An open marine record of the Toarcian oceanic anoxic event

,

event (the T-OAE). Associated with this event, there were pronounced perturbations in global weathering rates and seawater temperatures. Although it is commonly asserted that the T-OAE is a global event and that the distribution of black shales is likewise global, an isotopic and/or organic-rich expression of this event has as yet only been recognized on epi-and peri-continental Pangaean localities. To address this issue, the 15 carbon isotope composition of organic matter (δ 13 C org ) of Early Toarcian cherts from Japan that were deposited in the open Panthalassa Ocean was analysed. The results show the presence of a major (>6‰) negative excursion in δ 13 C org that, based on radiolarian biostratigraphy, is a correlative of the Early Toarcian negative CIE known from European epicontinental strata. Furthermore, a secondary ca. −2‰ excursion

Introduction
Over the past quinquennium there has been an increased effort to unravel the cause(s) 5 and consequences of the Early Toarcian oceanic anoxic event (T-OAE). Of enormous diagnostic significance over this interval is the presence of a pronounced, transient negative carbon isotope (δ 13 C) excursion that has been recognized in marine inorganic and organic matter (Jenkyns and Clayton, 1988;Schouten et al., 2000;Kemp et al., 2005;Suan et al., 2008), and terrestrial plant material (Hesselbo et al., 2000;10 2007). This feature has been postulated to be the result of an increase in 12 C in the carbon cycle, perhaps following the dissociation of methane hydrates (Hesselbo et al., 2000;Kemp et al., 2005). Associated with the T-OAE is concomitant evidence for: a sudden rise in seawater palaeotemperatures (Rosales et al., 2004;van de Schootbrugge et al., 2005), a three-fold increase in atmospheric CO 2 levels (McElwain et 15 al., 2005), a major increase in silicate weathering rates (Cohen et al., 2004;Waltham and Gröcke, 2006) and a biotic crisis affecting marine invertebrates and biocalcifying micro-organisms (Little and Benton, 1995;Mattioli et al., 2004;Tremolada et al., 2005). The lithological expression of the T-OAE may be locally very different. The development of anoxic, organic-rich facies was not generalized in the European Neotethyan 20 realm and, where developed, the petrology and organic richness of the sediments is variable. For instance, certain limestones in the H. falciferum ammonite Zone of northern Europe (Hallam, 1967) and of the Tethys (Jenkyns, 1985) are pink/red. These differences illustrate the importance of local environmental conditions that are ultimately more important in determining the local lithological expression of an OAE (Trabucho 25 Alexandre et al., 2010). Indeed, available biostratigraphical data indicate that the development and demise of organic-rich facies in Tethyan and Boreal provinces of Europe 387 Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | are diachronous (Wignall et al., 2005 and references therein): similar to the Cenomanian/Turonian boundary OAE (Tsikos et al., 2004). Hesselbo et al. (2000) suggested that the cause of the negative CIE was the dissociation, release and consequent oxidation of methane derived from continental shelf methane hydrates. More recently, a high-resolution δ 13 C org record across the T-OAE 5 obtained from mudrocks exposed in northeast England was generated by Kemp et al. (2005). These data show that the negative δ 13 C excursion was not a single event but three rapid events, the timing of which was paced by astronomically forced climate cycles. An alternative hypothesis recently put forward for the cause of the T-OAE is that 12 C-enriched thermogenic methane was released into the ocean-atmosphere system 10 due to an igneous intrusion into coaly organic-rich facies in the Karoo Basin, South Africa. (McElwain et al., 2005;Svensen et al., 2007). This mechanism has subsequently been questioned by Gröcke et al. (2009), who showed that both the physical nature of the contacts between intrusions and organic-rich facies and the geochemical pattern within these facies suggest little or no thermogenic methane generation from 15 the Karoo Basin.
A key controversy surrounding the T-OAE is that localities where the negative δ 13 C excursion has been recorded have primarily been restricted to epicontinental sections exposed in Europe. This has prompted speculation that the isotopic perturbation is a feature affecting only the shallow marine environment (i.e. <200 m water depth) and 20 driven by localized phenomena (van de Schootbrugge et al., 2005 In this study, the specific issue of the global extent of the Early Toarcian negative 5 δ 13 C excursion as well as the expression of the T-OAE in the open ocean are addressed by analysing a deep-sea chert record from Katsuyama, Japan (Fig. 1), which was located in the palaeo-Pacific (Panthalassic Ocean) during the Early Jurassic.

Geological setting
The studied samples were obtained from the Inuyama area, Mino Terrane, which is 10 one of the Mesozoic accretionary complexes presently exposed in Japan, extending to the north of Tokyo (Ashio Terrane) and lithologically and stratigraphically corresponding to the Chichibu Terrane ( Fig. 1). In Japan, these Mesozoic complexes consist of seamount-related sequences of volcanic rocks, limestones and pelagic sediments (bedded chert and siliceous mudstone) associated with trench-fill sediments. The sed- 15 iments were deposited on the seafloor of the Izanagi Plate from Carboniferous to Early Cretaceous and were subsequently accreted to the margin of Far East Asia during the Jurassic to Early Cretaceous (Matsuda and Isozaki, 1991). The stratigraphic section in the Inuyama area mainly consists of bedded cherts spanning the Early Triassic-Early/Middle Jurassic and has been well-constrained using con-20 odont and radiolarian biostratigraphy (e.g., Yao et al., 1982;Hori, 1990Hori, , 1993Matsuoka, 2004). Since the sedimentary complex of this area formed by accretionary processes at shallow depths (Kimura and Hori, 1993) undergoing extremely low-grade metamorphism (diagenetic zone of illite crystallinity index; Otsuka and Watanabe, 1992), it is considered that the cherts retain a primary record of the Upper Triassic The studied pelagic sediments in the Inuyama area are typically distal, open ocean sediments characterized by sedimentation rates of 1 m to 7 m Myr −1 and long sequences (>50 Myr) (Hori et al., 1993;Wignall et al., 2010). Although such sediments are generally considered to be analogous to present-day deep-sea radiolarian ooze (>3 km water depth), similar Mesozoic sequences drilled by the IODP offshore on Pa-5 cific seamounts are much shallower (within the mid-water oxygen minimum layer) and have been interpreted as reflecting deposition on seamounts under highly productive surface seawater below the equatorial divergence zone (Wilson et al., 1998;Jenkyns and Wilson, 1999;Robinson et al., 2004).
Palaeomagnetic studies reveal that these cherts contain primary magnetism and 10 record normal and reverse polarity zones during the Triassic (Shibuya and Samejima, 1986;Ando et al., 2001). The data imply that the Inuyama chert sequence was indeed deposited at low Southern Hemisphere latitudes during the Middle Triassic and migrated to the Northern Hemisphere during the Late Triassic-Early Jurassic, crossing the equatorial regions of the palaeo-Pacific (Fujii et al., 1993;Ando et al., 2001).

Katsuyama section
The biostratigraphy of the sequence at Katsuyama (Fig. 2), located in the northern part of the Inuyama area, has been well studied (Hori, 1990(Hori, , 1992(Hori, , 1997Takeuchi, 2001;Matsuoka, 2004). The studied sequence consists of a continuous succession of bedded cherts (>100 m thick), which based on radiolarian biostratigraphy (Matsuoka et al.,20 1994) spans the upper Lower Triassic (Spathian) to lower Middle Jurassic (Aalenian). Over the Lower Jurassic part of the succession, five radiolarian (sub)zones are recognized: the Parashuum simplum sSubzones I to IV, and the Mesosaturnalis hexagonus (=Hexasaturnalis hexagonus) Zone (in ascending order) (Hori, 1990;1997 Fig. 2). White chert (novaculite) may be the product of low-grade metamorphism of siliceous sequences or, alternatively, it may be the product of silicification of volcaniclastic rocks, which, given the geological setting appears to be more likely. The lithologies of this interval differ from the more predominant red cherts found above. Wignall et al. (2010) analysed samples from 5 Katsuyama and suggested the presence of anoxia based on pyrite framboid evidence. The correlation with their published log of the succession with the one presented here is ambiguous. The samples for the present isotopic study were obtained from the interval spanning the upper part of the Parashuum simplum IV (Ps IV) Subzone to the lower part of the Hexasaturnalis hexagonus (Hh) Zone, which is composed of a ca. 340 cm thick sequence of bedded chert. This equates to the radiolarian Zones in Carter et al. (2010) of Eucyrtidiellum nagaiae-Praeparvicingula tlellensis to Elodium pessagnoi-Hexasaturnalis hexagonus (see Fig. 2). On the basis of a 1 m Myr −1 sedimentation rate for the Lower Jurassic cherts of the Inuyama area (Hori et al., 1993), the studied 15 interval is calculated to have lasted a maximum of ca. 3.5 Myr.
Age assignments for the Katsuyama succession can be made based on the available radiolarian biostratigraphy. The Pliensbachian/Toarcian boundary is not wellconstrained using radiolarian biostratigraphy although the last occurrence of Eucyrtidiellum nagaiae is at ca. −30 cm (Fig. 2). This level is also about 2 m above the first 20 occurrence of Trillus elkhornensis in the Katsuyama section. On the basis of radiolarian data from east-central Oregon, North America, Trillus elkhornensis makes its first appearance in the Upper Pliensbachian Nicely Formation (Amaltheus margaritatus-Pleuroceras spinatum ammonite Zones) (Pessagno and Blome, 1980). In addition, the highest occurrence of Eucyrtidiellum sp. 2 (=E . sp. C2 of Nagai, 1986) has been reported from the Lower Toarcian Hyde Formation overlaying the Nicely Formation (Nagai, 1990 There is a significant gap in constraining the radiolarian biostratigraphy between ca. −30 cm and ca. 250 cm due to a lack of diagnostic species (Fig. 2). The lowest occurrence of the genus Parvicingula has been recognized in the middle Toarcian Whiteaves Formation (Hildoceras bifrons ammonite Zone) from the Queen Charlotte Islands, Canada (Carter et al., 1988) and Hexasaturnalis hexagonus is also obtained 5 from the upper/middle to lower/upper Toarcian sequence from the same area. Therefore at ca. 250 cm the first occurrence of Hexasaturnalis hexagonus suggests an age equivalent to the Harpoceras falciferum/Hildoceras bifrons ammonite Zone boundary of NW Europe (see Carter et al. 2010). This interval is also characterized by a remarkable faunal change, termed the Toarcian Radiolarian Event (Hori, 1997) (see Fig. 2).
Thus, if the T-OAE is present in the form of CIE then it should occur below ca. 240 cm (Fig. 2).

Analytical methodology
Bulk sediment samples were decalcified using 3M HCl in 50 ml centrifuge tubes for 16 h. Stable isotope measurements were performed at Durham University using a 15 Costech elemental analyser (ESC4010) coupled to a ThermoFinnigan Delta V Advantage isotope ratio mass spectrometer. Carbon isotope ratios are corrected for 17 O contribution (Craig, 1957) and reported in standard delta (δ) notation in per mil (‰) relative to VPDB. Data accuracy is monitored through routine analyses of international and in-house standards: the latter are stringently calibrated against the international 20 standards. Analytical uncertainty for δ 13 C org measurements was typically better than ±0.1‰ for standards and <0.2‰ on replicate sample analysis. Total organic carbon (TOC) data was obtained as part of this method. Because of the low TOC contents of the studied chert samples, the tin capsules used contained up to 100-120 mg of powdered chert and the Costech was set to Macro-O 2 . The isotope analyses were done in 25 no dilution mode and all results produced more than 1000 m.

The Katsuyama T-OAE record
The physical expression of the Early Toarcian OAE in the Katsuyama succession is defined through the presence of organic-rich facies (cf. Wignall et al., 2010). TOC contents through the section range between 0.01-2.75 wt %. Two marked increases in 5 TOC occur at ca. 45 cm and ca. 190 cm, where major lithologic changes are observed in the form of shifts from grey to black cherts (Fig. 3).
The siliceous sediments of the Katsuyama section are here interpreted to have been deposited in an open ocean setting on the seafloor of a volcanic seamount at mid-water depths. The seamount, which was originally emplaced in the Southern Hemisphere 10 (Fujii et al., 1993;Ando et al., 2001) was transported north-westwards to its current position in Far East Asia by seafloor spreading (cf. Wilson et al., 1998). During its journey, the seamount was covered by pelagic sediments and, within the equatorial divergence zone, siliceous black shales were deposited as part of the pelagic sequence. The black shales are organic-rich radiolarian cherts and, therefore, attest to conditions 15 of high productivity in the euphotic zone and of low oxygen at the seafloor. Whereas low oxygen conditions are a characteristic feature of the global oxygen minimum layer in the ocean, underneath the equatorial divergence zone these conditions would have been intensified. Due to the palaeogeographical setting of the Katsuyama section in the Early Jurassic, the T-OAE not only records the global perturbation in the carbon 20 cycle that is characteristic of this time interval and the record of which is known from sections worldwide, but it is also characterized by organic-rich open ocean lithologies.
Indeed, the presence of black shales in open ocean settings needs not necessarily to represent a spread of anoxia in the oceans, but rather the passage of an ideal substrate for the preservation of organic matter through a zone in the ocean which is 25 a priori anoxia-prone. This model helps to reconcile the presence of black shales in open ocean settings with evidence for upwelling in the open marine environment and for enhanced ocean overturning (Kajiwara et al., 1994;Suzuki et al., 1998 (Fig. 3), δ 13 C org values markedly increase from ca. −33‰ to ca. −25‰. This increase to less negative values is also coincident with a return to lower TOC values (Fig. 3).

Correlation of the Pliensbachian/Toarcian and T-OAE δ 13 C curve
The radiolarian biostratigraphy, in combination with the δ 13 C org record, permits a correlation to be made between the Katsuyama succession and Early Toarcian successions in Europe. Coupling of the age constraints of the Katsuyama succession to the TOC record and δ 13 C org variations strongly support the inference that the nega-20 tive CIE occurring between ca. 180 cm and ca. 250 cm height in the succession is the Early Toarcian negative CIE. Similarly, the data (Fig. 4) also supports the inference that the ca. 2‰ negative excursion occurring between ca. 60 cm and ca. 100 cm height represents the excursion found in European sections at the Pliensbachian/Toarcian boundary. Evidence for a Pliensbachian/Toarcian negative CIE from Europe has been 25 documented from sections in Yorkshire (Littler et al., 2009) and Peniche (Hesselbo et al., 2007;Suan et al., 2008) and is also apparent in the Mochras Farm record of Wales (Jenkyns et al., 2001) (Fig. 4) In the Argentinean δ 13 C org record presented by Al-Suwaidi et al. (2010), however, clear evidence for a negative shift at the Pliensbachian/Toarcian boundary was absent. As such, the data from the Katsuyama succession potentially represents the first evidence for the Pliensbachian/Toarcian δ 13 C org excursion outside Europe and in an open ocean setting, suggesting that the development of organic-rich facies was not restricted to epi-and peri-continental basins and thus more widespread than previously assumed. Interestingly, the rise in TOC occurs prior to the Pliensbachian/Toarcian CIE and thus differs from the T-OAE record, which typically indicates a TOC rise during the most negative part of the CIE (see Hesselbo et al., 2000;Jenkyns et al. 2002;Kemp et al., 2005).

Mechanistic Model for the T-OAE
The new geochemical record of the T-OAE presented in this study represents the first complete record of the excursion outside Europe and also the first open ocean expres-25 sion of the event. Using the "most" negative δ 13 C segment of the T-OAE as a tie-point, shows that the geochemical signature of this OAE is similar between epi-and pericontinental, and open ocean localities (Fig. 4). Moreover, because the Katsuyama succession is truly pelagic and, at the time of deposition, away from land influence, the negative excursion is shown to indeed represent a primary global signal. The T-OAE is therefore defined (and sediments deposited during the T-OAE associated with) by a 5 rapid negative excursion in the carbon isotope record that is recorded globally whatever the physical/lithological expression of the associated facies. In the open ocean, marine Katsuyama succession, the T-OAE interval is characterized by dysoxic, organic-rich facies (Fig. 2), with periodic anoxia and euxinia intervening (see also Wignall et al., 2010). 10 As noted in the introduction, the seemingly localized nature of the Early Toarcian negative δ 13 C excursion, apparently confined to epicontinental successions of the Neotethyan realm (Europe), has prompted a number of workers to question whether a global mechanism is capable of explaining the Early Toarcian excursion (e.g., Wignall et al., 2006;McArthur et al., 2007). An alternative hypothesis for the event has been offered instead which invokes local overturn of 12 C-enriched dissolved CO 2 from deeper waters to the surface, where photoautotrophic uptake of this carbon would consequently lead to a regional negative δ 13 C excursion (Kuspert, 1982;Rohl et al., 2000;Wignall et al., 2006;van de Schootbrugge et al., 2005). The overturn hypothesis was questioned by van Breugel et al. (2006) who analysed isorenieratene abundances and 20 δ 13 C across the event in the Paris Basin, France and concluded that respired CO 2 could not have contributed significantly to the negative CIE. Indeed, in light of the data presented here, and that of the recent Al-Suwaidi et al. (2010) study from Argentina, the overturn scenario is improbable given that it implies that overturn of a large amount of 12 C affected the global ocean synchronously.

25
These results notwithstanding, the purported lack of a negative δ 13 C excursion in records of belemnite calcite has been used by some authors to support the claim that isotopically light carbon did not dominate the ocean-atmosphere system during the T-OAE (Wignall et al., 2006;van de Schootbrugge et al., 2005) argue against the methane hydrate and igneous intrusion thermogenic hypotheses. This view is countered by evidence for a synchronous negative CIE in terrestrial plant material (Hesselbo et al., 2000(Hesselbo et al., , 2007. This observation, coupled with the results presented here, unequivocally demonstrates that isotopically light carbon dominated the global marine organic (both shallow and deep) and atmospheric reservoirs of car-5 bon. Furthermore, evidence exists for a well-developed negative excursion in carbonate rocks dominated by coccolith fossils (Hermoso et al., 2009) and brachiopods (Suan et al., 2008). This raises the question as to whether belemnites are reliable recorders of the isotopic composition of seawater during environmental perturbations such as those recorded during the T-OAE (cf. Hermoso et al., 2009). 10 The emplacement of the Karoo-Ferrar large igneous province in the latest Pliensbachian is likely to be of significance in the development of the negative δ 13 C excursion that defines the T-OAE. Long-term injection of volcanic greenhouse gases (primarily CO 2 ) into the atmosphere associated with the emplacement of large igneous provinces is thought to have led to an increase in global mean temperatures (Pálfy and 15 Smith, 2000). Although the Toarcian palaeotemperature record may have been influenced by a reduction in salinity (freshwater input) and/or latitudinal water flow through the NW European corridor (Bjerrum et al., 2001), belemnite-derived Mg/ca. and δ 18 O palaeotemperature calculations show an increase in seawater temperature through the latest Pliensbachian and Early Toarcian (McArthur et al., 2000;Rosales et al., 2004). 800%. The magnitude of the increase was subsequently reduced by modelling work on a strontium isotope record through the Jurassic by Waltham and Gröcke (2006). McArthur et al. (2007) argued, however, that the strontium isotope profile through the event was an artefact of changing sedimentation rates, rather than a primary climatic 397 Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | signal. Nevertheless, an increase in both seawater palaeotemperature and weathering rates are in agreement with an increase in atmospheric CO 2 . An increased nutrient flux driven by accelerated continental weathering regimes would have enhanced and promoted surface water productivity and led to algal and bacterial blooms (Watson et al., 2000) in proximal epi-and peri-continental environ-5 ments. δ 15 N tot values through the Mesozoic suggest that cyanobacteria were a significant contribution to oceanic productivity, with values fluctuating around 0‰ (Jenkyns et al., 2002). However, modern laboratory experiments have shown that cyanobacteria require five times as much iron over carbon when grown with N 2 (Kustka et al., 2003), thus indicating that iron supply must have been high during the Mesozoic and, 10 in particular, during OAEs. The concurrent input of methane into the oceanic and atmospheric carbon reservoirs, and its subsequent oxidation would have increased CO 2 and decreased O 2 concentrations -potentially in both the ocean and atmosphere (i.e., Dickens, 2000). Also, a subsequent rise in global productivity would also have led to increased consumption of O 2 in the oceanic and atmospheric reservoirs, enhancing 15 organic matter preservation, extinctions in the ocean (Cecca. and Macchioni, 2004) and floral radiations in the terrestrial realm (Vakhrameev, 1991). Based on the findings presented here and those of Wignall et al. (2010), these processes clearly not only affected the shallow marine system but also the deeper, open marine environment with enhanced TOC values and evidence of anoxia/euxinia in the Katsuyama cherts (Fig. 3, Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | margin clathrates and the Toarcian δ 13 C negative excursion. Indeed, many of the geochemical features of the T-OAE are readily apparent during the more recent PETM and these similarities have prompted direct comparisons to be made (Cohen et al., 2007). The results presented here further attest to the broad similarities between the two events because a key feature of the PETM is a well-developed expression of carbon 5 isotope changes in the atmosphere, and shallow and deep oceans.

Conclusions
The perturbation in the carbon cycle that occurred during the T-OAE was a global phenomenon that affected all carbon reservoirs and is recorded by a negative CIE in terrestrial and marine sediments (both shallow and deep ocean settings). Therefore, the forcing mechanism(s) behind the T-OAE must have been capable of driving global environmental change. These mechanisms may have included: the eruption of the Karoo-Ferrar continental flood basalts; the massive dissociation of continental margin clathrates; an increased hydrological cycle; and a transient greenhouse climate with increased CO 2 concentrations and palaeotemperatures. Apart from the T-OAE CIE, the 15 Pliensbachian/Toarcian CIE is recorded in the Katsuyama cherts suggesting another global carbon isotope excursion. These CIEs are important for understanding the Early Jurassic carbon cycle, as well as a tool for stratigraphic correlation between diverse palaeogeographic realms. The palaeogeography of black shale depositional environments during OAEs is a 20 matter of discussion. All previously known examples of Toarcian black shales were deposited on epi-and peri-continental shelves in environments proximal to Pangaea. In this study, a distal, truly pelagic, open ocean setting is described. These black shales were deposited on the seafloor of a volcanic seamount due to its passage under the highly productive waters of the equatorial divergence zone of Panthalassa. The shorter export path for organic matter resulting from the presence of a submarine topographical high in association with the expansion of the oxygen minimum layer due to high oxygen 399 Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Dera, G., Pellenard, P., Neige, P., Deconinck, J.-F., Pucéat, E., and Dommergues, J.-L.: Dis-tribution of clay minerals in Early Jurassic Peritethyan seas: Palaeoclimatic significance inferred from multiproxy comparisons, Palaeogeogr. Palaeoclim., Palaeoecol., 271, 39-51, 2009. Dickens, G. R.: Methane oxidation during the Late Palaeocene Thermal Maximum, Bull. Soc. 5 Geol. France, 171, 37-49, 2000. Fujii, J., Hattori, I., and Nakajima, T.: A study of radiolarian biostratigraphy and magnetostratigraphy of early Mesozoic red bedded chert, central Japan, News Osaka Micropal., 9, 71-89 , 1993. (in Japanese with English translation) Gorican, S., Sumuc, A., and Baumgartner, P. O.: Toarcian Radiolaria from Mt. Mangart 10 (Slovenian-Italian border) and their paleoecological implications, Mar. Micropal., 49, 275-301, 2003. Gröcke, D. R., Rimmer, S. M., Yoksoulian, L. E., Cairncross, B., Tsikos, H., and van Hunen, J.: No evidence for thermogenic methane release in coal from the Karoo-Ferrar large igneous province, Earth Planet. Sci. Lett., 277, 204-212, 2009.
Copyright of Solid Earth Discussions is the property of Copernicus Gesellschaft mbH and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.