzyuba et al., 2013; Kuznetsov et al., 2017]. Note that, following [Igolnikov et al., 2016; Igolnikov, 2019], we suggest that the Praetollia maynci ammonite Zone is localized at the bottom of the Ryazanian Stage of Siberia rather than the Chetaites sibiricus Zone, which was distinguished within that interval before. Note also that now the Berriasian Stage is divided into two substages, with the Tirnovella occitanica ammonite Zone belonging to the uppermost lower Berriasian [Szives et al., 2024].

The goals of this study include obtaining a comprehensive C, O, and Sr isotope characteristic of the Lower Cretaceous of northern East Siberia in the interval from Hectoroceras kochi Zone to Homolsomites bojarkensis Zone based on an isotope–geochemical study of carbonate material in the belemnite rostra originating from the Boyarka River, Nordvik Peninsula, and Anabar River outcrops (Fig. 1). It is particularly important to solve the problems associated with the lack of consensus on the position of the lower/upper Berriasian, lower/upper Valanginian, and Hauterivian in the Boreal sections. It would also be interesting to analyze the capabilities of the isotope chemostratigraphy method for assessing the correctness of the detailed (at the level of biostratigraphic zones and subzones and their separate intervals) correlation of the Boreal (Siberian) ammonite zonal standard with the SMAZ. All the studied sections belong to the category of Neocomian reference sections in northern East Siberia [Gol’bert, 1981].

The scheme illustrating the structural–facies zoning of the Lower Cretaceous deposits of the northern part of East Siberia confines the studied sections to the Boyarka (Boyarka River), Paksa (Nordvik Peninsula), and Anabar (Anabar River) regions (Figs. 1a, 2). These are the most important regions for Neocomian marine studies here because of the presence of sections of different facies, all of which together represent a complete succession of zones of the regional ammonite scale, crowned here by the Homolsomites bojarkensis Zone [Gol’bert, 1981; Bogomolov, 1989; Nikitenko et al., 2013]. The absence of long sedimentation gaps and the numerous remains of belemnites in the deposits of the sublittoral facies create excellent prospects for the use of many sections for chemostratigraphic constructions. According to paleogeographic reconstructions (https://paleolatitude.org), the region under study is located at the onset of the Cretaceous interval, approximately between 69º (Boyarka) and 75º (Nordvik, Anabar) of the northern paleolatitude. Parallel zonal scales for different groups of fauna, including belemnites, and flora have been suggested for the Neocomian formations of northern Asian Russia mainly on the basis of the study of East Siberian sections. The entire scale package was last revised a little over ten years ago [Nikitenko et al., 2013]. This package includes the results of the development of the zonal scale based on belemnites for the Volgian–Ryazanian boundary interval [Dzyuba, 2012]. Currently, the belemnite scale in the interval of the uppermost Ryazanian and Valanginian is being revised [Efremenko, 2023a]. The East Siberian scale based on Buchia (bivalves) is distinguished by the highest degree of stability in the interval under consideration [Zakharov, 1981, 1990]. It is also characterized by the widest territory of application: from the Arctic regions to the North Caucasus and Transcaucasia, the Russian Far East and northeastern China, and Northern California and Mexico [Urman et al., 2014; Zakharov and Rogov, 2020; Zakharov, 2022]. In this regard, each studied section is supplemented by buchiazones along with information on their subdivision based on ammonites.

Boyarka composite section. The Lower Cretaceous deposits on the Boyarka River are represented by numerous bedrock outcrops along the river channel, and only the outcrops of the middle part of the Ryazanian Hectoroceras kochi Zone, in which belemnite rostra are practically absent, exist on its tributaries: the Pravaya Boyarka (outcrop 25) and Levaya Boyarka (outcrop 15) rivers [Zakharov, 1970; Gol’bert, 1981; Igolnikov, 2019] (Fig. 1b, c). The total thickness of the Neocomian formations directly on the Boyarka River is over 260 m (Fig. 3). The Ryazanian section starts here from the uppermost H. kochi Zone and continuously extends until reaching the Tollia tolli Zone (outcrops 16 and 17) [Alekseev, 1984; Igolnikov, 2019]. After an observation gap (5–6 m), ammonites are represented by a complete zonal succession from Neotollia klimovskiensis regional Zone to Homolsomites bojarkensis Zone inclusive (outcrops 1–14 and 18) [Bogomolov, 1989]. Outcrops in which the same beds are duplicated are well correlated with each other on the basis of lithologic and biostratigraphic features [Zakharov, 1970; Gol’bert, 1981]; so, a composite section can be confidently compiled [Bogomolov, 1989] (Fig. 3). The subdivision of the Boyarka section based on buchias is somewhat less detailed, and the uncertainty intervals are wider. However, according to [Zakharov, 1981, 1990], the entire succession of buchiazones of the Boreal standard is recorded – from the uppermost Buchia unschensis Zone to the lowermost B. crassicollis Zone. The least fragmentary magnetic polar characteristics are published for the lower [Guzhikov and Baraboshkin, 2008] and upper [Pospelova and Larionova, 1971] parts of the section. Ryazanian deposits are represented by the Bukatyi clay–silty Formation, above which the Boyarka essentially sandy Formation is exposed [Gol’bert, 1981]. The Ryazanian, Valanginian, and early Hauterivian sediments were deposited under middle sublittoral, shallow-marine, and lagoon–marine conditions, respectively [Zakharov and Yudovnyi, 1974]. The Boyarka Formation contains erosion traces (sedimentation gaps) within and/or at the boundaries of ammonite zones. The comparatively thin (12.3 m) interval between the Euryptychites quadrifidus and E. astieriptychus undifferentiated zones (owing to the absence of characteristic taxa, the complete loss of the latter is not excluded) is probably associated with one of these erosions. In general, ammonites characterize almost the entire section, but the highest taxonomic diversity is observed in bivalves and foraminifers. Gastropods, scaphopods, brachiopods, serpulids, and others are identified, mainly in the Boyarka Formation; also, numerous ichnofossils are determined in the H. bojarkensis Zone [Gol’bert, 1981; Bogomolov, 1989]. Belemnites are abundant everywhere with the exception of separate small intervals. This is the reason for a relatively detailed chemostratigraphic sampling of the composite section, supplementing the data earlier obtained during the study of some outcrops [Nunn, 2007; Nunn et al., 2010] (Fig. 3).

Nordvik composite section. The studied material comes from the eastern coast of the Nordvik Peninsula, where the Lower Cretaceous deposits are exposed in a series of well-correlated outcrops (outcrops 31–33 and 35 in Fig. 1e). Above the Volgian rocks, there is a continuous succession from Ryazanian Praetollia maynci Zone to Valanginian Siberites ramulicosta Zone [Zakharov et al., 1983; Bogomolov, 1989; Nikitenko et al., 2013; Igolnikov, 2019]. The total thickness of these deposits is about 159 m (Fig. 4). New data [Igolnikov et al., 2016] suggest that the base of the Hectoroceras kochi Zone should be lowered to the level of the top of the M17n magnetozone. Previously, it was located at the level of the middle part of M16r [Bragin et al., 2013; Shurygin and Dzyuba, 2015]. Buchiazones in the studied interval are represented by a standard succession from the uppermost Buchia unschensis Zone to the lowermost B. sublaevis Zone with insignificant intervals of absence of finds of characteristic taxa [Zakharov, 1981, 1990; Zakharov et al., 1983]. The Ryazanian interval mainly corresponds to the lower subformation of the Paksa Formation (Figs. 2, 4). It is composed of predominantly argillite-like clays enriched in organic matter, which are relatively deep-sea (lower sublittoral conditions) in its lower part (beds 18–30) and shallow (middle sublittoral conditions) in its remaining part (beds 31–44) [Nikitenko et al., 2013]. The upper subformation of the Paksa Formation is exposed higher in the section and is composed of argillite-like siltstones and siltstones the material sedimentation of which occurred exclusively in the middle sublittoral settings. The Paksa Formation contains rich complexes of remains of organisms: bivalves, ammonites, fish, crustaceans, foraminifers, and palynoflora (the most abundant within the lower subformation). Belemnite rostra are relatively common in the lowermost Ryazanian and in the Ryazanian–Valanginian transitional deposits; they are rare in other intervals of the section, but are distributed more or less evenly. The northwestern part of the peninsula reveals a higher part of the section – the Homolsomites bojarkensis Zone [Zakharov et al., 1983], but the rarity and poor preservation of rostra have prevented using this material for research.

Anabar composite section. Numerous Valanginian outcrops are concentrated in the lower reaches of the Anabar River, in the area of Yuryung-Khaya settlement (Fig. 1d). The key outcrops are 8, 3, 4, 5, 5a, and 9, which form a composite section with a few observation gaps [Sanin, 1979; Gol’bert, 1981] (Fig. 5). The total thickness of the Valanginian interval together with observation gaps is more than 220 m. Outcrops 1 and 8 also represent the uppermost Ryazanian, but belemnites are not identified there, apparently, owing to the accumulation of this stratum in the coastal settings under water freshening conditions, which is reflected in the composition of bivalve assemblages [Sanin, 1979]. In the zonal subdivision of the Anabar section, both ammonites (from Neotollia klimovskiensis Zone to Dichotomites bidichotomus Zone) and buchias (from upper Buchia inflata Zone to B. sublaevis Zone) are characterized by long uncertainty intervals: Information about the ammonite and buchia zones in the present work is based on the data from [Gol’bert, 1981; Zakharov, 1981; Bogomolov, 1989] with the fixation of gaps in the finds of index species or representatives of the zonal complex. According to paleontological data, the lower/upper Valanginian boundary is located within the 5- to 6-m interval of the section between the last finds of the Buchia keyserlingi bivalves (in the complete absence of B. sublaevis) in bed 5 of outcrop 5 [Zakharov, 1981] and the first finds of the upper Valanginian ammonites of the Dichotomites bidichotomus Zone in bed 8 of outcrop 5, including Polyptychites сf. keyserlingi (=Polyptychites сf. polyptychus, according to [Bogomolov, 1989]), together with Neocraspedites kotschetkovi or similar forms [Gol’bert, 1981]. The studied part of the Valanginian section is represented by the Kharabyl Formation (Figs. 2, 5), characterized mainly by siltstones that are clayey at the bottom and more sandy at the top [Gol’bert, 1981; Nikitenko et al., 2013]. Sedimentation occurred mainly in the upper sublittoral zone with a single transition interval (members 7–10) to the inner part of the middle sublittoral zone [Nikitenko et al., 2013]. The lower Valanginian deposits contain rich fossil assemblages: Cephalopods, bivalves, gastropods, scaphopods, foraminifers, and ichnofossils are frequent; brachiopods, serpulids, and crustaceans are rare [Gol’bert, 1981; Nikitenko et al., 2013]. Higher in the section, in the transition zone to the Tigyan sandy Formation, fossils are extremely rare, which is associated with the gradual weakening of marine sedimentation and the transition to continental sedimentation in the Anabar paleoaquatic area during the late Valanginian [Sanin, 1979]. Belemnites are predominantly identified in the lower 75-m part of the section and the boundary layers of the lower and upper Valanginian, which is reflected in the chemostratigraphic sampling of this location.

The belemnite rostra involved in the isotope–geochemical studies are collected by O.V. Shenfil’ in all the studied sections (1985, 1988, and 1989) and by A.S. Alifirov on the Nordvik Peninsula (2009) during field work performed by the staff of the Trofimuk Institute of Petroleum Geology and Geophysics, Siberian Branch of the Russian Academy of Sciences (IPGG SB RAS, former Institute of Geology and Geophysics, Siberian Branch of the USSR Academy of Sciences) (Novosibirsk), and come from outcrops 1, 3, 4, 7–9, 11, 12, 14, 16, and 17 along the Boyarka River (43 specimens); outcrops 31a, 33, and 35 on the Nordvik Peninsula (30 specimens); and outcrops 1, 3, 5, and 8 along the Anabar River (29 specimens). Note that outcrop 31a is an isolated exposure (up to 180 m in length and 2.25 m in height) of Hectoroceras kochi Zone about 4.8 m in thickness, which is a northwestward continuation of outcrop 31 (Fig. 1). A total of 102 samples are studied, generally characterizing the interval between the Hectoroceras kochi and Homolsomites bojarkensis zones. Previously, samples for studies of the carbon and oxygen isotope composition were collected on the Boyarka River from outcrops 16, 17, and 13 (=beds 18–25, Fig. 3); outcrop 14 (=beds 23–33); and outcrops 1–4 (=beds 33–69) [Nunn, 2007; Nunn et al., 2010]. Only δ¹³C and δ¹⁸O characteristics were also reported for the Upper Jurassic–lowermost Lower Cretaceous of the Nordvik Peninsula, for which samples from outcrops 31 and 32 were studied [Žák et al., 2011; Dzyuba et al., 2013]. Belemnites in these sections are represented by the Boreal family Cylindroteuthididae [Dzyuba, 2012; Efremenko, 2023a,b] (Supplementary Materials, Tables S1 and S2). These are the inhabitants of shelf seas. The study involved rostra belonging to all paleoecologic groups of the family in the classification according to [Zakharov et al., 2014; Dzyuba et al., 2018], which are, respectively, interpreted as presumably nektobenthic forms (Acroteuthis, Liobelus, Simobelus, and Pachyteuthis acuta), moderately active swimmers probably preferring the bottom water layers (Arctoteuthis, Boreioteuthis, and Cylindroteuthis luljensis), and active swimmers (Cylindroteuthis knoxvillensis and Lagonibelus sp.) having an elongated, laterally compressed rostrum with a poorly developed ventral groove. The geographic distribution of a number of species was not limited to the Arctic seas, but also extended to the low-Boreal Pacific (Arctoteuthis porrectiformis, A. tehamaensis, and Cylindroteuthis knoxvillensis) or European (Acroteuthis arctica, A. explanatoides, Boreioteuthis hauthali, Liobelus acrei, and, according to unpublished data, Simobelus curvulus) waters [Dzyuba, 2012; Efremenko and Dzyuba, 2021] and even Tethyan waters (Liobelus acrei) [Mutterlose et al., 2022].

Isotope–geochemical studies were carried out using the carbonate rostra of belemnites. Preference was given to rostra characterized by a diameter range of 10–20 mm and lack of traces of abrasion or erosion on their surface. The alveolar and apical parts were excluded; the outer layer of the rostrum, the axial line, and the luminescent zones were removed using an engraver and diamond tips. The rostrum fragments preliminarily studied for the preservation of the microstructure under a Stemi 508 microscope were washed in distilled water, dried, crushed, and homogenized in an agate mortar. Sampling by drilling was not carried out. Separate portions of the same carbonate powder sample were used for chemical, δ¹³С, δ¹⁸О, and ⁸⁷Sr/⁸⁶Sr analyses. The procedure for monitoring visual, microscopic, and geochemical criteria for sample suitability, used to obtain high-quality material for the analysis of primary C, O, and Sr isotope systems, is considered in more detail in [Dzyuba et al., 2013, 2023; Kuznetsov et al., 2017, 2018]. Additionally, following [Ullmann and Korte, 2015], the correlation between the Mn/Ca, Fe/Ca, and Sr/Ca ratios and δ¹⁸O values was checked.

The elemental composition (Sr, Mg, Al, Ca, Mn, and Fe) was analyzed using a Finnigan Element II inductively coupled plasma mass spectrometer (ICP MS) in an acid extract of carbonate rocks obtained by dissolving the carbonate powder in 1 N hydrochloric acid for 48 h. The analytical error is less than 10% at a concentration of at least three times higher than the detection limit.

The stable isotope composition of carbon and oxygen in carbonate material was analyzed using a Finnigan MAT-253 IRMS equipped with GasBench-II, which transfers the analyzed gas to the mass spectrometer in a 6.0 helium flow. The accuracy in determining the carbon and oxygen stable isotope composition was controlled by measuring the δ¹³С_VPDB and δ¹⁸О_VPDB values in the NBS-19 international standard. The use of the two international standards, NBS-18 and NBS-19, yields close δ¹³С values, but, in the case of δ¹⁸О, the discrepancy with the values measured using the single NBS-19 standard varies from 0 to 1.9‰ (with an average deviation of 0.7‰). The choice of NBS-19 is due to the need to generalize the new δ¹⁸О data with those previously published for the Boyarka and Nordvik sections [Nunn, 2007; Dzyuba et al., 2013]. The error in the reproducibility of measurements is mainly <0.1‰ for δ¹³С and <0.2‰ for δ¹⁸О. The elemental ICP MS and C and O isotope analyses were performed at the Analytical Center for Multi-Elemental and Isotope Research, Siberian Branch of the Russian Academy of Sciences (Novosibirsk). The sample preparation and measurement and data interpretation methods were adapted within the framework of the state assignment (No. 122041400171-5) of V.S. Sobolev Institute of Geology and Mineralogy, Siberian Branch of the Russian Academy of Sciences (Novosibirsk).

The strontium isotopic composition is analyzed after the carbonate material is preliminarily leached in 0.1 M HCl [Kuznetsov et al., 2019]. The Sr isotopic composition is measured using a Triton TI multicollector mass spectrometer at the AIRES Analytical Center based on the Institute of Precambrian Geology and Geochronology, Russian Academy of Sciences (Saint Petersburg). The reproducibility of parallel measurements of ⁸⁷Sr/⁸⁶Sr of the NIST SRM-987 and EN-1 standards is 0.710277 ± 0.000004 (2σ; n = 25) and 0.709206 ± 0.000007 (2σ; n = 18), respectively. The resulting values are adjusted (normalized) to a value of 0.710248 for SRM-987, as recommended for the SIS [Gradstein et al., 2020].

Preservation of carbonate material. As the diagenesis stage is often associated with the input of Fe and Mn into the carbonate system and the depletion with Sr [Brand and Veizer, 1980], special attention has traditionally been paid to the content of these elements in the belemnite carbonate material. In many studies, the concentrations equal to Fe < 200 ppm, Mn < 100 ppm, and Sr > 800 ppm are considered to be characteristic of well-preserved rostra, including the Late Jurassic and Early Cretaceous belemnites [Price et al., 2000, 2018; Gröcke et al., 2003; Rud’ko et al., 2017]. All these criteria are met by 92 of 102 studied samples with Mn/Sr ≤ 0.07, Fe/Sr ≤ 0.15, an average Fe content of no more than 24 ppm in them, Mn = 14 ppm, and Sr = 1243 ppm (Suppl. Mat., Table S1). Moreover, 90% of these 92 samples have concentrations equal to Fe < 150, Mn < 50 ppm, and Sr > 1000 ppm, which correspond to stricter criteria for the preservation of rostrum material according to these parameters [McArthur et al., 2007; Dzyuba et al., 2013, 2023; Meissner et al., 2015; Möller et al., 2015; Kuznetsov et al., 2017]. The remaining 10% are characterized by either slightly elevated Fe contents (154–182 ppm) or Mn contents (54–83 ppm) or slightly lower Sr contents (881–990 ppm), with no overlaps in these parameters in the same samples, so they are not excluded from further consideration.

Ten samples collected within the framework of this study (six from the Nordvik Peninsula and four from the Boyarka River) do not meet the above criteria and are considered to have altered during diagenesis (Suppl. Mat., Table S2). In some of them, the values of the ratios Mn/Sr > 0.2 and Fe/Sr > 1.7 are obtained, which quite probably indicates the postsedimentary impact of meteoric waters [Brand and Veizer, 1980; Kuznetsov et al., 2017; Rud’ko et al., 2017; Zakharov et al., 2022, 2024]. The presence of Al is detected only in three of all the samples studied. Among them, only sample H-17 seems to be characterized by an increased Fe concentration (2507 ppm). This, along with an increased Mn content (111 ppm), indicates contamination of this sample not only owing to contamination of the analyzed solution with iron from the terrigenous component of the sample, but also with material from the recrystallized areas of the rostrum.

To further identify possible postsedimentary alterations, δ¹⁸O–δ¹³C, Fe–Mn, Sr–Mn, Sr–Fe, Mn/Ca–δ¹⁸O, Fe/Ca–δ¹⁸O, and Sr/Ca–δ¹⁸O binary diagrams are constructed for all 92 samples meeting the selection criteria, and the correlation coefficients (R²) between the values of the same parameters are calculated (Suppl. Mat., Table S3). In the case of secondary alteration of carbonate material in belemnite rostra, one should expect positive correlation between Fe and Mn, as well as the values of δ¹⁸O and δ¹³C; negative correlation between Sr and Mn, as well as between Sr and Fe; the presence of correlation between Mn/Ca and δ¹⁸O, Fe/Ca and δ¹⁸O, and Sr/Ca and δ¹⁸O values [Brand and Veizer, 1980; Ullmann and Korte, 2015]. However, a linear dependence between the compared measured parameters is practically absent or negligible. The calculated values of R² generally do not exceed 0.16, and the Anabar samples are associated with higher values of this coefficient for Sr/Ca and δ¹⁸O (R² = 0.26) and Fe and Mn (R² = 0.4). As a result, all 92 samples are found to be suitable for chemostratigraphic constructions.

δ¹³C, δ¹⁸O, and⁸⁷Sr/⁸⁶Sr variations. The δ¹³C values obtained in the present studies vary from –0.2 to 3.3‰ for the Boyarka section, from –0.8 to 1.5‰ for the Nordvik section, and from –0.3 to 3.9‰ for the Anabar section. The minimum values of this parameter are recorded in the Surites analogus Zone (from –0.8 to 2.1‰, average value 0.4‰) and Tollia tolli Zone (from –0.6 to 0.5‰, average value 0.1‰), and the maximum values are recorded in the Dichotomites bidichotomus Zone (from 1.8 to 3.7‰, average value 2.6‰) and the lowermost Homolsomites bojarkensis Zone (from 1.4 to 3.3‰, average value 2.2‰). Within the Ryazanian interval, the average δ¹³C value is 0.9‰ for the Boyarka section and 0.3‰ for the Nordvik section. Some positive deviation from these values is observed in the lowermost Hectoroceras kochi Zone and the transition interval between the S. analogus and Bojarkia mesezhnikowi zones (Figs. 3, 4). A sharp increase in the δ¹³C values by 2.3–2.4‰, based on the Boyarka section data, is recorded from the uppermost Siberites ramulicosta Subzone of the Valanginian interval (Fig. 3). The same stratigraphic interval, apparently, accounts for a similar shift (on average by 2.3‰) in the δ¹³C curve in the Anabar section, which is somewhat weakly characterized by ammonites (Fig. 5). The Valanginian interval preceding the positive excursion of δ¹³C is characterized by the following average δ¹³C values: 0.9‰ in the Boyarka section, 0.7‰ in the Nordvik section, and 1.0‰ in the Anabar section.

The δ¹⁸O values are in the following ranges: from –2.5 to 2.5‰ in the Boyarka section, from –1.1 to 1.2‰ in the Nordvik section, and from –2.2 to 1.0‰ in the Anabar section. The lowest values are obtained for the Neotollia klimovskiensis Zone (from –2.5 to 1.1‰, average –0.7‰) and the Siberites ramulicosta Zone (from –2.2 to –0.2‰, average –1.0‰), and the highest values are obtained in the Hectoroceras kochi Zone (–0.7 to 1.6‰, average 0.1‰) and the uppermost Tollia tolli Zone (0.3 to 0.7‰, average 0.5‰). The average δ¹⁸O values established for the remaining ammonite zones vary from –0.5 to 0‰. In general, the average values of δ¹⁸O obtained for the studied part of the Ryazanian interval (–0.2‰ in the Boyarka section and 0.3‰ in the Nordvik section) are higher than those obtained for the rest of the Lower Cretaceous section interval (–0.7‰ in the Boyarka section, –0.5‰ in the Nordvik section, and –0.6‰ in the Anabar section). Taking into account the sampling features of each section, the main trends in the δ¹⁸О curves are generally maintained for the entire region under study. Synchronous negative shifts in the δ¹³C and δ¹⁸O values, which could indicate postsedimentary changes in the belemnite carbonate material [Ullmann and Korte, 2015], are not observed (Figs. 3–5).

The ⁸⁷Sr/⁸⁶Sr ratio increases upward in the section (Figs. 3–5). The ⁸⁷Sr/⁸⁶Sr values vary within a range of 0.707216–0.707257 (with a single deviation up to 0.707273) in the Hectoroceras kochi and Surites analogus zones, 0.707256–0.707279 in the Bojarkia mesezhnikowi Zone, 0.707260–0.707289 in the Tollia tolli Zone, 0.707281–0.707330 in the Neotollia klimovskiensis Zone, 0.707316–0.707343 (with a single deviation up to 0.707279) in the Euryptychites quadrifidus Zone, 0.707320–0.707347 in the Euryptychites astieriptychus Zone, 0.707321–0.707365 in the Siberites ramulicosta Zone, 0.707350–0.707366 in the Dichotomites bidichotomus Zone, and 0.707369–0.707416 in the Homolsomites bojarkensis Zone. In each section studied, the Sr isotope curve tends to a gradual increase in values without sharp jumps, thereby generally agreeing well with the sedimentation rate in the paleobasins. At the same time, the interval between the H. kochi and B. mesezhnikowi zones is characterized by a stable perturbation (increase–decrease–increase) of the ⁸⁷Sr/⁸⁶Sr ratio of yet unclear nature; it is most pronounced in the Nordvik section, so that the differences in the ⁸⁷Sr/⁸⁶Sr values between the H. kochi and S. analogus zones are negligible. Yet, the shift in the ⁸⁷Sr/⁸⁶Sr ranges is quite explicit between the lowermost H. kochi Zone (0.707216–0.707245) and the uppermost part of the same zone (0.707234–0.707257 and up to 0.707273). There is a strong overlap in the ⁸⁷Sr/⁸⁶Sr ranges in the E. quadrifidus and E. astieriptychus zones as well as in the Polyptychites beani Subzone and the D. bidichotomus Zone.

In total, the studied Lower Cretaceous interval in northern East Siberia for the three sections is characterized fairly uniformly by the values of δ¹³C, δ¹⁸O, and ⁸⁷Sr/⁸⁶Sr, with the exception of gaps in the isotopic record occurring in the transitional interval between the Bojarkia mesezhnikowi and Tollia tolli zones as well as at the uppermost Dichotomites bidichotomus Zone.

In the Cretaceous history of the global carbon cycle associated with СО₂ exchange between the atmosphere and the ocean, one of the most famous events well recognized both by the isotope composition of carbon in carbonates (belemnite rostra and sedimentary carbonate rocks) and in fossil organic matter is the Weissert event [Erba et al., 2004; Price et al., 2016, 2018; Galloway et al., 2020; Gradstein et al., 2020; Jelby et al., 2020]. Three main phases of this event have been described on the basis of the study of bulk carbonate rocks of a number of Tethyan sections, and a time interval has been determined for each of them: (1) rapid increase in the δ¹³С values (early–late Valanginian transition); (2) stable δ¹³С values (late Valanginian, partially); and (3) smooth decrease in the δ¹³С values to the pre-excursion state (the end of the late Valanginian and the early Hauterivian) [Martinez et al., 2015]. It is noteworthy that the δ¹³С values reach a peak and usually remain at approximately the same level within the Verrucosum–Peregrinus chrons [Martinez et al., 2015; Price et al., 2016]. In some Boreal organic carbon isotope records, the signal stabilizes at near-peak values for longer time periods, notably until the end of the Hauterivian or early Barremian in the Lower Cretaceous sections of Svalbard [Jelby et al., 2020]. However, this is not a general situation based on the δ¹³С_org values from other localities [Galloway et al., 2020; Jelby et al., 2020] as well as δ¹³С data published for belemnites from Boreal sections. Thus, the peak value interval is quite limited in the Boreal δ¹³С curves based on the study of belemnites from northwestern Europe (from the uppermost Prodichotomites hollwedensis Zone to the lowermost Stoicoceras tuberculata Zone of the ammonite scale) [Meissner et al., 2015], Greenland (from the uppermost BC4 Zone to the lowermost BC5 Zone of the nannofossil scale) [Möller et al., 2015], and northern Russia, based on a combination of data from the Izhma and Boyarka sections (the Polyptychites polyptychus Zone and, at least, the lowermost Dichotomites bidichotomus Zone of the ammonite scale) [Nunn et al., 2010]. At the time of the Weissert event, the Boreal seas were connected to the carbon cycle in lower-latitude seas owing to a series of open straits: Water exchange between the northern and southern seas was facilitated mainly by the Norwegian–Greenland Seaway (Viking Corridor) in the Boreal–Atlantic sector and by the Anyui, Alaska, and Dawson straits in the Boreal–Pacific sector [Mutterlose et al., 2003, 2022; Baraboshkin et al., 2007].

Based on a generalization of previously obtained [Nunn et al., 2010] and new δ¹³С data for the Boyarka section, all three phases of the Weissert event are established (Fig. 3). Phase 1 is noted for the Siberites ramulicosta Zone near the boundary between the S. ramulicosta and Polyptychites beani subzones. Phase 2 comprises the P. beani Subzone and is reliably recorded for the Polyptychites triplodiptychus and Dichotomites bidichotomoides subzones of the Dichotomites bidichotomus Zone. Phase 3 presumably covers the entire beds with Neocraspedites kotschetkovi (this interval contains only one δ¹³С value [Nunn et al., 2010]) and clearly the Homolsomites bojarkensis Zone (within the interval studied). As the peak δ¹³С values in the Tethyan sections are usually reached starting from the basal part of the upper Valanginian (from the Verrucosum Chronozone) and they are previously clearly established on the Izhma River in northern Russia, starting from the Polyptychites polyptychus Zone [Nunn et al., 2010], which is a biostratigraphic analog of the Siberian P. beani Subzone [Baraboshkin, 2004], the position of the base of the upper Valanginian in the sections of northern Russia can be considered quite definite. The new data obtained here do not contradict this. When it comes to the Anabar section, the scarce material allows only for a conditional distinction between phase 1 of the Weissert event and the moment at which phase 2 begins (Fig. 5).

The nature of the Weissert event is still debatable. The proposed relation with early phases of volcanic activity in the Paraná–Etendeka igneous province, which might have led to a significant increase in atmospheric CO₂ [Erba et al., 2004], is contrary to numerous paleoclimatic reconstructions [Price and Mutterlose, 2004; McArthur et al., 2007; Meissner et al., 2015; Möller et al., 2015; Price et al., 2016, 2018]. According to [Price et al., 2018], the Weissert event only coincides with trap eruptions in the Paraná–Etendeka province. It is associated with some climate cooling, especially clearly established for its main phase by data from the Tethyan sections using the δ¹⁸O and Mg/Ca paleothermometry methods. This is also evidenced by glendonites [Rogov et al., 2023]. As suggested in [Price et al., 2018], the cooling reflects a significant reduction in atmospheric CO₂, implying the presence of an alternative factor contributing to increased ocean bioproductivity. For example, tectonic activity associated with the breakup of Gondwana might have introduced hydrothermal biolimiting elements from the spreading ridge zone and triggered a global accumulation of nutrients [Erba et al., 2004]. The global sea level rise recorded for the late Valanginian–Hauterivian [Haq, 2014] might also have contributed to increased runoff and nutrient fluxes into the oceans. According to the data obtained from the Boyarka and Anabar sections, the increase in the δ¹⁸O values was observed during the transition from phase 1 of the Weissert event to its subsequent phases (Figs. 3, 5). However, this increase in the δ¹⁸O values, which marks a decrease in paleotemperatures, is not quite strong as, incidentally, in West Siberia [Price and Mutterlose, 2004].

The first half of the Early Cretaceous is characterized by a global increase in the ⁸⁷Sr/⁸⁶Sr ratio in the ocean [McArthur and Howarth, 2024]. It is generally assumed that the Sr continental flow is the main factor determining the change in the ⁸⁷Sr/⁸⁶Sr ratio in the seas at the start of the Early Cretaceous. At the same time, there are many unanswered questions in matters of spreading rates or high hydrothermal activity episodes not directly related to the formation rate of oceanic crust (ridge “jumps” or changes in ridge orientation) as well as in assessment of the balance of continental weathering and Sr hydrothermal flows [Price et al., 2016; Gradstein et al., 2020]. The published Sr isotope values increase from 0.707191–0.707192 near the base of the Berriasian to 0.707219 at the base of the Ryazanian ([Kuznetsov et al., 2017], corrected to the standard SRM-987 = 0.710248). Then the ⁸⁷Sr/⁸⁶Sr ratio continues to increase from 0.707237 (on average) at the base of the upper Berriasian to 0.707289 at the base of the Valanginian, to 0.707352 at the base of the upper Valanginian, and to 0.707383 at the base of the Hauterivian [McArthur et al., 2007; McArthur and Howarth, 2024]. The values increase monotonously, with the exception of the late Valanginian plateau, which is even more pronounced when we use the absolute age scale according to [Gradstein et al., 2020] (Fig. 6).

The results obtained for the Boyarka, Nordvik, and Anabar sections of the Lower Cretaceous are in good agreement with the literature data both on the values of ⁸⁷Sr/⁸⁶Sr isotope ratios in belemnites and on the general trend of their change (Fig. 6). This suggests a number of important conclusions. The ⁸⁷Sr/⁸⁶Sr values obtained in the Boyarka and Nordvik sections in combination with the available paleomagnetic data for the same sections [Guzhikov and Baraboshkin, 2008; Bragin et al., 2013] and taking into account the refined position of the base of the Ryazanian Hectoroceras kochi Zone relative to the magnetozones on the Nordvik Peninsula [Igolnikov et al., 2016] indicate that the base of the upper Berriasian corresponds to the middle part of the H. kochi Zone. As previously revealed by analyzing the biostratigraphic criteria of the Boreal–Tethyan correlation, the base of the H. kochi Zone cannot be older than the base of the Dalmasi Subzone of the Occitanica Chronozone. This, however, does not exclude the late Berriasian age of some part of this zone, especially since there is evidence of the correspondence of the Buchia okensis Zone, which is related to the uppermost H. kochi ammonite Zone, to a fragment of the Paramimounum Subzone of the Boissieri Chronozone [Bragin et al., 2013].

As suggested by the ⁸⁷Sr/⁸⁶Sr data obtained, the base of the Valanginian in the Siberian sections should be located somewhat higher than the base of the Neotollia klimovskiensis Zone, the basal beds of which on the Nordvik Peninsula have the characteristics (0.707281–0.707288) of the Berriasian Alpillensis Chronozone: from 0.707263 to 0.707289, as reported in [McArthur et al., 2007; McArthur and Howarth, 2024]. It seems that the age reinterpretation of the boundary between Tollia tolli and N. klimovskiensis zones implies the necessity of simultaneous age reinterpretation of the top of the Ryazanian as a whole. This is evidenced by the noticeable deviation of ⁸⁷Sr/⁸⁶Sr toward lower values obtained for the uppermost Ryazanian in northwestern Europe [Jones et al., 1994] and on the Russian Plate [Gröcke et al., 2003] as compared with the uppermost Berriasian data [McArthur et al., 2007], as reported in [Kuznetsov et al., 2017]. In the case of the Boyarka River, the basal layers of the N. klimovskiensis Zone, apparently, fall on the observation gap between Tollia tolli Zone and N. klimovskiensis Zone. An interval of the N. klimovskiensis Zone available for observation has Sr isotope characteristics (0.707299–0.707330) of the Valanginian Pertransiens Chronozone (0.707289–0.707330, as reported in [McArthur et al., 2007; McArthur and Howarth, 2024]). The basal layers of the N. klimovskiensis Zone are not represented on the Anabar River either: The ⁸⁷Sr/⁸⁶Sr values at the base of the section studied here are basal Valanginian (0.707289–0.707293).

The interval corresponding to the Euryptychites quadrifidus Zone, the Euryptychites astieriptychus Zone, and the Siberites ramulicosta Subzone in the Siberian ammonite zonal standard, most likely, corresponds to the Neocomiensiformis–Inostranzewi chronozones of the lower Valanginian in terms of the ⁸⁷Sr/⁸⁶Sr ratio (Fig. 6). Further comparisons within the Valanginian are complicated by the insufficient amount of the ⁸⁷Sr/⁸⁶Sr data. The Homolsomites bojarkensis Zone crowns the studied stratigraphic interval. Based on the ⁸⁷Sr/⁸⁶Sr data obtained for the lowermost part of this zone on the Boyarka River (0.707369–0.707416), the H. bojarkensis Zone is, undoubtedly, early Hauterivian in age and clearly has the best age overlap with the Radiatus Chronozone. The last one is characterized by the ⁸⁷Sr/⁸⁶Sr values of 0.707383 to 0.707398, as reported in [McArthur et al., 2007]. It is possible that the base of the Hauterivian falls on beds with Neocraspedites kotschetkovi, in which, according to [Bogomolov, 1989], the index species of the overlying H. bojarkensis Zone first appears. This is indicated by the paleomagnetic characteristics of the uppermost Boyarka section [Pospelova and Larionova, 1971], corresponding to the top parts of the beds with N. kotschetkovi and the H. bojarkensis Zone (Fig. 3), taking into account the position of the M10n magnetozone relative to the SMAZ at the present time [Gradstein et al., 2020].

The results of the study of the belemnite carbon material from the Boyarka, Nordvik, and Anabar sections are applied to obtain a comprehensive C, O, and Sr isotope characteristic of the greater part of the Neocomian rocks of Arctic Siberia for the first time. The chemostratigraphic data serve as the basis for an updated correlation scheme of the Boreal (Siberian) zonal standard with the SMAZ zones in the upper Berriasian–lowermost Hauterivian interval. In particular, the analysis of the resulting values of ⁸⁷Sr/⁸⁶Sr in combination with bio- and magnetostratigraphic data makes it possible to correct the age of a number of Siberian ammonite zones and subzones and their separate intervals. This clarifies the entire “package” of parallel zonal scales (based on different fauna and flora groups) of the Boreal zonal standard. The results of the studies indicate that the upper Berriasian in the Siberian sections begins approximately from the middle of the Hectoroceras kochi Zone and the Valanginian starts slightly above the base of the Neotollia klimovskiensis Zone. The base of the Hauterivian can be located in the interval of beds with Neocraspedites kotschetkovi. The age of the overlying Homolsomites bojarkensis Zone is reliably identified as early Hauterivian. Moreover, it is shown that the upper boundary of the Ryazanian interval (the top of the Tollia tolli Zone or its Boreal stratigraphic analogs) is located below the level of the upper boundary of the Berriasian interval (the top of the Alpillensis Chronozone), so that it does not coincide with the level corresponding to the lower boundary of the Valanginian interval (the base of the Pertransiens Chronozone).

At the current stage of the study, the interval from the Polyptychites beani Subzone to the Dichotomites bidichotomus Zone in Arctic Siberia has been studied by the SIS method in the least detail. Nevertheless, the available ⁸⁷Sr/⁸⁶Sr data, including the results obtained for the underlying and overlying deposits, combined with the δ¹³С variation trend, suggest that this interval is late Valanginian (probably excluding the beds with Neocraspedites kotschetkovi or their top). For the first time, the resulting δ¹³С variation allows for the recognition of separate phases of the Weissert event for northern East Siberia. The calculated amplitude of the δ¹³С shift at the beginning of this event, corresponding to the first of the largest positive δ¹³С excursions in the Cretaceous, is about 2.3‰. Based on the δ¹⁸О data, it is revealed that the global climate cooling accompanying the Weissert event in Siberia is negligible.

The authors are grateful to O.V. Shenfil’ and A.S. Alifirov, who provided the belemnite rostra collections along with field diaries (1985, 1988, and 1989) or extracts (2009); to E.Yu. Baraboshkin and M.A. Rogov for discussing the work and providing valuable recommendations; to I.V. Nikolaeva, an employee of the Analytical Center for Multi-Elemental and Isotope Research; and to T.S. Zaitseva, head of the AIRES Analytical Center.

The study was supported by the Russian Science Foundation, grant No. 22-17-00228 (https://rscf.ru/project/22-17-00228/), at the IPGG SB RAS. Previously, as part of the work on the state assignment for the IPGG SB RAS, detailed ideas about the structure of the studied sections had been formed (project FWZZ-2022-0004).