1. Introduction
Much attention has focused on the events surrounding the Early Toarcian Oceanic Anoxic Event (OAE: c. 183 Ma) in the Early Jurassic (Jenkyns, Reference Jenkyns1985, Reference Jenkyns1988). The relative enrichment with organic matter of certain strata of Early Toarcian age, together with the marine and terrestrial carbon-isotope record, indicates that a large perturbation in the carbon cycle took place at that time. There is less of a consensus on the exact nature and extent of the perturbation, in particular, the significance of the sharp negative carbon-isotope excursion (CIE) which punctuates a broad positive excursion, seen in marine organic-matter (δ13Corg) of between ~ −5 ‰ and −7 ‰, reported from the Boreal (NW European) falciferum Zone (Jenkyns & Clayton, Reference Jenkyns and Clayton1997; Hesselbo et al. Reference Hesselbo, Gröcke, Jenkyns, Bjerrum, Farrimond, Morgans Bell and Green2000; Schouten et al. Reference Schouten, Van Kaam-Peters, Rijpstra, Schoell and Sinninghe Damsté2000; Röhl et al. Reference Röhl, Schmid-Röhl, Oschmann, Frimmel and Schwark2001; Jenkyns et al. Reference Jenkyns, Jones, Gröcke, Hesselbo and Parkinson2002; Cohen et al. Reference Cohen, Coe, Harding and Schwark2004; Kemp et al. Reference Kemp, Coe, Cohen and Schwark2005). A similar but in most cases more subdued (~ −3.5 ‰ or less) excursion in hemipelagic carbonate is recorded from the corresponding levisoni Zone in Portugal (Hesselbo et al. Reference Hesselbo, Jenkyns, Duarte and Oliveira2007; Suan et al. Reference Suan, Pittet, Bour, Mattioli, Duarte and Mailliot2008b), from coeval strata in France (Hermoso et al. Reference Hermoso, Le Callonnec, Minoletti, Renard and Hesselbo2009), and from time-equivalent Tethyan deep-water pelagic and shallow-water platform-carbonate sections in Italy (Jenkyns & Clayton, Reference Jenkyns and Clayton1986; Jenkyns, Gröcke & Hesselbo, Reference Jenkyns, Gröcke and Hesselbo2001; Woodfine et al. Reference Woodfine, Jenkyns, Sarti, Baroncini and Violante2008; Sabatino et al. Reference Sabatino, Neri, Bellanca, Jenkyns, Baudin, Parisi and Masetti2009).
The positive carbon-isotope excursion associated with the Early Toarcian OAE has been attributed to increased sequestration of isotopically light marine organic matter in ocean sediments under oxygen-depleted conditions, leading to a corresponding ‘heavy’ signal preserved in materials such as marine carbonates, marine organic matter, and fossil wood. This interpretation is supported by the widespread occurrence of organic-rich black shales deposited during this time interval, which essentially define the OAE, and from biomarker evidence for photic-zone sulphate reduction in horizons that show this organic enrichment (Schouten et al. Reference Schouten, Van Kaam-Peters, Rijpstra, Schoell and Sinninghe Damsté2000; Pancost et al. Reference Pancost, Crawford, Magness, Turner, Jenkyns and Maxwell2004; Bowden et al. Reference Bowden, Farrimond, Snape and Love2006). In contrast, the negative carbon-isotope excursion has been interpreted by most workers as resulting from input of isotopically light carbon to the shallow ocean, atmosphere and biosphere, from a source such as gas hydrate or thermally metamorphosed shale and coal (Hesselbo et al. Reference Hesselbo, Gröcke, Jenkyns, Bjerrum, Farrimond, Morgans Bell and Green2000, McElwain, Wade-Murphy & Hesselbo, Reference McElwain, Wade-Murphy and Hesselbo2005; Kemp et al. Reference Kemp, Coe, Cohen and Schwark2005; Svensen et al. Reference Svensen, Planke, Chevallier, Malthe-Sørenssen, Corfu and Jamtveit2007). Alternative purely palaeoceanographic mechanisms, such as recycling of waters from below a chemocline, are untenable because they are unsupported by modern studies of present-day anoxic basins and fail to explain expression of the isotope excursion in terrestrial materials (Küspert Reference Küspert, Einsele and Seilacher1982; Van Breugel et al. Reference Van Breugel, Baas, Schouten, Mattioli and Sinninghe Damsté2006a,Reference Van Breugel, Schouten, Paetzel and Sinninghe Damstéb; cf. McArthur et al. Reference McArthur, Cohen, Coe, Kemp, Bailey and Smith2008).
The Pliensbachian–Toarcian boundary preceding the OAE has received less attention, but remains an important interval in terms of changes in carbon cycling, changes in sea level, faunal turn-over and evolving Early Jurassic palaeoclimate. A sea-level rise across the stage boundary of several tens of metres, marked in many sections by a change from shallower to deeper water facies, has previously been inferred (Hallam, Reference Hallam1981, Reference Hallam1997; Hesselbo & Jenkyns, Reference Hesselbo, Jenkyns, de Graciansky, Hardenbol, Jacquin, Farley and Vail1998). The boundary coincides with a relative minimum in seawater strontium-isotope values, with 87Sr/86Sr ratios falling to the least radiogenic values seen throughout the Early Jurassic (Jones et al. Reference Jones, Jenkyns, Coe and Hesselbo1994; Jones & Jenkyns, Reference Jones and Jenkyns2001). The transition from the Pliensbachian into the Toarcian is marked by the initiation of the mass-extinction event, suggesting that changes in climate and palaeoceanography started to occur well before the onset of the OAE (Hallam, Reference Hallam1986; Little & Benton, Reference Little and Benton1995; Harries & Little, Reference Harries and Little1999; Aberhan & Fürsich, Reference Aberhan and Fürsich2000; Macchioni & Cecca, Reference Macchioni and Cecca2002; Cecca & Macchioni, Reference Cecca and Macchioni2004; Wignall, Newton & Little, Reference Wignall, Newton and Little2005; Zakharov et al. Reference Zakharov, Shurygin, Il'ina and Nikitenko2006; Wignall & Bond, Reference Wignall and Bond2008).
All of the well-characterized mass extinction events in the geological record were accompanied by perturbations in the global carbon cycle (e.g. Keller & Lindinger, Reference Keller and Lindinger1989; Arens & Jahren, Reference Arens and Jahren2000; Payne et al. Reference Payne, Lehrmann, Wei, Orchard, Schrag and Knoll2004; Hesselbo et al. Reference Hesselbo, Robinson, Surlyk and Piasecki2002; Riccardi et al. Reference Riccardi, Kump, Arthur and D'Hondt2007). Such a perturbation has recently been reported for the first time from a Pliensbachian–Toarcian boundary section at Peniche, Portugal, whose ammonite faunas show both Boreal and Tethyan affinities, and where an excursion of ~ −2 ‰ has been seen in the δ13Ccarbonate record, in both bulk carbonate and brachiopod calcite (Hesselbo et al. Reference Hesselbo, Jenkyns, Duarte and Oliveira2007; Suan et al. Reference Suan, Mattioli, Pittet, Mailliot and Lécuyer2008a). The excursion, with a magnitude of ~ −3 ‰, is also visible in the corresponding δ13C values of macroscopic fossil wood. In this study, we present new data that show a similar negative CIE in the δ13Corg record from NE Yorkshire, England. We also supply new δ13Cbelemnite data from fossils collected near the peak of the δ13Ccarbonate excursion at Peniche.
2. Geological setting
2.a. Hawsker Bottoms, England
Samples were collected from the predominantly argillaceous cliff and foreshore sections at Hawsker Bottoms, NE Yorkshire, England (Fig. 1). This section consists of sideritic ironstones at the base, overlain by mudstones with varying degrees of organic enrichment and containing distinctive levels of concretionary limestone. This site has excellent exposure and has been the subject of many previous studies into Early Jurassic palaeoceanography (e.g. Jenkyns & Clayton, Reference Jenkyns and Clayton1997; Hesselbo et al. Reference Hesselbo, Gröcke, Jenkyns, Bjerrum, Farrimond, Morgans Bell and Green2000; McArthur et al. Reference McArthur, Donovan, Thirlwall, Fouke and Mattey2000; Bailey et al. Reference Bailey, Rosenthal, McArthur, Van de Schootbrugge and Thirlwall2003; Cohen et al. Reference Cohen, Coe, Harding and Schwark2004; Kemp et al. Reference Kemp, Coe, Cohen and Schwark2005; Van de Schootbrugge et al. Reference Van de Schootbrugge, McArthur, Bailey, Rosenthal, Wright and Miller2005; Wignall, Newton & Little, Reference Wignall, Newton and Little2005). It is possible to follow individual beds with confidence for some lateral distance along the coast, because of the excellent exposure and easy identification of prominent marker beds (Howarth, Reference Howarth1973). The Hawsker Bottoms exposure features sediments deposited in the Cleveland Basin, which constituted part of the NW European epicontinental shelf during Early Jurassic times (e.g. Hesselbo & Jenkyns, Reference Hesselbo, Jenkyns and Taylor1995; Bjerrum et al. Reference Bjerrum, Surlyk, Callomon and Slingerland2001). The fully marine, clay-rich strata are thought to have been deposited in a fairly shallow, periodically anoxic or dysoxic environment (Hallam, Reference Hallam1967). The transition from the inferred very shallow conditions in the latest Pliensbachian to the deeper water conditions associated with the Early Toarcian transgression are represented by a lithological change from dominantly oolitic ironstones and mudstones in the Cleveland Ironstone Formation, to dark, locally laminated mudstones with sideritic nodules in the overlying Grey Shale Formation. A thin bed just above the Pliensbachian–Toarcian boundary (Bed 26 in Fig. 2), known as the Sulphur Band (Howarth, Reference Howarth1973), is an organic-rich level, and probably records a brief but intense period of anoxia in the earliest Toarcian (Dactylioceras tenuicostatum Zone, Protogrammoceras paltum Subzone) (Wignall Reference Wignall1994).
Map to show the palaeogeography of the NW European epicontinental shelf region during the Late Pliensbachian–Early Toarcian interval, and location of study sites. Adapted from Ziegler (Reference Ziegler1988). 1 – Yorkshire, 2 – Peniche, AM – Armorican Massif, IBM – Iberian Massif, MC – Massif Central, LBM – London–Brabant Massif, IM – Irish Massif.
High-resolution data across the Pliensbachian–Toarcian boundary from Hawsker Bottoms. (a) Percentage carbonate (%CaCO3) data. (b) Percentage total organic carbon (%TOC) record. (c) Carbon-isotope record with δ13Corg and δ13Cwood data shown. Graphic log of the Hawsker Bottoms section, ammonite zones and bed numbers as for Figure 3. Long dashed line = Pliensbachian–Toarcian boundary. For data table see online Appendix Table A1, at http://multimediafluids.uam.elogim.com/journals/geo.
Samples of Late Pliensbachian age are taken from the Cleveland Ironstone Formation (Bed 38 to Bed 43 of Howarth, Reference Howarth1955). Toarcian samples are taken from the Grey Shale Formation (Bed 1 to Bed 19 of Howarth, Reference Howarth1973), which were collected to a point just above the top of the Red Nodule Beds (Beds 7 to Bed 17, Fig. 2). These correspond to the Pleuroceras hawskerense Subzone within the Pleuroceras spinatum Zone, through the P. paltum and lower D. clevelandicum subzones of the D. tenuicostatum Zone. Stratigraphic level and ammonite biostratigraphy were determined by reference to Howarth (Reference Howarth1955, Reference Howarth1973, Reference Howarth1991) and Howard (Reference Howard1985) and are reported relative to an arbitrary level within the ‘Pseudopecten bed’ (Bed 28 of Howarth, Reference Howarth1973) in the lower P. paltum Subzone, approximately 50 cm above the Pliensbachian–Toarcian boundary. The position of the stage boundary in Yorkshire is fixed at a level between the stratigraphically highest Pleuroceras and the stratigraphically lowest Dactylioceras, and is placed at the base of the Sulphur Band (Bed 26) following Howarth (Reference Howarth1973, Reference Howarth1991).
2.b. Peniche, Portugal
The section at Peniche (Fig. 1), where the negative carbon-isotope event at the stage boundary was first identified, is the sole candidate for Global Stratotype Section and Point (GSSP) for the base of the Toarcian (Duarte et al. Reference Duarte, Perilli, Dino, Rodrigues and Paredes2004; Elmi, Reference Elmi2006). This section represents an exposed fragment of the Lusitanian Basin, with a sedimentary history ranging from the mid-Triassic to late Callovian (Duarte, Reference Duarte1997). The Lemede and Cabo Carvoeiro formations that span the Pliensbachian–Toarcian boundary were deposited in a marine setting, leading to the deposition of hemipelagic, carbonate-dominated, coccolith-bearing marls and limestones (Hesselbo et al. Reference Hesselbo, Jenkyns, Duarte and Oliveira2007). The excellent coastal exposure of sediments, only weakly affected by diagenesis, has allowed the generation of a continuous high-resolution chemostratigraphic profile.
2.c. Correlation between Yorkshire and Portugal
A major issue, when comparing datasets from widely spaced sections that belong to different faunal provinces, is achieving an accurate and high-resolution stratigraphic correlation between the two locations. Correlation is hampered for some strata of Early Jurassic age by the fact that different ammonite fauna from distinct geographic locations are used to construct regional biostratigraphic schemes. Correlation between Peniche and Hawsker Bottoms requires temporal control between the mixed Tethyan/Boreal (Peniche section) and the Boreal NW European (Hawsker section) domains. Although some of the crucial ammonite taxa are present in both regions during the Late Pliensbachian–Early Toarcian (P. spinatum for example), some ammonite zones are defined on the basis of different taxa. Here, the Boreal D. tenuicostatum Zone is taken as broadly age-equivalent to the D. polymorphum Zone, and the overlying Boreal falciferum Zone equivalent to the levisoni Zone in Portugal (Page, Reference Page2004; Cecca & Macchioni, Reference Cecca and Macchioni2004). The Pliensbachian–Toarcian boundary at Peniche (Fig. 4) is placed at the base of Bed 15e, 22 cm below the top of the Lemede Formation (Elmi, Reference Elmi2006) and has been proposed as the GSSP for the base of the Toarcian. Bed 15e has the lowest occurrence of the Dactylioceras (Eodactylites) at this location and marks the base of the polymorphum Zone. The first occurrence of the Dactylioceras in NW Europe is deemed to be time-equivalent, although in Yorkshire the species is indeterminate (Howarth, Reference Howarth1991). It should be noted, however, that Dactylioceras (Eodactylioceras) has been reported co-occurring with the typically Late Pliensbachian P. hawskerense in the western Carpathians (Rakus, Reference Rakus1995; Elmi, Reference Elmi2006).
3. Methods and materials
3.a. Location of samples
Sixty-five powdered bulk samples from Hawsker Bottoms were collected using a battery-operated hand-drill at a resolution of between 2 cm and 4 cm over the key interval (between −20 cm and +170 cm, either side of the stage boundary), and at a lower resolution in the uppermost P. hawskerense Subzone (every 10 to 20 cm). This collection was supplemented with twenty-two hand specimens taken from higher in the section (Beds 3 to 19: Fig. 2) to extend the record further into the P. paltum and D. clevelandicum subzones. Five macroscopic wood samples were collected from the exposure at Hawsker Bottoms, all from the lower P. paltum Subzone of the Early Toarcian. Some of these wood samples are now preserved as coal, while others have retained their original woody texture. Five additional belemnites from ~ 1 m above the Pliensbachian–Toarcian boundary (lower polymorphum Zone) at Peniche were also collected and analysed to supplement the existing chemostratigraphic database.
3.b. CaCO3, TOC and carbon isotopes
The weight percentage of carbonate (%CaCO3) and total organic carbon (%TOC) present in each sample were determined using a Strohlein Coulomat 702, by comparing the total carbon content of bulk sediment with samples whose organic matter had been removed by roasting at 420 °C (details in Jenkyns, Reference Jenkyns1988). Drilled bulk samples were analysed directly, whereas hand specimens and wood specimens were first crushed in an agate pestle and mortar. To obtain δ13Corg values, bulk sediment samples were first decarbonated by reacting ~ 2 g of sediment in a plastic test-tube with 3M HCl in a water bath at ~ 60 °C. After two hours, the solute was decanted and the process repeated twice more with fresh acid. The sediment was then washed repeatedly with de-ionized water until neutrality was reached. The wood samples were first cleaned using a scalpel to remove any sediment or pyrite and then decarbonated in the same manner as the bulk sediment. The carbon-isotope ratios of the organic matter in the decarbonated samples were determined using a Europa Scientific Geo 20–20 mass spectrometer connected to a Carlo Erba 1108 elemental analyser. The samples were referred to a ‘Nylon 66’ standard (δ13Cnylon = −26.16 ± 0.21 ‰), and the results expressed in δ notation, as a per mil (‰) deviation relative to the Vienna Pee Dee Belemnite (VPDB) standard. The calibration was obtained by analysing two standards between every seven sediment samples. Results are accurate to better than ±0.2 ‰, based on repeat standard results.
3.c. Belemnite samples
Belemnites were washed in distilled water and cleaned using a scalpel to remove any indurated sediment or pyrite from the surface. Samples were fragmented to ~ 3 mm sized pieces, before being cleaned using first 0.6 M then 0.3 M HCl in a sonic bath. The fragments were inspected by eye to identify the least altered samples and pieces of < ~ 1 mm size were selected. Calcite from the exterior, apical line, or any dark-coloured growth bands was avoided, as such material is considered to be most prone to diagenetic alteration (Sælen, Reference Sælen1989; McArthur et al. Reference McArthur, Donovan, Thirlwall, Fouke and Mattey2000; Gomez, Goy & Canales, 2008). The δ13Cbelemnite data were generated by crushing selected sub-samples to a fine powder in the pestle and mortar, which were then treated with hydrogen peroxide and acetone and dried at 50 °C for at least one hour. The samples were analysed using a Prism II mass spectrometer with an on-line VG Isocarb common acid-bath preparation system. In the instrument, samples were reacted with purified phosphoric acid at 90 °C. Calibration was made daily using the Oxford in-house (NOCZ) Carrara marble standard. Reproducibility of replicated standards is better than 0.2 ‰. Each belemnite was analysed three times for δ13Ccarbonate and the mean value of the three determinations is reported.
4. Results
4.a. Percentage CaCO3 and TOC, Hawsker Bottoms
The sediments from the studied interval at Hawsker Bottoms are generally low in CaCO3 with an average value of 3.96% (Fig. 2a). The higher resolution data, from 20 cm below the Pliensbachian–Toarcian boundary to 170 cm above this datum, suggest a possible cyclicity in CaCO3 values with a mean spacing of 52 cm between peaks in carbonate content. The sediments at Hawsker Bottoms have varying %TOC values with an average value of 1.38%, reaching maximum values of 5.89% in the Sulphur Band (Bed 26; Fig. 2b). Highest %TOC values coincide with the darkest and most strongly laminated mudstones.
4.b. Bulk organic-matter carbon-isotope record, Hawsker Bottoms
A complex negative δ13Corg excursion, comprising two negative peaks, each with a magnitude of ~ −2.5 ‰and reaching minimum values of −28.4 ‰, can be seen in the Pliensbachian–Toarcian boundary interval at Hawsker Bottoms (Fig. 2c). The first component of the CIE occurs in the upper P. hawskerense Subzone in the latest Pliensbachian, culminating in a value of −28.4 ‰ in the Sulphur Band. The earliest Toarcian samples (base of P. paltum Subzone) are characterized by a return to pre-excursion values of ~ −26 ‰, which are then followed by the second negative CIE with a peak in Bed 2 at a minimum value of ~ −27.8 ‰. The return to more positive values of ~ −25.5 ‰ in Bed 4 persists until the onset of the larger negative carbon-isotope excursion in the late D. tenuicostatum Zone that is associated with the Early Toarcian OAE (Fig. 3). The entire boundary negative excursion, from pre-excursion values of ~ −26 ‰ in the P. hawskerense Subzone to post-excursion values of ~ −25.5 ‰ in the P. paltum Subzone, occurs over 3.6 m of section.
Comparison between carbon-isotope records at Hawsker Bottoms, Yorkshire and Peniche, Portugal. Existing Hawsker Bottoms δ13Corg record (P. paltum to H. falciferum subzones) from Cohen et al. (Reference Cohen, Coe, Harding and Schwark2004) and Kemp et al. (Reference Kemp, Coe, Cohen and Schwark2005), integrated with new high-resolution dataset from upper P. hawskerense to base of D. clevelandicum subzones (P. spinatum – Pleuroceras spinatum, P. hawskerense – Pleuroceras hawskerense, P. paltum – Protogrammoceras paltum, D. cl – Dactylioceras clevelandicum, D. ten – Dactylioceras tenuicostatum, D. semi – Dactylioceras semicelatum, C. exaratum – Cleviceras exaratum, H. falciferum – Harpoceras falciferum, Cl. Ironstone – Cleveland Ironstone). δ13Corg record compared with carbonate (δ13Ccarbonate) and wood data (δ13Cwood) from Peniche, Portugal (Hesselbo et al. Reference Hesselbo, Jenkyns, Duarte and Oliveira2007) and with brachiopod calcite data from Peniche (δ13Cbrachiopod) from Suan et al. (Reference Suan, Mattioli, Pittet, Mailliot and Lécuyer2008a). Graphic log of Hawsker Bottoms section based on field observations and adapted from Cohen et al. (Reference Cohen, Coe, Harding and Schwark2004). Graphic log of the Peniche section adapted from Hesselbo et al. (Reference Hesselbo, Jenkyns, Duarte and Oliveira2007) (Plien – Pliensbachian stage, P. spin – Pleuroceras spinatum Subzone). Bed numbers and ammonite zonation at Hawsker Bottoms taken from Howarth (Reference Howarth1955) (bracketed bed numbers) and Howarth (Reference Howarth1973) (unbracketed bed numbers). Portuguese ammonite biostratigraphy and bed numbers taken from Duarte (L. V. Duarte, unpub. Ph.D. thesis, Universidade de Coimbra, 1995) (Bed number 1) and Mouterde (Reference Mouterde1955) (Bed number 2). Dashed numbered lines in Yorkshire section represent approximate location of extinction steps in Jurassic benthic fauna, as described in Harries & Little (Reference Harries and Little1999). Correlation between sections is based on the assumed age equivalence of the NW European tenuicostatum Zone and the Tethyan polymorphum Zone, and on GSSP placement of Pliensbachian–Toarcian boundary in Peniche (Elmi, Reference Elmi2006).
Importantly, there is an inverse correlation between %TOC values and δ13Corg values in the Toarcian part of the section. Beds with higher %TOC have the most negative δ13Corg values. In the P. paltum Subzone, the beds with the most negative δ13Corg values tend to be the more laminated, dysoxic mudstones such as the Sulphur Band and Bed 2.
4.c. Wood carbon-isotope record, Hawsker Bottoms
The δ13C values of fossil-wood samples have a range from ~ −24 to ~ −26 ‰, which is in line with previous determinations from the Yorkshire sections (Hesselbo et al. Reference Hesselbo, Gröcke, Jenkyns, Bjerrum, Farrimond, Morgans Bell and Green2000). The small number of samples makes comparison with the δ13Corg curve from the bulk sediment difficult, but it is noted that the lowest carbon-isotope values seen in the wood (~ −26 ‰) correspond to the negative excursion (~ −28.5 ‰) observed in the δ13Corg record within the Sulphur Band (Fig. 2c). The persistent ~ 2 ‰ offset in carbon-isotopes between bulk organic-matter and fossil-wood records has been seen in other examples, in which the wood is consistently isotopically heavier than the bulk organic matter (Hesselbo et al. Reference Hesselbo, Gröcke, Jenkyns, Bjerrum, Farrimond, Morgans Bell and Green2000).
4.d. Belemnite carbon-isotope record, Peniche
The δ13C values of five belemnites from above the Pliensbachian–Toarcian boundary at Peniche, Portugal, were analysed and the results combined with existing data from Hesselbo et al. Reference Hesselbo, Jenkyns, Duarte and Oliveira2007 (Fig. 4a). The new samples are situated ~ 1 m above the stage boundary, with the majority clustered ~ 25 cm above the maximum negative carbon-isotope excursion seen in the accompanying δ13Ccarbonate record. δ13Cbelemnite values from these samples range from ~ −0.6 to ~ +0.7 ‰ and are generally more negative than nearly all other Late Pliensbachian–Early Toarcian belemnites from the same section reported in Hesselbo et al. (Reference Hesselbo, Jenkyns, Duarte and Oliveira2007). However, as the belemnite record at Peniche does not represent a mono-specific assemblage, it may be that a proportion of the variation in the δ13Cbelemnite values could be attributed to inter-species variation. It is difficult to determine whether the clear CIE excursion seen in the carbonate record has also been captured in the belemnite record, because no belemnites that coincide precisely with the lowest part of the negative CIE have been collected at Peniche.
(a) Close-up of boundary sequence at Peniche showing δ13Ccarbonate and δ13Cbelemnite data taken from Hesselbo et al. (Reference Hesselbo, Jenkyns, Duarte and Oliveira2007), with new δ13Cbelemnite data points from just above the peak negative excursion interval. Close-up of δ13Cwood data also from Hesselbo et al. (Reference Hesselbo, Jenkyns, Duarte and Oliveira2007) and δ13Cbrachiopod data from Suan et al. (Reference Suan, Mattioli, Pittet, Mailliot and Lécuyer2008a) is shown in (b). Bed numbering as for Figure 3. For new δ13Cbelemnite data see online Appendix Table A2, at http://multimediafluids.uam.elogim.com/journals/geo.
5. Discussion
5.a. Origin of the negative CIE at the stage boundary
The newly recognized complex excursion at the Pliensbachian–Toarcian boundary has several similarities with the negative CIE that is well known from the Early Toarcian OAE, so that the proposed mechanisms to explain its presence will be similar. The various models put forward to explain the perturbation at the OAE have recently been discussed in Cohen, Coe & Kemp (Reference Cohen, Coe and Kemp2007) and can be divided into three broad hypotheses: (1) overturning or upwelling of a stratified water mass (‘Kuspert model’), (2) massive dissociation of methane clathrates and (3) mechanisms directly relating to magmatism in the Karoo–Ferrar Large Igneous Province (LIP). The relative strengths and weaknesses of these various theories are discussed at length in Cohen, Coe & Kemp (Reference Cohen, Coe and Kemp2007), and in the subsequent discussions in McArthur et al. (Reference McArthur, Cohen, Coe, Kemp, Bailey and Smith2008), and will not be discussed in great detail here.
For the Pliensbachian–Toarcian boundary excursion, the Küspert model (Küspert Reference Küspert, Einsele and Seilacher1982), in which isotopically light respired CO2 is recycled in a restricted basin, is consistent with the apparent absence of the excursion in low-resolution belemnite records from Hawsker Bottoms and the inverse relationship between the %TOC and δ13Corg (Fig. 2b, c) (cf. Sælen et al. Reference Sælen, Tyson, Talbot and Telnæs1998; Van de Schootbrugge et al. Reference Van de Schootbrugge, McArthur, Bailey, Rosenthal, Wright and Miller2005; Wignall et al. Reference Wignall, McArthur, Little and Hallam2006; Hesselbo et al. Reference Hesselbo, Gröcke, Jenkyns, Bjerrum, Farrimond, Morgans Bell and Green2000; Jenkyns et al. Reference Jenkyns, Jones, Gröcke, Hesselbo and Parkinson2002; Hesselbo et al. Reference Hesselbo, Jenkyns, Duarte and Oliveira2007). In contrast, the organic geochemical evidence of Van Breugel et al. (Reference Van Breugel, Baas, Schouten, Mattioli and Sinninghe Damsté2006a,Reference Van Breugel, Schouten, Paetzel and Sinninghe Damstéb) suggests that recycled CO2 from anoxic deep waters plays only a negligible role in influencing the δ13C of phytoplankton in the overlying waters and the δ13Corg record preserved in the sediment.
Of course, the δ13Corg value represents the isotopic value of bulk organic-matter in the sediment, and therefore any changes in this value may reflect a change in the relative contribution of one type of organic-matter over another. The correlation between the most organically enriched beds and the most isotopically depleted samples at Hawsker Bottoms, suggests the issue may merit further investigation. However, the similarity in the carbon-isotope records of the Late Pliensbachian to Early Toarcian interval between the Peniche and Hawsker Bottoms sections (Fig. 3), in materials ranging from bulk organic-matter and bulk carbonate to fossil-wood and brachiopod calcite, strongly suggests expression of the carbon-cycle perturbation in at least a regional context. The boundary excursion at Peniche is recorded primarily in δ13Ccarbonate and occurs entirely within carbonate facies, and therefore cannot be explained merely by changes in local organic-matter provenance.
The presence of the stage boundary CIE in the fossil-wood record from Peniche indicates that the perturbation must have affected the atmospheric as well as the marine carbon reservoir. The methane dissociation hypothesis has in the past been the preferred explanation of many authors for the Early Toarcian falciferum Zone CIE, because it would have affected the atmospheric δ13C of CO2 and can also explain other reported phenomena such as high global temperatures and an increased intensity in weathering inferred from the Os-isotope record (Hesselbo et al. Reference Hesselbo, Gröcke, Jenkyns, Bjerrum, Farrimond, Morgans Bell and Green2000; Beerling, Lomas & Gröcke, Reference Beerling, Lomas and Gröcke2002; Jenkyns, Reference Jenkyns2003; Cohen et al. Reference Cohen, Coe, Harding and Schwark2004; Kemp et al. Reference Kemp, Coe, Cohen and Schwark2005; Cohen, Coe & Kemp, Reference Cohen, Coe and Kemp2007). The apparent abrupt onset of the stage boundary CIE at Hawsker Bottoms (Fig. 2c), and the initiation of an extinction step near the same level, supports the notion of a catastrophic event such as release of methane hydrate. The volume of clathrate in the Early Jurassic and its potential susceptibility to destabilization is, however, a major unknown factor. Neither the speed of onset nor the duration of the stage-boundary CIE can be determined, due to the unconstrained sedimentation rate at this section during the Early Jurassic period. Attempts to quantify the sedimentation rate, and therefore the timing of this event, are frustrated by a lack of age constraint. Attempts to quantify relative time using strontium-isotopes are controversial, and the errors associated with utilizing the available radiometric dates are larger than the likely duration of the event itself (McArthur et al. Reference McArthur, Donovan, Thirlwall, Fouke and Mattey2000). Therefore, at present, it is difficult to use the relative duration of the event to decide which of the suggested mechanisms is most likely to have initiated the CIE.
Recently, comparisons have been made between the Early Toarcian falciferum Zone CIE and the Palaeocene–Eocene Thermal Maximum (PETM: ~ 55.8 Ma), in terms of their geochemical signatures and possible causative mechanisms (Cohen, Coe & Kemp, Reference Cohen, Coe and Kemp2007). The Palaeocene–Eocene Thermal Maximum is marked by a pronounced negative CIE of up to −2.5 ‰ in bulk carbonate, and up to −7 ‰ in bulk organic-matter (e.g. Kennett & Stott, Reference Kennett and Stott1991; Pagani et al. Reference Pagani, Pedentchouk, Huber, Sluijs, Schouten, Brinkhuis, Sinninghe Damsté and Dickens2006). The event is also associated with benthic faunal extinctions, extensive carbonate dissolution and evidence for a large and rapid global warming event (e.g. Zachos et al. Reference Zachos, Röhl, Schellenberg, Sluijs, Hodell, Kelly, Thomas, Nicolo, Raffi, Lourens, McCarren and Kroon2005; Sluijs et al. Reference Sluijs, Schouten, Pagani, Woltering, Brinkhuis, Sinninghe Damsté, Dickens, Huber, Reichart, Stein, Matthiessen, Lourens, Pedentchouk, Backman and Moran2006, Reference Sluijs, Bowen, Brinkhuis, Lourens, Thomas, Williams, Haywood, Gregory and Schmidt2007; Bowen et al. Reference Bowen, Bralower, Delaney, Dickens, Kelly, Koch, Kump, Meng, Sloan, Thomas, Wing and Zachos2006). The negative CIE associated with the Palaeocene–Eocene Thermal Maximum has also been attributed to a sudden methane clathrate dissociation event, due to its apparently short duration (~ 200 ka) and associated temperature spike (e.g. Dickens et al. Reference Dickens, O'Neil, Rea and Owen1995; Dickens, Reference Dickens2000; Thomas et al. Reference Thomas, Zachos, Bralower, Thomas and Bohaty2002; Katz et al. Reference Katz, Cramer, Mountain, Katz and Miller2001). However, the potential role of methane clathrate release at the Palaeocene–Eocene Thermal Maximum has recently been called into question (e.g. Higgins & Schrag, Reference Higgins and Schrag2006). The magnitude of the CIE constrains the maximum amount of biogenic methane (δ13C = ~ −60 ‰) which could have been released into the ocean–atmosphere system during this event, to < 2000 GtC (Gigatons of carbon) (e.g. Dickens, Reference Dickens2000). However, the estimated extent of carbonate dissolution in the oceans during the event requires a far greater carbon input than this (~ 4500 GtC), thus suggesting that biogenic methane cannot be solely responsible for the perturbation (Zachos et al. Reference Zachos, Röhl, Schellenberg, Sluijs, Hodell, Kelly, Thomas, Nicolo, Raffi, Lourens, McCarren and Kroon2005). These considerations may have a bearing on the interpretation of the Early Toarcian, by suggesting that the Jurassic CIEs may also have been triggered by more than one casual mechanism.
It is well documented that the Karoo–Ferrar LIP on southern Gondwana (centred on modern day South Africa and Antarctica) reached a period of peak volcanic activity at c. 183 Ma (Pálfy & Smith, Reference Pálfy and Smith2000; Pálfy, Smith & Mortensen, Reference Pálfy, Smith, Mortensen, Koeberl and MacLeod2002), and was therefore broadly coincident with the Late Pliensbachian–Early Toarcian CIEs. A third hypothesis suggests that Karoo-Ferrar sills intruded Permian organic-rich sediments or coals, releasing large amounts of isotopically depleted greenhouse gases to the Toarcian atmosphere (McElwain, Wade-Murphy & Hesselbo, Reference McElwain, Wade-Murphy and Hesselbo2005; Svensen et al. Reference Svensen, Planke, Chevallier, Malthe-Sørenssen, Corfu and Jamtveit2007). A broad range of dates for peak lava emplacement from c. 184 to c. 178 Ma has been calculated for the Karoo region, indicating that the stage boundary lies within the main extrusive phase of the LIP (Jourdan et al. Reference Jourdan, Féraud, Bertrand, Kampunzu, Tshoso, Watkeys and Le Gall2005, Reference Jourdan, Féraud, Bertrand, Watkeys and Renne2008). It is possible, therefore, that an earlier pulse of thermogenic methane from this region could be partly responsible for the stage boundary CIE seen at Yorkshire and Peniche. Recent work by Gröcke et al. (Reference Gröcke, Rimmer, Yoksoulian, Cairncross, Tsikos and van Hunen2009) that calls into question the causative role of thermogenic methane from the Karoo basin in the falciferum Zone CIE may be of limited relevance because it considers only the case of dyke intrusion into coal, rather than sill intrusion into coal or black shale, the latter being likely to be very much more effective in the generation of large amounts of methane. As suggested in Suan et al. (Reference Suan, Pittet, Bour, Mattioli, Duarte and Mailliot2008b), the negative CIE might be attributed directly to volcanic outgassing of isotopically light CO2. The average carbon-isotopic value of the mantle is generally taken to be only ~ −6 ‰, hence requiring the release of unrealistically large volumes of volcanogenic CO2 in order to cause such a large perturbation in the carbon cycle. However, mantle xenolith compositions are compatible with an isotopically heterogeneous mantle that may contain reservoirs of very depleted carbon (~ −30 ‰), and flood basalts can exhibit carbon-isotopic values much lighter (~ −24 ‰) than the assumed average mantle value (Deines, Reference Deines2002; Hansen, Reference Hansen2006). In addition, other authors have identified fractionation processes during degassing as a potential source of isotopically light carbon in basalts (e.g. Mattey, Reference Mattey1991).
The suggestion that Large Igneous Province volcanism played a direct causative role in initiating the stage boundary and Early Toarcian carbon-isotope excursions has, however, been suggested to be incompatible with the Milankovitch cyclicity reported during the falciferum Zone Oceanic Anoxic Event excursion (Kemp et al. Reference Kemp, Coe, Cohen and Schwark2005). This assertion is put forward on the basis that it is difficult to envisage how a mantle-derived source of depleted carbon could cause fluctuations in the carbon cycle that were apparently paced in time with orbital cycles. This line of reasoning may be countered by supposing that a longer-term global isotopic shift could be modified by Milankovitch-controlled local processes. Notwithstanding this debate, since no such cyclicity has yet been identified within the carbon-isotope record at the stage boundary, and the duration and speed of onset of the event is unclear, a volcanogenic origin for this CIE cannot be discounted.
A reconstructed palaeotemperature record from the Late Pliensbachian–Early Toarcian interval could be used to decipher which mechanism is the most applicable to the stage boundary CIE, since massive methane dissociation should be associated with a temperature maximum (as is seen at the Palaeocene–Eocene Thermal Maximum). Oxygen-isotope ratios from belemnite calcite can be interpreted in terms of changes in palaeotemperature, although care must be taken to consider the many other factors which can affect this ratio, such as taxon-specific effects and unknown initial seawater chemistry. The δ18Obelemnite records from central and northern Spain suggests a gradual decrease in δ18O values (~ 0 ‰) in the latest Pliensbachian P. spinatum Subzone, to between −1.5 and −2.5 ‰ in the H. serpentinus Zone (which is considered to be the southern European equivalent of the Boreal falciferum Zone) (Gómez, Goy & Canales, Reference Gómez, Goy and Canales2008). This change is interpreted as a gradual temperature rise from ~ 13 °C in the latest Pliensbachian to peak temperatures of ~ 25 to 28 °C during the late H. serpentinum Zone, thus approximately coinciding with the negative CIE of the Toarcian OAE. Such a temperature trend is broadly in agreement with other δ18Obelemnite data from Spain, which show a similar change from ~ 13 to ~ 25 °C across the same time interval (Rosales, Quesada & Robles, Reference Rosales, Quesada and Robles2004a). Other work using belemnite oxygen isotopes from Yorkshire, however, shows an opposite trend across the stage boundary itself, with δ18O values changing from ~ −4 ‰ in the latest Pliensbachian to ~ −1 ‰ in the P. paltum Subzone of the Early Toarcian (Bailey et al. Reference Bailey, Rosenthal, McArthur, Van de Schootbrugge and Thirlwall2003), which could be interpreted as a drop in temperature. This disparity in oxygen-isotope values during the Late Pliensbachian–Early Toarcian has been interpreted in terms of a north–south salinity gradient within the North European (Boreal) seaway at this time (Rosales, Quesada & Robles, Reference Rosales, Quesada and Robles2004b, Reference Rosales, Quesada and Robles2006). It would seem, therefore, that inferred palaeotemperature trends from the Early Jurassic do not currently provide unequivocal evidence in support of one mechanism over another to explain the stage boundary CIE.
6. Conclusions
A sharp negative δ13Corg isotope excursion of somewhat complex character and with a magnitude of ~ −2.5 ‰ is present in sedimentary successions across the Pliensbachian–Toarcian boundary in NE Yorkshire, England. This excursion can be correlated with a similar phenomenon of lesser magnitude in the carbonate sedimentary record from coastal Portugal, suggesting a perturbation in the carbon cycle synchronous on at least a regional scale. This excursion will likely provide a useful chemostratigraphic marker for the Pliensbachian–Toarcian boundary. The presence of the excursion in materials ranging from bulk carbonate and bulk organic matter, to brachiopod calcite and fossil wood, suggests the perturbation must have affected the entire ocean–atmosphere system and should be registered in appropriate materials in coeval strata in all parts of the world. Indeed, coincidence with the first phase of extinction in marine benthos at the stage boundary, well before the main phase of anoxia associated with the Early Toarcian OAE, further suggests the event was not simply a localized anomaly. Further high-resolution work across the stage boundary from sites outside of the European epicontinental shelf region is needed to determine the probable cause.
Acknowledgements
Organic-carbon isotope data were generated at the Department of Archaeology, Oxford University, by Peter Ditchfield, for whose assistance we are grateful. Thanks to Norman Charnley for his efforts in generating carbonate carbon-isotope data and to Steve Wyatt for laboratory assistance. Thanks to Julian Littler for fieldwork assistance and to Stuart Robinson for scientific discussions that improved this manuscript. We are grateful to two anonymous reviewers for their detailed comments on this manuscript.
