1. Introduction
Striking stratigraphic fluctuations in the carbon isotopes of carbonates, and marine and terrestrial organic matter, have provided an essential time-control reference for the Upper Cretaceous marine strata. The chemostratigraphic framework is critical as a basis for the clarification of the spatio-temporal climatic and biotic evolution on a global scale, thus detailed Upper Cretaceous δ13C stratigraphy has been established intensively for marine carbonates in Europe and North America (e.g. Scholle & Arthur, Reference Scholle and Arthur1980; Gale et al. Reference Gale, Jenkyns, Kennedy and Corfield1993; Jenkyns, Gale & Corfield, Reference Jenkyns, Gale and Corfield1994; Mitchell, Paul & Gale, Reference Mitchell, Paul, Gale, Howell and Aitken1996; Voigt & Hilbrecht, Reference Voigt and Hilbrecht1997; Stoll & Schrag, Reference Stoll and Schrag2000; Tsikos et al. Reference Tsikos, Jenkyns, Walsworth-Bell, Petrizzo, Forster, Kolonic, Erba, Premoli Silva, Baas, Wagner and Sinninghe Damsté2004; Bowman & Bralower, Reference Bowman and Bralower2005; Jarvis et al. Reference Jarvis, Gale, Jenkyns and Pearce2006). Recently, Jarvis et al. (Reference Jarvis, Gale, Jenkyns and Pearce2006) compiled the δ13C curve for marine carbonates of the Upper Cretaceous sequence in Europe, and documented the use of 72 carbon isotope events for interregional correlation through the Cenomanian–Santonian sequence. More marine and terrestrial δ13C data from different regions should be accumulated and these data correlated with the reference δ13Ccarbonate profile for the establishment of high-resolution event stratigraphy of the Upper Cretaceous sequence.
The forearc siliciclastic sequence of the Aptian–Paleocene Yezo Group is exposed along a 1200 km long belt that extends from southern Hokkaido in Japan, northward to Sakhalin Island in Russia, along the NW Pacific margin (Okada, Reference Okada, Hashimoto and Uyeda1983; Takashima et al. Reference Takashima, Kawabe, Nishi, Moriya, Wani and Ando2004; Shigeta & Maeda, Reference Shigeta, Maeda, Maeda and Shigeta2005). Because of the extensive distribution with abundant marine macro- and microfossils, the Yezo Group has been employed as a chronostratigraphic reference unit for the circum-Pacific region. Over the past decade, various carbon isotope stratigraphies of terrestrial organic matter have been proposed for the Yezo Group (Hasegawa & Saito, Reference Hasegawa and Saito1993; Hasegawa, Reference Hasegawa1997, Reference Hasegawa2003a; Ando et al. Reference Ando, Kakegawa, Takashima and Saito2002, Reference Ando, Kakegawa, Takashima and Saito2003; Hasegawa et al. Reference Hasegawa, Pratt, Maeda, Shigeta, Okamoto, Kase and Uemura2003; Ando & Kakegawa, Reference Ando and Kakegawa2007; Uramoto et al. Reference Uramoto, Fujita, Takahashi and Hirano2007). These studies compared marine and terrestrial carbon isotopes, and established the δ13C stratigraphy by assuming an isotopic linkage between marine and terrestrial carbon reservoirs within the earth surface subsystems of ocean–atmosphere–terrestrial biosphere; thus, time-stratigraphic terrestrial organic carbon isotope fluctuations for Cretaceous time are assumed to reflect δ13C variations in atmospheric CO2 (Hasegawa, Reference Hasegawa1997; Ando et al. Reference Ando, Kakegawa, Takashima and Saito2002; Gröcke et al. Reference Gröcke, Price, Robinson, Baraboshkin, Mutterlose and Ruffell2005). For the Upper Cretaceous sequence of the Yezo Group, several studies have outlined the overall pattern of the δ13C record (e.g. Hasegawa, Reference Hasegawa1997; Hasegawa et al. Reference Hasegawa, Pratt, Maeda, Shigeta, Okamoto, Kase and Uemura2003), and global correlations of the chemostratigraphic events have been documented for the Cenomanian and Turonian sequences (Hasegawa, Reference Hasegawa2003a; Ando & Kakegawa, Reference Ando and Kakegawa2007; Uramoto et al. Reference Uramoto, Fujita, Takahashi and Hirano2007). Further examination of the stratigraphic δ13C record for the Yezo Group will contribute to improvement of the Upper Cretaceous chemostratigraphic framework in the NW Pacific.
The Albian–Campanian sequence of the Cretaceous Yezo Group is widely present throughout the Obira area of Hokkaido (Fig. 1). Several faults recognized in the area record only a minor offset (Fig. 1b), meaning that the complete stratigraphic sequence is preserved. For this reason, many studies have analysed the stratigraphy, palaeontology and geochemistry of the Yezo Group in the Obira area (e.g. Igi et al. Reference Igi, Tanaka, Hata and Sato1958; Tsushima et al. Reference Tsushima, Tanaka, Matsuno and Yamaguchi1958; Tanaka, Reference Tanaka1963; Tanabe et al. Reference Tanabe, Hirano, Matsumoto and Miyata1977; Matsumoto et al. Reference Matsumoto, Muramoto, Hirano and Takahashi1981; Sekine, Takagi & Hirano, Reference Sekine, Takagi and Hirano1985; Hasegawa & Saito, Reference Hasegawa and Saito1993; Hasegawa, Reference Hasegawa2001; Nishi et al. Reference Nishi, Takashima, Hatsugai, Saito, Moriya, Ennyu and Sakai2003; Takashima et al. Reference Takashima, Kawabe, Nishi, Moriya, Wani and Ando2004; Funaki & Hirano, Reference Funaki and Hirano2004; Kaneko & Hirano, Reference Kaneko and Hirano2005; Oizumi et al. Reference Oizumi, Kurihara, Funaki and Hirano2005; Takahashi, Reference Takahashi2005; Nishimura, Maeda & Shigeta, Reference Nishimura, Maeda and Shigeta2006; Uramoto et al. Reference Uramoto, Fujita, Takahashi and Hirano2007). In terms of carbon isotope stratigraphy, Hasegawa & Saito (Reference Hasegawa and Saito1993) presented the first report of the δ13C record around the Cenomanian–Turonian boundary within the Yezo Group, along with the δ13C data for the Oyubari area of central Hokkaido. Uramoto et al. (Reference Uramoto, Fujita, Takahashi and Hirano2007) established global chemostratigraphic correlations for the Cenomanian Stage of the Yezo Group, based on the recently refined macro- and microfossil biostratigraphy of the Upper Cretaceous in the Obira area (Nishi et al. Reference Nishi, Takashima, Hatsugai, Saito, Moriya, Ennyu and Sakai2003; Funaki & Hirano, Reference Funaki and Hirano2004; Oizumi et al. Reference Oizumi, Kurihara, Funaki and Hirano2005). However, stratigraphic δ13C data have yet to be presented and analysed in terms of global chemostratigraphy, for the rocks in the Obira area that overlie the Turonian sequence.
Figure 1. (a) Index map showing the distribution of the Yezo Group (Takashima et al. Reference Takashima, Kawabe, Nishi, Moriya, Wani and Ando2004) and locality of the Obira area in Hokkaido, Japan. (b) Geological map of the Obira area (modified after Funaki & Hirano, Reference Funaki and Hirano2004; Oizumi et al. Reference Oizumi, Kurihara, Funaki and Hirano2005; Uramoto et al. Reference Uramoto, Fujita, Takahashi and Hirano2007). (c) Schematic stratigraphy of the Obira area. Lithostratigraphic divisions are after Funaki & Hirano (Reference Funaki and Hirano2004) and Oizumi et al. (Reference Oizumi, Kurihara, Funaki and Hirano2005). Age assignments are from macro- and microfossil biostratigraphy, and carbon isotope stratigraphy by Tanabe et al. (Reference Tanabe, Hirano, Matsumoto and Miyata1977), Matsumoto et al. (Reference Matsumoto, Muramoto, Hirano and Takahashi1981), Sekine, Takagi & Hirano (Reference Sekine, Takagi and Hirano1985), Hasegawa & Saito (Reference Hasegawa and Saito1993), Nishi et al. (Reference Nishi, Takashima, Hatsugai, Saito, Moriya, Ennyu and Sakai2003), Funaki & Hirano (Reference Funaki and Hirano2004), Oizumi et al. (Reference Oizumi, Kurihara, Funaki and Hirano2005) and Uramoto et al. (Reference Uramoto, Fujita, Takahashi and Hirano2007). Co – Coniacian; Ca – Campanian.
The purpose of this study is to provide a global correlation of the Upper Cretaceous δ13C records between Japan and Europe. We will establish a carbon isotope stratigraphy of sedimentary organic matter for the Upper Cretaceous sequence of the Obira area and then, δ13C patterns are correlated within the Upper Cretaceous marine sequences in NW Pacific and Europe, using the available biostratigraphic framework for the calibration.
2. Geological setting
The middle to upper part of the Yezo Group is widely distributed in the Obira area (Fig. 1b). The sedimentary succession consists of a series of siliciclastic marine sequences up to about 6000 m in maximum thickness. The succession is unconformably overlain by the Tertiary Jugosenzawa Formation.
The lithostratigraphic framework for the Obira area has been presented in previous works (Igi et al. Reference Igi, Tanaka, Hata and Sato1958; Tsushima et al. Reference Tsushima, Tanaka, Matsuno and Yamaguchi1958; Tanaka, Reference Tanaka1963; Funaki & Hirano, Reference Funaki and Hirano2004; Oizumi et al. Reference Oizumi, Kurihara, Funaki and Hirano2005). In this study, we employ the recent lithostratigraphic divisions for the Obira area defined by Funaki & Hirano (Reference Funaki and Hirano2004) and Oizumi et al. (Reference Oizumi, Kurihara, Funaki and Hirano2005).
The studied successions of the Yezo Group consist of the conformable Tenkaritoge, Saku and Haborogawa formations, in ascending stratigraphic order. The Tenkaritoge Formation is mainly represented by dark grey siltstone and occasional intercalated thin sandstone. The Saku Formation consists of alternating beds of sandstone and siltstone. The Haborogawa Formation is mainly characterized by dark grey massive siltstone with intercalated sandstone. Occurrences of thick-bedded sandstone (> 1 m) and slump deposits characterize the lower part of the Haborogawa Formation, making it useful for regional correlations (Kamikinenbetsu Sandstone Member: Funaki & Hirano, Reference Funaki and Hirano2004).
The ages of study units range from Cenomanian to Santonian, based on macro- and microfossil biostratigraphy, and carbon isotope stratigraphy (Fig. 1c). Uramoto et al. (Reference Uramoto, Fujita, Takahashi and Hirano2007) recognized the first occurrence of the Late Cenomanian indicator Inoceramus pictus minus Matsumoto in the basal part of the study succession within the Tenkaritoge Formation. Hasegawa & Saito (Reference Hasegawa and Saito1993) presented the age-diagnostic microfossil biostratigraphy and carbon isotope stratigraphy of terrestrial organic matter within the upper part of the Tenkaritoge Formation, and reported a conspicuous positive δ13C excursion that encompassed the Cenomanian–Turonian boundary. Occurrences of global age-diagnostic macrofossils such as Vascoceras cf. durandi (Thomas & Peron), Collignoniceras woollgari Mantell, Subprionocyclus cf. neptuni (Geinitz), Barroisicers spp. and Didymotis costatus (Fric) characterize the Turonian–Coniacian sequences (Tanabe et al. Reference Tanabe, Hirano, Matsumoto and Miyata1977; Matsumoto et al. Reference Matsumoto, Muramoto, Hirano and Takahashi1981; Funaki & Hirano, Reference Funaki and Hirano2004). Oizumi et al. (Reference Oizumi, Kurihara, Funaki and Hirano2005) reported the first occurrence of the Santonian ammonoid Texanites (Plesiotexanites) kawasakii (Kawada) from the uppermost part of the study succession in the Haborogawa Formation.
3. Materials and Methods
A total of 62 fresh mudstone samples were collected along the Kanajirizawa, Obirashibegawa, Okufutamatazawa and Akanosawa rivers in the Obira area (Fig. 2).
Figure 2. Map showing the sample locations of mudstone in this study.
For the evaluation of organic matter, we conducted total organic carbon (TOC) content analyses and Rock-Eval pyrolysis on five selected samples for the Upper Turonian–Santonian sequence. Previous studies have described the compositional and geochemical characteristics of the kerogen for the Cenomanian–Middle Turonian sequence in the Obira area (Hasegawa & Saito, Reference Hasegawa and Saito1993; Hasegawa, Reference Hasegawa2001; Uramoto et al. Reference Uramoto, Fujita, Takahashi and Hirano2007).
Powdered bulk samples were acidified with 6N solution of HCl for 24 hours to decompose carbonates. The elemental composition of 20–30 mg of each subsample was then analysed using a Yanaco CHN-Corder MT-3 at JAPEX Research Centre (Chiba, central Japan), employing antipyrine (C11H12N2O) as the standard. The elemental composition of each acid-processed subsample was corrected based on the weight-percent of removed carbonates, and then the TOC content of the whole rock was obtained. We used the TOC values in calculating the Hydrogen Index and Oxygen Index for the following Rock-Evalpyrolysis.
Rock-Eval pyrolysis was conducted using a VINCI Technologies model 6 device at JAPEX Research Centre. From each powdered bulk subsample, 100 mg were pyrolysed from 300 to 650°C in a nitrogen atmosphere. The amount of free hydrocarbons (S1), hydrocarbons released during heating (S2), CO2 released during pyrolysis to 390°C (S3), the temperature at which the maximum amount of S2 hydrocarbons were generated (T max), the Hydrogen Index (= 100·S2/TOC), and the Oxygen Index (= 100·S3/TOC) were recorded.
For δ13C analyses of organic matter, acid-processed bulk subsamples were treated with a mixture of benzene and methanol (7:3) to eliminate free hydrocarbons. Analyses of δ13C ratios were performed using a VG Isotech SIRA series II mass spectrometer (precision: ±0.02‰) and a GV Instruments Isoprime EA mass spectrometer (precision: ±0.10‰) at the JAPEX Research Centre. Carbon isotope ratios were expressed as permil deviation from the PeeDee belemnite (PDB) standard.
4. Results
4.a. Type and maturity of sedimentary organic matter
Table 1 shows the results of total organic carbon content analysis and Rock-Eval pyrolysis, and Figure 3 shows a plot of Hydrogen Index versus T max. The Hydrogen Index values vary from 21 to 34 mg HC/g TOC and T max values range from 433 to 438°C (Table 1). These ranges correspond to type III/IV kerogen, and the degree of organic maturity is comparable to vitrinite reflectance values between 0.5 and 0.9%Ro (Fig. 3) (Mukhopadhyay, Reference Mukhopadhyay, Mukhopadhyay and Dow1994; Hunt, Reference Hunt1996).
Table 1. Results of total organic carbon content analysis and Rock-Eval pyrolysis of selected samples
TOC – total organic carbon content; HI – Hydrogen Index; OI – Oxygen Index
Figure 3. A plot of Hydrogen Index versus T max of selected mudstone samples in the Obira area.
4.b. Carbon isotope records
The obtained δ13C values of organic matter (δ13Corg) are −25.3‰ to −21.8‰ for the Upper Cretaceous sequence in the Obira area (Table 2), and we plotted our data with the previously reported δ13C data from the Obira area (Hasegawa & Saito, Reference Hasegawa and Saito1993; Uramoto et al. Reference Uramoto, Fujita, Takahashi and Hirano2007) (Fig. 4). The overall patterns of the Upper Cretaceous δ13Corg profile in the Obira area are independent of lithological variations.
Table 2. Carbon isotope values in study samples
Samples with asterisk were analysed with a GV Instruments Isoprime EA mass spectrometer, and other samples with a VG Isotech SIRA Series II mass spectrometer.
Figure 4. Carbon isotope profile of sedimentary organic matter for the Upper Cretaceous sequence in the Obira area. Macrofossil data are after Tanabe et al. (Reference Tanabe, Hirano, Matsumoto and Miyata1977), Matsumoto et al. (Reference Matsumoto, Muramoto, Hirano and Takahashi1981), Sekine, Takagi & Hirano (Reference Sekine, Takagi and Hirano1985), Funaki & Hirano (Reference Funaki and Hirano2004), Oizumi et al. (Reference Oizumi, Kurihara, Funaki and Hirano2005), Nishimura, Maeda & Shigeta (Reference Nishimura, Maeda and Shigeta2006) and Uramoto et al. (Reference Uramoto, Fujita, Takahashi and Hirano2007). Planktonic foraminiferal zones are after Hasegawa & Saito (Reference Hasegawa and Saito1993) and Nishi et al. (Reference Nishi, Takashima, Hatsugai, Saito, Moriya, Ennyu and Sakai2003). Ksm – Kamikinenbetsu Sandstone Member. For legend see Figure 1.
The Upper Cretaceous δ13Corg profile for the Obira area shows remarkable variations at specific horizons (Fig. 4). The most notable feature of the δ13Corg profile in the Obira area is characterized by a +3.5‰ positive interval from the uppermost part of the planktonic foraminifer Rotalipora cushmani Zone to the middle part of the Whiteinella archaeocretacea Zone around the Cenomanian–Turonianboundary.
Following the positive excursion, δ13Corg values vary between ~ −25‰ and ~ −23.5‰, superimposed by several short-term positive anomalies of ~ 1‰ through Turonian–Santonian sequences.
5. Evaluation of sedimentary organic matter
The sedimentary organic matter in the present samples is the type III/IV kerogen (Fig. 3), indicating that sedimentary organic matter with an origin from terrestrial plants predominates in the modal composition of the kerogen (Hunt, Reference Hunt1996).
Based on the estimated vitrinite reflectance values of about 0.5–0.9%Ro (Fig. 3), the organic maturation level of the samples corresponds to the catagenesis stage (Mukhopadhyay, Reference Mukhopadhyay, Mukhopadhyay and Dow1994; Hunt, Reference Hunt1996). It is known that significant changes in the δ13C values of kerogen do not occur below the metamorphic stage (Teerman & Hwang, Reference Teerman and Hwang1991; Whiticar, Reference Whiticar1996). Our results demonstrate that the maximum maturity of the kerogen in our samples did not reach the metamorphic stage; thus, we can rule out any influence of the maturity on the δ13Corg values of our samples.
The kerogen character for the Upper Turonian–Santonian sequence in the Obira area is compositionally and geochemically identical to the kerogen for the underlying Cenomanian–Middle Turonian sequences (Hasegawa & Saito, Reference Hasegawa and Saito1993; Hasegawa, Reference Hasegawa2001; Uramoto et al. Reference Uramoto, Fujita, Takahashi and Hirano2007). Therefore, the δ13C values of the kerogen in study samples represent fluctuations in the carbon isotope values of terrestrial plants in the hinterland of the NE Asian region.
6. Discussion
6.a. Correlation of δ13C profiles of terrestrial organic matter within the Yezo Group
We compare significant carbon isotope fluctuations of terrestrial organic matter from the Upper Cretaceous sequence in the Obira area with previously reported δ13C profiles for terrestrial organic matter from Japan and Russia (Hasegawa, Reference Hasegawa1997, Reference Hasegawa2003a; Hasegawa & Hatsugai, Reference Hasegawa and Hatsugai2000; Hasegawa et al. Reference Hasegawa, Pratt, Maeda, Shigeta, Okamoto, Kase and Uemura2003) (Fig. 5). Correlations of δ13C profiles were validated by the biostratigraphic intrabasinal correlation in the global chronological time scale on the basis of worldwide macro- and microfossils, as well as biohorizons of regional marker inoceramids in the NW Pacific.
Figure 5. Correlation of the Upper Cretaceous δ13C curves of terrestrial organic matter (δ13CTOM) of the Yezo Group in Hokkaido, Japan and Sakhalin, Russia (Oyubari area: composite curve of Hasegawa, Reference Hasegawa1997 and Tsuchiya, Hasegawa & Pratt, Reference Tsuchiya, Hasegawa and Pratt2003, published in Hasegawa, Reference Hasegawa2003a; Obira area: this study, Hasegawa & Saito, Reference Hasegawa and Saito1993, and Uramoto et al. Reference Uramoto, Fujita, Takahashi and Hirano2007; Kotanbetsu area: Hasegawa & Hatsugai, Reference Hasegawa and Hatsugai2000; Naiba area: Hasegawa et al. Reference Hasegawa, Pratt, Maeda, Shigeta, Okamoto, Kase and Uemura2003). Correlative carbon isotope fluctuations are connected with grey bands and a dashed line. Solid lines are the first occurrence of the regional marker inoceramid of the Yezo Group. The ‘IH spike’ in the Oyubari and Naiba areas is a Middle Turonian positive δ13C event reported by Hasegawa (Reference Hasegawa2003a). Macrofossil data for the Obira area are after Tanabe et al. (Reference Tanabe, Hirano, Matsumoto and Miyata1977), Matsumoto et al. (Reference Matsumoto, Muramoto, Hirano and Takahashi1981), Sekine, Takagi & Hirano (Reference Sekine, Takagi and Hirano1985), Funaki & Hirano (Reference Funaki and Hirano2004), Oizumi et al. (Reference Oizumi, Kurihara, Funaki and Hirano2005), Nishimura, Maeda & Shigeta (Reference Nishimura, Maeda and Shigeta2006) and Uramoto et al. (Reference Uramoto, Fujita, Takahashi and Hirano2007). Planktonic foraminifera zones for the Obira area are after Hasegawa & Saito (Reference Hasegawa and Saito1993) and Nishi et al. (Reference Nishi, Takashima, Hatsugai, Saito, Moriya, Ennyu and Sakai2003).
The correlation suggests that terrestrial organic δ13C profiles show notable fluctuations at specific stratigraphic horizons and all the fluctuations are closely associated with the occurrences of age-diagnostic fossils (Fig. 5). The value and amplitude of each δ13C fluctuation are similar among sections, so the δ13C fluctuation is interpreted to represent the averaged δ13C fluctuation of terrestrial plants in the hinterland of NE Asian region.
The Cenomanian–Turonian boundary δ13C fluctuation is correlated with the sharp positive δ13C excursion (~ +2.5‰) in the Oyubari area (Hasegawa, Reference Hasegawa1997), because the maximum δ13C value in this positive event is recorded within the planktonic foraminifera Whiteinella archaeocretacea Zone. The δ13C values observed in the Obira area (from −25.3‰ to −21.8‰) are similar to the δ13C fluctuations in the Kotanbetsu area (from −25.0‰ to −21.8‰) (Hasegawa & Hatsugai, Reference Hasegawa and Hatsugai2000), thus suggesting the maximum amplitude of terrestrial organic δ13C fluctuation during the Cenomanian–Turonian boundary event in the hinterland. The δ13C fluctuation also corresponds to the positive excursion of ~ +1.5‰ recorded in the lower part of the Naiba area (Hasegawa et al. Reference Hasegawa, Pratt, Maeda, Shigeta, Okamoto, Kase and Uemura2003). While previous studies have reported that the Cenomanian–Turonian boundary positive δ13C fluctuation is characterized by a single or double spike event in the Yezo Group (Hasegawa & Saito, Reference Hasegawa and Saito1993; Hasegawa, Reference Hasegawa1997; Hasegawa & Hatsugai, Reference Hasegawa and Hatsugai2000; Hasegawa et al. Reference Hasegawa, Pratt, Maeda, Shigeta, Okamoto, Kase and Uemura2003), the present study obtained δ13C data with higher resolution, thereby identifying high-frequency δ13C fluctuations. The significant short stratigraphic range of δ13C fluctuation (Hasegawa, Reference Hasegawa1997) and abrupt change in lithology reported for the Cenomanian–Turnian boundary sequence (Kurihara & Kawabe, Reference Kurihara and Kawabe2003; Takashima et al. Reference Takashima, Kawabe, Nishi, Moriya, Wani and Ando2004) in the Oyubari area denote the absence of part of the stratigraphic sequence due to hiatus or a winnowing process (Stow, Reading & Collinson, Reference Stow, Reading, Collinson and Reading1996). The different δ13C patterns in the Kotanbetsu and Naiba areas (Hasegawa & Hatsugai, Reference Hasegawa and Hatsugai2000; Hasegawa et al. Reference Hasegawa, Pratt, Maeda, Shigeta, Okamoto, Kase and Uemura2003) are due to the relative scarcity of exposure and the wide sampling interval.
Subsequently, the δ13C profile of terrestrial organic matter in the Yezo Group shows several positive-and-negative anomalies of ~ 1‰ in the Middle Turonian, upper Turonian–lower Coniacian and Santonian sequences (Fig. 5). The positive δ13C fluctuation in association with the occurrence of Middle Turonian ammonoid Collignoniceras woollgari in the Obira area is correlated with the ‘IH spike’ that is a positive δ13C event in the Middle Turonian sequence of the Yezo Group documented by Hasegawa (Reference Hasegawa2003a) (Fig. 5). Occurrences of Inoceramus hobetsensis Nagao and Matsumoto in the Yezo Group support this regional chemostratigraphic correlation. Our correlation of the ‘IH spike’ reinforces the significance of this δ13C fluctuation for stratigraphic correlation. In addition, biostratigraphic calibration also indicates that another δ13C positive feature is also present above the ‘IH spike’. The fluctuation is characterized by a small positive peak of ~ +0.7‰ in the Oyubari area (Hasegawa, Reference Hasegawa2003a) and this feature seems to be present in the data from the Obira, Kotanbetsu and Naiba areas (dashed line in Fig. 5). However, this feature is recognized by few data points except for the Oyubari area. Further δ13C measurements are expected to confirm this correlation.
A positive-and-negative δ13C fluctuation pattern which is characterized by a negative notch within an overall positive δ13C fluctuation is recognizable with occurrences of Didymotis costatus and Barroisiceras spp. in the upper Turonian–lower Coniacian sequence of the Obira area. Correlative counterparts of this fluctuation are present in association with the first occurrence of Inoceramus uwajimensis Yehara of the Yezo Group in the Oyubari and Naiba areas (Hasegawa, Reference Hasegawa2003a; Hasegawa et al. Reference Hasegawa, Pratt, Maeda, Shigeta, Okamoto, Kase and Uemura2003) (Fig. 5). The positive-and-negative δ13C pattern is important for correlation. However, because the negative notch is represented by few data, addition of terrestrial δ13C data is necessary to confirm this correlation.
A Santonian positive δ13C shift that is closely associated with the first occurrence of Texanites (Plesiotexanites) kawasakii in the Obira area is comparable with the positive δ13C shift just above the first occurrence of Inoceramus amakusensis Nagao and Matsumoto in the Naiba area (Hasegawa et al. Reference Hasegawa, Pratt, Maeda, Shigeta, Okamoto, Kase and Uemura2003) (Fig. 5). Our correlation of the Santonian terrestrial organic δ13C fluctuations provides a new chemostratigraphic constraint on the Upper Cretaceous sequence of the Yezo Group.
6.b. δ13C event correlations in marine carbonates of Europe and terrestrial organic matter of the NW Pacific
The above arguments demonstrate the correlation of terrestrial organic δ13C events of the Upper Cretaceous sequence at a regional scale. In the following discussion, we demonstrate the global chemostratigraphic correlation of these terrestrial organic δ13C fluctuations to the carbonate δ13C record in Europe, based on biostratigraphic data (Figs 6, 7).
Figure 6. Correlation of the Upper Cretaceous δ13C curves of terrestrial organic matter (δ13CTOM) of the Yezo Group in the Obira area, Hokkaido, Japan (this study; Hasegawa & Saito, Reference Hasegawa and Saito1993; Uramoto et al. Reference Uramoto, Fujita, Takahashi and Hirano2007), and the reference δ13Ccarbonate curve in Europe (Jarvis et al. Reference Jarvis, Gale, Jenkyns and Pearce2006). Grey shaded areas and a dashed line mark the correlative carbon isotope fluctuations. Detailed correlation of the Cenomanian–Turonian Boundary carbon isotope event between Europe and Japan is shown in Figure 7. Macrofossil data for the Obira area are after Tanabe et al. (Reference Tanabe, Hirano, Matsumoto and Miyata1977), Matsumoto et al. (Reference Matsumoto, Muramoto, Hirano and Takahashi1981), Sekine, Takagi & Hirano (Reference Sekine, Takagi and Hirano1985), Funaki & Hirano (Reference Funaki and Hirano2004), Oizumi et al. (Reference Oizumi, Kurihara, Funaki and Hirano2005) and Uramoto et al. (Reference Uramoto, Fujita, Takahashi and Hirano2007). Planktonic foraminifera zones for the Obira area are after Hasegawa & Saito (Reference Hasegawa and Saito1993) and Nishi et al. (Reference Nishi, Takashima, Hatsugai, Saito, Moriya, Ennyu and Sakai2003).
Figure 7. Detailed correlation of the Cenomanian–Turonian boundary δ13C event (CTBE) recorded in marine carbonates (δ13Ccarb) in Eastbourne, England, and terrestrial organic matter (δ13CTOM) in the Obira area, Hokkaido, Japan. Grey shaded area is the chemostratigraphic range of the CTBE. Correlative δ13C peaks are connected with dashed lines. Positive peaks at a, b and c are defined by global correlation of CTBE between GSSP and England (Jarvis et al. Reference Jarvis, Gale, Jenkyns and Pearce2006). See text for the correlation of positive peak b’ and negative carbon isotope fluctuations. Carbon isotope data for Eastbourne are after Paul et al. (Reference Paul, Lamolda, Mitchell, Vaziri, Gorostidi and Marshall1999) and biostratigraphic data are after Keller et al. (Reference Keller, Han, Adatte and Burns2001) and Gale et al. (Reference Gale, Kennedy, Voigt and Walaszczyk2005). Carbon isotope data for the Obira area are after this study, Hasegawa & Saito (Reference Hasegawa and Saito1993) and Uramoto et al. (Reference Uramoto, Fujita, Takahashi and Hirano2007). Macrofossil data for the Obira area are after Sekine, Takagi & Hirano (Reference Sekine, Takagi and Hirano1985), Funaki & Hirano (Reference Funaki and Hirano2004), and Uramoto et al. (Reference Uramoto, Fujita, Takahashi and Hirano2007). Planktonic foraminifera zones for the Obira area after Hasegawa & Saito (Reference Hasegawa and Saito1993) and Nishi et al. (Reference Nishi, Takashima, Hatsugai, Saito, Moriya, Ennyu and Sakai2003).
6.b.1. Cenomanian–Turonian boundary δ13C fluctuation
The isotopic fluctuations that encompass the Cenomanian–Turonian boundary are the best documented δ13C event in the Cretaceous sequence (e.g. Pratt & Threlkeld, Reference Pratt, Threlkeld, Stott and Glass1984; Gale et al. Reference Gale, Jenkyns, Kennedy and Corfield1993; Pratt et al. Reference Pratt, Arthur, Dean, Scholle, Caldwell and Kauffman1993; Hasegawa, Reference Hasegawa1997; Paul et al. Reference Paul, Lamolda, Mitchell, Vaziri, Gorostidi and Marshall1999; Keller et al. Reference Keller, Han, Adatte and Burns2001, Reference Keller, Stübe, Berner and Adatte2004; Tsikos et al. Reference Tsikos, Jenkyns, Walsworth-Bell, Petrizzo, Forster, Kolonic, Erba, Premoli Silva, Baas, Wagner and Sinninghe Damsté2004; Bowman & Bralower, Reference Bowman and Bralower2005; Jarvis et al. Reference Jarvis, Gale, Jenkyns and Pearce2006; Li et al. Reference Li, Jenkyns, Wang, Hu, Chen, Wei, Huang and Cui2006; Voigt et al. Reference Voigt, Aurag, Leis and Kaplan2007). The most detailed carbon isotope records of the Cenomanian–Turonian boundary event have been reported for marine carbonates and marine organic matter from North America and Europe (Paul et al. Reference Paul, Lamolda, Mitchell, Vaziri, Gorostidi and Marshall1999; Keller et al. Reference Keller, Stübe, Berner and Adatte2004; Tsikos et al. Reference Tsikos, Jenkyns, Walsworth-Bell, Petrizzo, Forster, Kolonic, Erba, Premoli Silva, Baas, Wagner and Sinninghe Damsté2004; Bowman & Bralower, Reference Bowman and Bralower2005; Voigt et al. Reference Voigt, Aurag, Leis and Kaplan2007). Based on the correlation between North American and European data, Jarvis et al. (Reference Jarvis, Gale, Jenkyns and Pearce2006) proposed that the following three positive peaks characterize the δ13C event of the Cenomanian–Turonian boundary (peak a–c in Jarvis et al. (Reference Jarvis, Gale, Jenkyns and Pearce2006); Fig. 7): (1) peak a, the first positive peak; (2) peak b, the positive peak in the middle part of the positive excursion; and (3) peak c, the positive peak at the top of the positive excursion.
Biostratigraphic calibration suggests that the δ13C patterns of the Cenomanian–Turonian boundary event are highly conformable between marine carbonates and terrestrial organic matter (Fig. 7), so the three peaks are correlated with discrete terrestrial organic δ13C peaks in Cenomanian–Turonian boundary event of the Obira area, Hokkaido (Fig. 7): (1) the sharp positive δ13C peak in the uppermost part of the Rotalipora cushmani Zone with peak a; (2) the positive δ13C peak in the lower part of the Whiteinella archaeocretacea Zone with peak b; and (3) the positive peak that corresponds with the occurrence of Lower Turonian ammonoid Vascoceras cf. durandi with peak c.
Consistent δ13C patterns allow further correlation of positive-and-negative fluctuations of the Cenomanian–Turonian boundary event: a small positive carbon isotope fluctuation below the first occurrences of Turonian index fossils (peak b’ in Fig. 7); minimum carbon isotope value between the each positive peak; and top and bottom of the Cenomanian–Turonian boundary event. Therefore, a total of nine chemostratigraphic horizons are proposed for the key to the high-resolution chemostratigraphy of the Cenomanian–Turonian boundary sequence (Fig. 7).
6.b.2. Middle Turonian δ13C fluctuation
The Middle Turonian δ13Ccarb profiles for Europe show a longer-term negative fluctuation. This fluctuation is superimposed by two shorter-term positive isotopic events (Jarvis et al. Reference Jarvis, Gale, Jenkyns and Pearce2006). Previous studies termed the two positive fluctuations as the Round Down Event and the Pewsey Event, in ascending order (Gale, Reference Gale, Hesselbo and Parkinson1996; Jarvis et al. Reference Jarvis, Gale, Jenkyns and Pearce2006).
The Round Down Event and Pewsey Event are tentatively correlated to the ‘IH spike’ and the above positive fluctuations in the Yezo Group (Fig. 6), based on the occurrences of Middle Turonian ammonoid Collignoniceras woollgari and biostratigraphic position of the terrestrial organic δ13C fluctuations. Recently, Hasegawa (Reference Hasegawa2003a) correlated the ‘IH spike’ to the Pewsey Event in Europe. We re-interpret the correlation on the basis of the biostratigraphic calibration, and our new global correlation suggests that the patterns of conspicuous positive δ13C of terrestrial organic matter for the Middle Turonian sequence of the Yezo Group are conformable with the carbonate δ13C curve in Europe. However, this correlation should be confirmed by further δ13C analyses of the Yezo Group, because the correlative counterpart of the Pewsey Event in the Yezo Group is represented by few data points.
6.b.3. Upper Turonian–lower Coniacian δ13C fluctuation
The δ13Ccarb fluctuations through the Upper Turonian–Lower Coniacian sequence display a δ13C minimum at the Turonian–Coniacian boundary that is sandwiched by the positive values in the Upper Turonian and Lower Coniacian (Voigt & Hilbrecht, Reference Voigt and Hilbrecht1997; Voigt, Reference Voigt2000; Jarvis et al. Reference Jarvis, Gale, Jenkyns and Pearce2006). These fluctuations were previously termed the Hitch Wood Event (Upper Turonian), Navigation Event (Turonian–Coniacian boundary) and Beeding Event (Lower Coniacian) (Gale, Reference Gale, Hesselbo and Parkinson1996; Jarvis et al. Reference Jarvis, Gale, Jenkyns and Pearce2006).
The positive-and-negative patterns of δ13Ccarb fluctuation are correlated with the upper Turonian–lower Coniacian δ13C fluctuation in the Yezo Group, based on the combined occurrences of Didymotis costatus and Barroisiceras spp. (Fig. 6); however, this correlation remains tentative, because the negative notch within the overall positive fluctuation in the Yezo Group is represented by few data points. Additional terrestrial organic δ13C analyses are necessary to confirm this correlation.
6.b.4. Santonian δ13C fluctuation
The Santonian δ13Ccarb profiles for Europe show an overall increasing trend with several superimposed positive-and-negative isotopic events (Jarvis et al. Reference Jarvis, Gale, Jenkyns and Pearce2006). The most conspicuous positive δ13Ccarb shift occurs between the negative δ13Ccarb event of the ‘Haven Brow Event’ and the positive δ13Ccarb event of the ‘Horseshoe Bay Event’. Association of the first occurrence of ammonoid Texanites (Plesiotexanites) kawasakii with the Santonian positive δ13C fluctuation of terrestrial organic matter supports the correlation of the positive shift in the NW Pacific to the European positive shift (Fig. 6), thus underlining the significance of the δ13C fluctuation for the global correlation of the Santonian marine sequences.
6.c. Implications for δ13C fluctuations of terrestrial organic matter
We demonstrate the global correlation of short-term δ13C fluctuation of the Upper Cretaceous sequence between marine and terrestrial records based on the biostratigraphic calibration. The correlation suggests that the patterns of remarkable δ13C fluctuations in marine carbonates are conformable in terrestrial organic δ13C fluctuations. This observation indicates that the terrestrial organic δ13C data mirror the global isotopic patterns in the carbon reservoir of the ocean–atmosphere–terrestrial biosphere system during the global marked δ13C fluctuation period. The factors for fundamental similarity in the δ13C fluctuation between marine and terrestrial records are generally considered to have been induced by the organic carbon burial process changes in association with primary production and sea-level controlled accommodation space changes (e.g. Arthur, Dean & Pratt, Reference Arthur, Dean and Pratt1988; Jarvis et al.Reference Jarvis, Gale, Jenkyns and Pearce2006).
However, it is noteworthy that our correlation also indicates that the amplitude of several terrestrial organic δ13C fluctuations is larger than that of the coeval carbonate δ13C fluctuations (Figs 6, 7). Particularly for the Cenomanian–Turonian boundary excursion, the amplitude of the δ13C event in terrestrial records of the Yezo Group (+3.5‰ in the Obira area: this study; +3.2‰ in the Kotanbetsu area: Hasegawa & Hatsugai, Reference Hasegawa and Hatsugai2000) is significantly higher than European marine carbonate records (~ +2.5‰: Jarvis et al. Reference Jarvis, Gale, Jenkyns and Pearce2006). This observation suggests that palaeoenvironmental parameters which affect terrestrial organic carbon isotopes on a regional scale may play a role in forming such different fluctuations. Factors such as irradiance, nutrients, age (juvenile v. adult), and seasonal variation have been cited as possible factors in controlling the terrestrial organic δ13C value (e.g. Arens, Jahren & Amundson, Reference Arens, Jahren and Amundson2000). These factors, however, are homogenized in the case of the Hokkaido sections because terrigenous materials are derived from a sufficiently large hinterland in the NE Asian region (Hasegawa, Reference Hasegawa2001).
Recently, Hasegawa (Reference Hasegawa2003b) inferred that pCO2 changes, and temperature and humidity-related carbon cycle processes in forests, affected the terrestrial organic δ13C values of the Yezo Group, based on the discrepancies of δ13C fluctuation within the middle Cenomanian–early Turonian sequences between Japan and Europe. However, although oxygen isotope and TEX86 palaeothermometry have suggested that the climatic optimum during Cretaceous times was in the Late Turonian age (e.g. Bornemann et al. Reference Bornemann, Norris, Frierich, Beckmann, Schouten, Sinninghe Damsté, Vogel, Hofmann and Wagner2008), significant δ13C discrepancies between marine carbonates and terrestrial organic matter are not present in Late Turonian δ13C records (Fig. 6). Therefore, it is likely that other mechanisms also affected the terrestrial organic δ13C fluctuations of the Yezo Group.
Additional key factors for terrestrial organic δ13C fluctuation are the CO2 emission into the atmosphere on a regional scale by release of terrestrial methane hydrate (Archer, Reference Archer2007) or biomass burning in the terrestrial realm (Kurtz et al. Reference Kurtz, Kump, Arthur, Zachos and Paytan2003; Finkelstein, Pratt & Brassell, Reference Finkelstein, Pratt and Brassell2006). Because δ13C values of methane hydrate and terrestrial organic matter (terrestrial vegetation, soil organic matter and peat) are lower than those of atmospheric CO2, the above CO2 emission processes generate light δ13C values of terrestrial organic matter. In the case of the Hokkaido sections, terrestrial organic δ13C values tend to be negative, especially in Late Cenomanian time (this study; Hasegawa, Reference Hasegawa2003b; Uramoto et al. Reference Uramoto, Fujita, Takahashi and Hirano2007), as opposed to the coeval positive trend of European carbonate δ13C records (Fig. 6). Such negative and amplified terrestrial organic δ13C fluctuations of the Yezo Group can be explained by the above CO2 emission into the atmosphere in the Yezo Group hinterland of the NE Asian region.
7. Conclusions
We obtained a carbon isotope profile of terrestrial organic matter for the Upper Cenomanian–Santonian sequence in the Obira area of Hokkaido, northern Japan. The terrestrial organic δ13C profile shows remarkable fluctuations for the Cenomanian–Turonian boundary, the Middle Turonian, the Turonian–Coniacian boundary, and the Santonian sequences. Based on the presence of internationally recognizable macro- and microfossils, these short-term terrestrial organic δ13C events are correlated with the previously reported δ13C events in Japan and Europe. These correlations reinforce the utility of these isotopic events in terms of global chemostratigraphic correlations. In particular, correlation of the Cenomanian–Turonian boundary event demonstrates that highly conformable δ13C patterns are recognizable between marine carbonates and terrestrial organic matter. The fundamental similarity of δ13C fluctuations between marine and terrestrial data indicates that the terrestrial organic δ13C data from the Cenomanian–Turonian boundary mirror the global isotopic patterns in the carbon reservoir of the ocean–atmosphere–terrestrial biosphere system. In addition, our correlation of short-term δ13C fluctuations of the Upper Cretaceous sequence between marine and terrestrial records also suggests that the magnitude of several terrestrial organic δ13C events is greater than that of coeval marine carbonate δ13C events. Especially for the Cenomanian–Turonian boundary δ13C excursion, the amplitude of terrestrial organic δ13C fluctuations of the Yezo Group (+3.5‰ in the Obira area: this study; +3.2‰ in the Kotanbetsu area: Hasegawa & Hatsugai, Reference Hasegawa and Hatsugai2000) is significantly higher than European carbonate δ13C fluctuations (+2.5‰ in Europe: Jarvis et al. Reference Jarvis, Gale, Jenkyns and Pearce2006). This observation is interpreted to indicate that the regional CO2 emission into the atmosphere by release of terrestrial methane hydrate or biomass burning of terrestrial organic matter amplified the δ13C profiles of terrestrial organic matter in the Yezo Group to a magnitude greater than those of the global ocean-atmosphere δ13C trend.
Acknowledgements
We deeply appreciate IGCP507 members for discussion, M. Ito and S. Yoshida (Chiba University) for support and encouragement, H. Nishi (Hokkaido University) and T. Hasegawa (Kanazawa University) for critical reading of the earlier version of the manuscript, T. Ohta (Waseda University, Tokyo) for fruitful comments on the manuscript, and K. Kurihara (Mikasa City Museum) for helpful suggestions. We conducted analyses at JAPEX Research Centre, Chiba, Japan. We are grateful to M. Yamamura and M. Miyairi for providing us an opportunity to use apparatus, S. Kato, N. Takeda and A. Waseda for support throughout this study, and Y. Kajiwara, H. Nishita, M. Aoyama, K. Kuriyama and K. Yuki for analytical support. We also thank H. Funaki for discussion and cooperation throughout this study while he was a student at Waseda University, R. Tsuruta, M. Kaneko, M. Oizumi, S. Tanaka, T. Fujita, F. Yajima, R. Murakami, K. Seike and H. Suda for much help in the fieldwork, and R. Kato (Rumoi City) for heartful support during our fieldwork. The manuscript was greatly improved by the insightful reviews of S. Voigt and two anonymous reviewers. This study was financially supported by a Grant-in-aid for the scientific research of the Japan Society for the Promotion of Science to HH (B, 15340178 for 2003–2005; B, 19340157 for 2007–2009), Waseda University Grant to HH (2000A-096; 2001A-536), and Sasakawa Scientific Research Grant from The Japan Science Society to GU (19-625, 2007).
