No evidence for thermogenic methane release in coal from the Karoo-Ferrar large igneous province

Earth and Planetary Science Letters 277 (2009) 204–212

Contents lists available at ScienceDirect

Earth and Planetary Science Letters

j ourna l homepage: www.e lsev ie r.com/ locate /eps l

No evidence for thermogenic methane release in coal from the Karoo-Ferrar largeigneous province

Darren R. Gröcke a,⁎, Susan M. Rimmer b, Lois E. Yoksoulian b, Bruce Cairncross c,Harilaos Tsikos d, Jeroen van Hunen e

a Department of Earth Sciences, Durham University, Science Laboratories, Durham, DH1 3LE, UKb Department of Earth and Environmental Sciences, University of Kentucky, Lexington KY 40506-0053, USAc Department of Geology, University of Johannesburg, PO Box 524, Auckland Park 2006, Gauteng, South Africad Department of Geology, Rhodes University, PO Box 94, Grahamstown, 6140, South Africae Department of Earth Sciences, Durham University, Science Laboratories, Durham, DH1 3LE, UK

⁎ Corresponding author.E-mail addresses: [email protected] (D.R. Grö

(S.M. Rimmer), [email protected] (L.E. Yoksoulian), [email protected]@ru.ac.za (H. Tsikos), jeroen.van-hunen@durham

0012-821X/$ – see front matter © 2008 Elsevier B.V. Adoi:10.1016/j.epsl.2008.10.022

a b s t r a c t

Article history:

Editor: M.L. Delaney

Keywords:carbon isotopesthermogenic methanevitrinite reflectancedyke intrusionKaroo-Ferraroceanic anoxic eventToarcian

oceanic anoxic event (T-OAE) and concurrent negative carbon-isotope (δ13C)excursion have recently been attributed to either the release of methane (CH4) clathrates or thermogenic CH4

gas associated with the Karoo-Ferrar large igneous province (LIP) into coals and organic-rich shales. 12C-enriched thermogenic CH4 production associated with the Karoo-Ferrar would result in residual materialbeing 12C-depleted nearer the intrusions. In this study, geochemical analyses (carbon isotopes, volatile matter(VM), vitrinite reflectance (Ro)) are reported for two coal transects associated with dykes intruding the No. 4Lcoal in the Highveld Coalfield, Karoo Basin, South Africa. VM decreases from over 35% to around 15% in onetransect, and the second transect shows a less pronounced decrease (from N25% to ~16%). Accompanying thedecrease in VM content is an increase in Ro from background levels of around 0.7% to over 4% adjacent to thedyke; used as a palaeo-geothermometer, Ro values indicate background temperatures of ~100 °C increasingto N300 °C close to the contact. Despite changes in VM and Ro, there are no significant changes in δ13C,certainly not of the magnitude that would be expected associated with large-scale thermogenic CH4

generation. These and other Gondwanan coals have low vitrinite and liptinite contents (components moreprone to CH4 generation), in part explaining the modest decreases in VM adjacent to the dykes. This,combined with the relatively narrow metamorphic aureole surrounding the intrusions and the likelihoodthat at least some of the volatiles generated by the intrusion were trapped as coalbed CH4 or condensed aspyrolytic carbon, suggests only limited CH4 release. In addition, based on original estimates of moisturecontents in these coals and the depth at time of intrusion (1,000–2,000 m) the dykes would have lost most oftheir energy heating and evaporating water, thus having very little remaining energy to generatethermogenic CH4.

© 2008 Elsevier B.V. All rights reserved.

1. Introduction

Over the past quinquennium there has been an increased effort tounravel the cause and consequences of the Early Jurassic Toarcianoceanic anoxic event (T-OAE). Renewed interest has been generated bythe interpretationput forwardbyHesselboet al. (2000) as to the cause ofthe negative δ13C excursions associatedwith the onset of the T-OAE andwidespread distribution of organic-rich black shales. Hesselbo et al.(2000) generated a high-resolution bulk δ13C organic curve through theT-OAE from the Hawsker Bottoms section in Yorkshire, which alsocontained Jet – a form of fossilized wood. Both sample types recorded

cke), [email protected] (B. Cairncross),.ac.uk (J. van Hunen).

ll rights reserved.

the negative δ13C excursion, suggesting that the cause of the excursionwas great enough to affect both the oceanic and atmospheric reservoirs,and therefore global in nature. An additional terrestrial record from ashallow marine-succession at Bornholm, Denmark, also providedunequivocal evidence for the terrestrial realm recording this carbonisotopic excursion (Hesselbo et al., 2000).

More recently, Hesselbo et al. (2007) have shown that this negativeδ13C excursion associated with the T-OAE has been recorded in fossilwood and bulk marine carbonate from Peniche, Portugal – a GSSPcandidate for the basal Toarcian. An even higher-resolution study on theHawsker Bottoms section by Kemp et al. (2005) indicates that thenegative anomaly is more complex and actually includes three distinct,sharp negative excursions, akin to the Paleocene–Eocene ThermalMaximum (PETM). Hence, both Hesselbo et al. (2000, 2007) and Kempet al. (2005) have proposed that the release ofmethane gashydrateswasa cause and/or component of the T-OAE that led to highly negative δ13C

205D.R. Gröcke et al. / Earth and Planetary Science Letters 277 (2009) 204–212

signatures. Although there is some dispute over the cause of the T-OAE,due to the absence of this negative δ13C excursion in belemnites (seework by van de Schootbrugge et al., 2005) this is primarily due to a‘belemnite gap’ in this time interval (Hesselbo et al., 2007). Recentevidence by Suan et al. (2008a) and Hermoso et al. (2009) suggests thatthe negative δ13C excursion is also recorded in brachiopods andcarbonate size fractions respectively. In addition, multiple lines ofevidence are emerging suggesting major environmental changesassociated with the T-OAE, such as an increase in oceanic temperaturesandhigh atmospheric CO2 levels (McElwain et al., 2005) andweatheringflux to the oceans (Cohen et al., 2004; Waltham and Gröcke, 2006).

Of particular importance is the interpretation put forward byMcElwain et al. (2005), using the data from Hesselbo et al. (2000),which states that the negative δ13C excursion was caused by intrusion oftheKaroo-Ferrar large igneous province (LIP) into coalfields. Svensen et al.(2007) suggest that contact metamorphism by dykes and sills associatedwith the Karoo-Ferrar LIP of organic-rich sediments, such as theWhitehillFormation,were apotential sourceof thermogenicmethane, as evidencedfromvitrinite reflectance and total organic carbon contents. Svensen et al.(2007) extrapolate their data to predict basin-wide gas generation, eventhough theyclearly indicate thismaynotbe anaccurate assumption.Morerecently, Summons et al. (2008) suggest that themajority of theWhitehillFormation is unaffected by dolerite magmatism, thus exhibiting imma-ture organicmatter. Therefore, the basin-wide extrapolation of Svensen etal. (2007) requires amore accuratemodel to account for regional variationin contact metamorphism and aureole effects. In addition, according toKemp et al. (2005) and Suan et al. (2008b), the negative carbon-isotopeexcursions associated with the T-OAE are cyclic and very rapid (b20 kyr),and thus there is no current reasonwhy dyke and/or sill intrusionwouldbe tuned to a cyclic parameter. An obvious target for creating massiveamounts of CO2 and CH4 to produce the isotopic excursion at the T-OAEwould be through the thermal alteration of coal, which has far greatertotal organic carbon contents (typically N50%) than the WhitehillFormation and thus would require less material to be volatized.Notwithstanding these issues, if the model of Svensen et al. (2007) is tobe applied (i.e., that contact metamorphism caused the transformation of

Fig. 1. Locality map of the coalfields of South Africa within and surrounding the main Karoo Bwere taken is noted by the square box under Secunda.

organic-rich sediments (shales andcoal) to isotopically light CO2andCH4),then we should expect to see a transformation in Karoo Basin coals.

Experimental studies have demonstrated that during thermogenicgas production, 12C–12C bonds rupture more readily than do 13C–12Cbonds (Sackett et al.,1970); in the case of methane generation from coalat 400 °C, this leads to the evolution of 12C-enriched gas that isapproximately 25‰ lighter than the coal (i.e., ~−50‰; Sackett, 1978).Despite the large volume of carbon in coal and organic-rich shales, thisgas release can lead to 12C-depleted residues, as has been reported forboth intruded shales (Simoneit et al., 1981; Clayton and Bostick, 1986;Saxby and Stephenson, 1987) and coals (Meyers and Simoneit, 1999;Cooper et al., 2007). Thus, it can be hypothesized that if the model ofMcElwain et al. (2005) and Svensen et al. (2007) is to be applied to the T-OAE, the release of substantial thermogenic methane (12C-enriched)from organic-rich sediments and/or coals should have left a residualmaterial that is 12C-depleted. In this study, we report geochemicalanalyses (stable carbon isotopes, proximate and ultimate analyses, andvitrinite reflectance) for two coal transects in the Highveld Coalfield,Karoo Basin that are intruded by Karoo doleritic dykes. A subsequentdiscussion of the geochemical and physical evidence against substantialthermogenicmethaneproduction fromdyke intrusion into coal (or evensill intrusion in contact with organic-rich shales) is put forward.

2. Regional geology

The study site is situated close to Secunda, ~140 km southeast ofJohannesburg, in the Highveld Coalfield, northeastern main KarooBasin, South Africa (Fig. 1). In this area, SASOL (Pty) Ltd runs theworld's largest oil-from-coal extraction plant. All of southern Africa'seconomic coal is contained within the Palaeozoic strata of the KarooSupergroup (Cairncross 1989, 2001), and in particular, along thenorthern passive margin of the foreland basin (Cadle et al., 1993;Catuneanu et al., 2005). Coal rank varies from high volatile bituminousto anthracite (Snyman, 1998).

The coal seams are interbedded in the Vryheid Formation, a clasticsuccession of shale, siltstone, immature sandstone (arkosic arenites,

asin, and in particular the Highveld Coalfield. The study area in which the coal transects

Fig. 2. General stratigraphic column of the coal-bearing Vryheid Formation atMiddelbult Colliery. Modified from Hagelskamp (1987).

206 D.R. Gröcke et al. / Earth and Planetary Science Letters 277 (2009) 204–212

lithic arenites), andminor conglomerate. TheVryheid Formation forms awedge that surrounds the northeastern, northern and northwesternperimeterof theKaroo Basin that pinches out basinward (Hobday,1978).It is thickest in thewest (~500m), and thins towards the north, east andsouth. In the study area, the coal-bearing succession attains an averagethickness of 150 m (Hagelskamp et al., 1988) and the main economicseam is theNumber 4 Lower (No. 4L) coal seam; it is this coal seam fromwhich we have sampled two coal transects away from dykes (Fig. 2).Transect M was taken from No. 4L which was cut by a dyke 1.2 m inthickness that pinched out once it passed through the coal seam.Transect P was sampled where a dyke (1.5 m thick) cut through coalseam No. 4L coal face with no apparent change in thickness (Fig. 3).Palaeoenvironmental studies of this region of the Highveld Coalfield(Hagelskamp, 1987) suggest that the lower sections were deposited byglaciofluvial and glaciolacustrine processes overlain by mixed-load andbedload fluvial systems with associated intermittent peat swamps. Theage of the coal-bearing sequence is Artinskian-Kungurian – Late EarlyPermian (Cairncross, 2001).

Apart from the sedimentary succession, younger dolerites (gen-erally tholeiitic or olivine dolerites) have pervasively intruded into andthrough the entire Karoo Supergroup (Bell and Germy, 2000; Duncanand Marsh, 2006). Recent U/Pb ages derived from zircons in a sill fromthe Loeriesfontein area provide an emplacement age of 182.5±0.4 Ma(Svensen et al., 2007). However, it should be clearly stated here thatthese sills were not emplaced in one single event but as multipleintrusions, as evidenced by cross-cutting relationships between dykesand sills (Anhaeusser and Maske, 1986; van Zijl, 2006).

As a result of different episodes of dolerite intrusion, different typesof dolerite exist, as described by De Oliveira and Cawthorn (1999) fromtheMajuba Colliery, located ~80 km southeast of the present study area.Dyke intrusion for the Highveld Coalfield shows that it is much moreextensive than sill intrusion, thus dyke/coal contact has more potentialfor producing thermogenic methane and/or carbon dioxide. Whiledolerites are frequently associated with the geological successioncontaining the Permian coal seams in the Karoo Basin, it is extremelyrare for a dolerite sill to actually lie in direct contact with either thehanging wall or footwall of a coal seam (Anhaeusser and Maske, 1986).Thus, dykes would be the only source for generating thermogenicmethane from the vaporization of coal.

3. Samples and analytical methods

Two underground areas were selected at theMiddelbult Colliery inthe Highveld Coalfield at Secunda. The two sample sites are 7 km apartand were chosen because near-vertical dolerite dykes had beenexposed in the No. 4L coal seam working face. Using the general rulethat the dykes affect the adjacent coal to a distance equal to 1–2 timesthe dyke thickness (Bostick and Pawlewicz,1984; Snyman and Barclay,1989; Barker et al., 1998), samples were collected from the intrusionout to a distance that was expected to exceed that of the contactaureole. Due to the inherent hardness of the coal in the alteration halo,grab samples were collected rather than whole-seam channelsamples.

Bulk coal samples were decalcified with 3 M HCl for 6 h and thenrinsed until neutrality. The samples were then dried in an oven set at60 °C overnight. Carbon-isotope analyses were performed on aThermoFinnigan DeltaPlus XP coupled with a Costech elementalanalyzer. Up to 8 international and internal standards were analyzedmultiple times during the generation of this isotopic dataset andanalytical precision of δ13C was 0.1‰. Replicate coal analyses weretypically better than 0.2‰.

Proximate and ultimate analyses of coal samples were performed induplicate at the University of Kentucky (utilizing facilities at theKentuckyGeological Surveyand theCenter forAppliedEnergyResearch)using a LECO TGA 701 and a LECO CHN-2000 analyzer, respectively,according to standard ASTM procedures (D-3172-89 and D-3176-89,

Fig. 3. Cross sections for transect M and transect P through the No. 4L coal seam. Note the pinching out of the dyke (V symbols) in transect M: coal samples were takenwhere the dykewas ~1.2 m in thickness. The dyke at transect P was 1.5 m and showed no thinning or thickening through the outcrop.

207D.R. Gröcke et al. / Earth and Planetary Science Letters 277 (2009) 204–212

respectively). Inorganic carbonwas determined by coulometry, using anUIC Coulometrics CM 5014 equipped with an acidification module, andtotal organic carbon (TOC) was calculated by difference (%C fromultimate analysis less inorganic carbon). Mean random vitrinitereflectance analyses (% Ro, oil) (50 readings per sample)were performedusing a Zeiss Universal scope fitted with a photomultiplier tube,standardized against known glass and SiC standards; data collectionwas controlled by custom computer hardware and software. Readingswere collected on vitrinite or the coked equivalents in the thermallyaltered coals. Quantitative organic petrographywasbasedon500 countson each of two pellets per sample, using reflected light microscopy(under both white-and blue-light).

4. Results

In the transects studied, the M series shows a decrease in VM (dry,ash-free basis, daf) from over 35% to around 15%, with a slight increase(23%) directly adjacent to the dyke; transect P shows a less pronounceddecrease (from N25% to around 16%) (Tables 1 and 2 respectively).Accompanying the decreased volatile content is an increase in Ro frombackground levels of around 0.7% to over 4% adjacent to thedyke (Fig. 4).In coal samples adjacent to thedyke, a vesicular texture (devolatilizationvacuoles) is observed in the vitrinite providing additional petrographicevidence for VM loss (Fig. 5). The major maceral components arevitrinite (preservedwoody tissue) and inertinite (material that has been

Table 1Geochemical results from the M-series coal

ID Metres (m) Vm † Ash † Fc † Vm ⁎ Fc ⁎ % C % H % N

M1 0 14.7 35.2 50.1 22.7 77.3 56.7 1.5 0.8M2 0.3 11.3 26.1 62.6 15.3 84.7 64.9 1.7 0.8M3 0.5 14.3 25.1 60.6 19.1 80.9 65.1 2.0 1.1M4 0.9 15.6 22.8 61.6 20.2 79.8 67.0 2.7 1.6M5 1.3 16.3 18.6 65.2 20.0 80.0 69.8 3.2 1.9M6 1.8 21.3 23.8 54.9 28.0 72.0 65.0 3.6 1.8M7 2.3 26.2 29.3 44.5 37.0 63.0 70.6 4.0 1.8M8 2.8 22.8 19.7 57.5 28.4 71.6 62.1 3.1 1.5M9 3.3 28.3 17.7 54.0 34.4 65.6 63.9 5.5 1.6M10 4.3 27.8 10.8 61.4 31.2 68.8 68.8 4.2 1.7

Metres represent distance from dyke. Ro = vitrinite reflectance (% oil). Vit = vitrinite; Lipt =

partially combusted or possibly oxidized). The relative amounts of eachdonot change systematically approaching the dyke, but inertinite showstwo increases through the transects at the expense of vitrinite, thusresulting in a somewhat sinusoidal pattern (Fig. 6; Table 2). Coal canhave considerable lateral and vertical variability inmaceral compositionreflecting conditions at the time of deposition, thus this variability islikely to be primary in origin rather than due to any effects of theintrusion. δ13C shows a similar sinusoidal pattern (Fig. 6) possiblyreflecting, in part, this shift from higher to lower vitrinite contents:where the coal is higher in inertinite, δ13C becomes heavier which isconsistent with previous work which has shown that inertinite isheavier than vitrinite (Rimmer et al., 2006). Thus, patterns in δ13C seemto reflect a combination of maceral and maturation effects.

5. Discussion

It is generally thought that coal rank does not influence δ13C until veryhigh levels of maturity, a stability thought to reflect, in part, the fact thatlittlemethane (whichwould be depleted relative to thewhole coal) is lostuntil high rank (Sackett, 1978). At anthracite levels, δ13C values becomesignificantly heavier (McKirdy and Powell,1974; Hoefs and Frey,1976) as aresult of methane generation. For example, at high metamorphic levels(greenschist and amphibolite facies), organicmattermay have δ13C valuesof −15‰ to −10‰ compared to −28‰ to −24‰ for unmetamorphosedorganicmatter (Hoefs and Frey,1976).Minor shifts (on the order of 1–2‰)

C/N C/H % CaCO3 Ro δ13Corg TOC (wt.%) Vit % Lipt % Inert %

87.5 3.3 1.71 4.5 −22.7 55.6 37.2 0.0 62.889.5 3.2 1.30 3.6 −22.5 63.2 11.2 0.0 88.867.7 2.7 1.84 3.0 −22.3 77.8 19.8 0.0 80.249.4 2.1 1.10 2.5 −22.6 79.0 21.2 0.0 78.841.8 1.9 0.65 2.3 −23.1 61.4 40.2 0.0 59.841.3 1.5 0.61 1.2 −23.3 82.1 29.8 0.0 70.244.8 1.5 0.56 0.9 −22.9 72.3 30.0 4.0 66.048.3 1.7 0.94 0.7 −22.5 76.9 11.6 3.2 85.247.5 1.0 0.75 0.7 −23.0 81.7 27.6 3.2 69.248.4 1.4 0.22 0.7 −23.1 72.7 33.8 3.2 63.0

liptinite; Inert = inertinite. † = dry; ⁎ = dry and ash free.

Table 2Geochemical results from the P-series coal

ID Metres Vm † Ash † Fc † Vm ⁎ Fc ⁎ % C % H % N C/N C/H % CaCO3 Ro δ13Corg TOC (wt.%)

P1 0 11.3 32.5 56.2 16.7 83.3 56.5 1.6 0.3 202.0 2.9 1.1 4.50 −22.7 70.7P2 0.3 15.4 23.8 60.8 20.2 79.8 65.2 2.4 0.9 81.6 2.3 1.7 3.77 −22.6 62.2P3 0.6 13.0 30.0 57.0 18.6 81.4 59.9 2.6 1.0 67.6 2.0 0.4 2.69 −22.4 58.4P4 0.9 11.5 27.8 60.6 16.0 84.0 62.8 2.8 1.2 58.6 1.9 0.2 2.68 −22.7 67.1P5 1.4 16.6 35.0 48.4 25.6 74.4 48.8 2.4 1.0 56.8 1.7 0.2 1.96 −22.7 63.8P6 1.9 16.6 20.9 62.5 21.0 79.0 70.7 3.1 1.5 54.0 1.9 0.8 1.95 −23.2 65.8P7 2.4 16.4 18.0 65.6 20.0 80.0 72.2 3.2 1.6 52.7 1.9 0.8 1.82 −22.6 64.8P8 2.9 19.3 17.7 63.0 23.5 76.5 70.9 3.3 1.7 47.7 1.8 0.7 1.80 −22.9 86.8P9 3.4 19.4 15.7 64.8 23.1 76.9 73.1 3.4 4.7 18.0 1.8 0.6 1.79 −23.0 67.2P10 3.9 21.2 34.6 44.2 32.4 67.6 54.7 2.7 1.4 45.1 1.7 1.7 1.58 −22.9 59.7P11 4.4 15.5 26.1 58.4 21.0 79.0 63.8 2.9 1.5 49.3 1.8 0.3 1.59 −22.6 81.9P12 4.9 17.9 23.4 58.7 23.4 76.6 64.8 3.2 1.8 42.5 1.7 0.5 1.59 −22.7 65.6P13 5.4 14.7 36.3 49.0 23.0 77.0 52.6 2.6 1.3 47.1 1.7 0.6 0.75 −22.7 43.5P14 5.9 13.6 45.9 40.4 25.2 74.8 43.2 2.4 1.1 46.2 1.5 0.5 0.74 −22.6 50.7

Metres represent distance from dyke. Ro = vitrinite reflectance (% oil). † = dry; ⁎ = dry and ash free.

208 D.R. Gröcke et al. / Earth and Planetary Science Letters 277 (2009) 204–212

are seen with increased maturation of lower rank coals in the datareported by Boudou et al. (1984) and Mastalerz and Schimmelmann(2002).

Numerous studies have shownan increase in vitrinite reflectance (Ro)adjacent to intrusions (e.g., Bostick and Pawlewicz, 1984; Stewart et al.,2005; Cooperet al., 2007) and the thicknessof thealterationhalo appearsto be a function of the size of the intrusion (Bostick and Pawlewicz,1984),thermal conductivity, permeability and heating duration (Saxby andStephenson, 1987), and the style of heat transfer (convection vs.conduction) (Barker et al., 1998). If significant gas generation occurredduring intrusion, one would expect to see a decrease in volatile matter(VM) contentof the coal adjacent to intrusionsaccompanying an increasein Ro. If this gas generation involved preferential loss of 12C-enrichedmethane, then thealterationhalo of low-VM,high-Ro coal should containcoal with heavier δ13C values. VM valuesmay be influenced by carbonatemineralization, since the carbonates break down at temperaturesexperienced during the proximate analysis. Carbonates were observedin the most altered coals adjacent to the dyke, but appear to representb2% of the coal, so any impact on VM would be negligible.

Fig. 4. Geochemical profiles of transects “M” and “P” from the No. 4L coal seam. Note that as vishows very little change in the entire transect. See test for discussion.

Assuming the temperature of the dolerite at the time of intrusionwas ~1,200 °C and that of the coal was ~50 °C (Snyman and Barclay,1989) then as a first approximation, the temperature of the coal at theintrusion contact (Tcontact) can be estimated to have been as high as~625 °C based on the equation:

Tcontact = Tmagma + Thost� �

=2

where, Tmagma and Thost are the initial temperature of the magma andhost rock (in this case, coal), respectively (Carslaw and Jaeger, 1959).Mean random vitrinite reflectance (Rv−r) can be converted tomaximum temperature attained (Tpeak) using the relationship estab-lished by Barker and Pawlewicz (1994), providing an independentmeasure of the coal/dyke contact temperature:

Tpeak = ln Rv−rð Þ + 1:19ð Þ=0:00782

This palaeo-geothermometer is thought to be reliable up totemperatures of 300 °C (Barker et al., 1998). The conversion shows

trinite reflectance increases towards the dyke, volatilematter decreases, whereas δ13Corg

Fig. 5. Representative photomicrographs of unaltered and altered coal (all photomicrographs taken under polarized, reflected white-light using a 40x antiflex oil immersionobjective; image (d) taken under crossed polarizers). Unaltered coal samples, for example M9 (a) and M10 (b), have high inertinite contents, containing considerably moresemifusinite (Sf) than fusinite (Fus) and have relatively low vitrinite contents (Vit). Heat-altered coals (such as M1) adjacent to the intrusion (c) show development of devolatilizationvacuoles (DV) within the vitrinite, which, in turn, has altered to an isotropic coke. Note that under crossed polarizers (d) (same field of view) no anisotropy is visible in the alteredvitrinite. See text for discussion.

209D.R. Gröcke et al. / Earth and Planetary Science Letters 277 (2009) 204–212

background temperatures of ~100 °C increasing to N300 °C close to thecontact (within ~0.5 m) (Fig. 4). Vitrinite reflectance values (andinferred temperatures) return to background levels within a distanceof 1.5 to 2 times the dyke width suggesting cooling by convection thatmay have reduced temperatures from the calculated value (625 °C) atthe contact (Barker et al., 1998). Thus, most of the contact aureolereached temperatures between 100 °C and 250 °C, temperatures at

Fig. 6. Petrographic composition of transect M coal samples adjacent to dyke comparedwith δ13Corg. Note the predominance of inertinite and possible influence of petrographiccomposition on isotopic composition.

which one would expect to see methane production with significantfractionation (Cramer et al., 1998) leaving a residue carbon lessdepleted in 12C (Barker et al., 1998). However, despite the pronouncedchange in both VM and Ro seen in the two transects, there are nosignificant changes in δ13C (Fig. 4), certainly not of the magnitude thatmight be expected associated with large-scale CH4 generation.

Inertinite macerals have higher initial carbon contents and lowervolatile matter contents than vitrinite (Hessley et al., 1986) and thus,inertinite-rich coals would undergo less alteration than vitrinite-richcoals. As noted for other Gondwanan South African coals (Snyman andBotha,1993; Taylor et al., 1998), these coal samples are relatively high ininertinite macerals (~60–90%), particularly semi-inertinite (Fig. 5). Thisis significant as both VM content and bulk δ13C can be influenced bymaceral content.Within a single coal, differentmacerals can show up toa 2.5‰ difference in δ13C; in one example from aMiddle Pennsylvaniancoal, liptinitewasdepletedby ~2‰ relative to thevitrinite, and inertinitewas ~0.5‰ heavier than vitrinite (Rimmer et al., 2006). In addition, forcoals with reflectances that range between 0.6% and 1.3% Rm), Whiticar(1996) reported a slight shift (1‰) with increased reflectance towardsheavier δ13C values in both liptinites and vitrinites, but not in inertinites;the differences in maceral response likely being a function of variationsin volatile constituents. Therefore, thermal alteration of coals initiallyenriched in inertinite would tend to show more modest changes ineither VM or δ13C compared to coals more enriched in vitrinite andliptinite. In terms of maceral composition, these coals have very lowliptinite contents (b4%) and the few liptinite macerals that are presentare lost inside the alteration halo (disappearing by a Ro level of 1.35%).

There is significant methane content in Karoo coals (Auret andBuck, 1992) including the coals of the Highveld Coalfield (up to 1.3 m3/ton) (Lloyd and Cook, 2004), especially in the vicinity of the igneous

210 D.R. Gröcke et al. / Earth and Planetary Science Letters 277 (2009) 204–212

intrusions (Saghafi et al., 2008), and there have been reports ofmethane outbursts (Davies et al., 2000) and methane emissions fromwells intersecting the seams of the Highveld Coalfield (Snyman andBotha, 1993). Any discussion of the impact of 12C-enriched methanerelease must also consider the timing of methane production and howmuch of this gasmayhave been retained in the seam either as coal-bedmethane or in a condensed form such as pyrolytic carbon deposits.Pyrolitic carbon can influence bulk δ13C and has been suggested as thecause of lighter isotopic compositions directly adjacent to intrusions(Meyers and Simoneit, 1999; Cooper et al., 2007): this has also beenobserved in coals adjacent to dolerite dykes in South Africa (Falcon andSnyman, 1986).

Contrary to the opinion that coal rank variations in the Permianseams of South Africa are due primarily to the dolerite intrusions,Snyman and Barclay (1989) suggest that the intrusions resulted in onlythe initial phase of coal maturation and that subsequent burialmaturation superimposed on the contact metamorphism produced aregional increase in coal rank. Based on the low rank of coals from theFree State that were unaffected by dolerite intrusions and applying theapproach of Bostick et al. (1979), they estimate that the coal was onlylignitic in rank (with an estimated Ro of ~0.45%) at the time of intrusion.The present-day backgroundRo ~0.7% and the presence of isotropic coke(Fig. 5) are consistent with a very low rank at the time of intrusion. Thelow pre-intrusion rank is significant in that much of the magmatic heatwas likely consumed in driving off moisture from the coal (Snyman andBarclay,1989) and also influenced the type of gases emitted. Up to a rankequivalent to the medium-volatile/low-volatile boundary (Ro=1.5%), asmuch as twice the volume of CO2 as CH4 is generated, beyond whichpoint CH4 is the predominant gas evolved (Hunt, 1996). The largemolecular size of CO2 limits itsmobility andmostmigrates in a dissolvedstate (Rightmire, 1984), possibly contributing to carbonate formation inthe intruded areas (Taylor et al., 1998).

Any methane generation would have been limited to a very narrowzone (a fewmetres at most) immediately adjacent to the intrusions andat least some of this CH4 would have been retained in the coal structureadsorbed within the coal's matrix pore system. Assuming that thecoalificationmodel of Snyman and Barclay (1989) is correct, some of themethane seen today in these coals could have been generated post-intrusion associated with regional maturation, thus post-dating theKaroo-Ferrar LIP. Finally, the high inertinite content of these coalswouldindicate only limited CH4 production; because of their low initial Hcontent, inertinite macerals have relatively little hydrocarbon generat-ing potential (Scott, 2002). Therefore, the low vitrinite and liptinitecontent of these coals (components that would be more prone to CH4

generation; Scott, 2002), the modest decreases in VM adjacent to thedykes due to high initial inertinite content, the relatively narrowmetamorphic aureole surrounding the intrusions, and the likelihoodthat at least some of the volatiles generated by the intrusion weretrapped in the coal as coal-bed methane or condensed as pyrolyticcarbon in coal adjacent to the intrusions suggest that nomethane or onlya limited amount of methanewas released from coals during the Karoo-Ferrar LIP.

A possible alternative explanation is that the coal recording theintrusion of the dykes has been vaporized and thus the record is notpreserved. This is a possibility since theheatof the intruding dykewouldbe ~1,200 °C, compared to the maximum temperature recorded by thepalaeo-geothermometer of N300 °C. When the hot dyke intruded, itprobably heated up the coal in the vicinity, but in order for it to vaporizethe coal first needs to be dried. Based on Snyman and Barclay (1989) weassume that the coals were rich in moisture (up to 30%). A roughestimate of the energy available for coal vaporization can be obtainedfrom a simple energy budget calculation. The dykes cut the sub-horizontal coal layers almost vertically, so a first-order one-dimensionalapproach is considered which is symmetrical around the centre of thedyke.We consider the followingprocess: at formation, themagma in thedyke is hot (~1,200 °C above the background temperature of 60 °C) and

in a liquid state. The energy released from this hot, liquid dykeheats partof the coal layer that directly surrounds the dyke. At the timeof intrusionthe coals were at a burial depth of between 1,000–2,000 m (Johnsonet al., 2006), and thus at a pressure of 20–40 MPa, so that the boilingtemperature of water is 350–410 °C. This suggests that palaeo-geothermometer values were close to the boiling temperature ofwater, which makes drying the coal an unlikely process.

In that case, a significant part of the intrusion energy would beconsumed by the evaporation of pore water (which has a very highvaporization heat) in the immediate vicinity of the dyke. For example,heating 1 m of coal on either side of a 1 metre-wide intrusion to theboiling temperature of water, and subsequently boiling off the water init, would consume enough energy to solidify the intrusion, and reduceits excess temperature by a factor of ~2 (Turcotte and Schubert, 2002).Furthermore, a porousmediumwith30%pore space has generally averyhigh permeability, which makes the pore fluid very mobile. Therefore,hydrothermal circulation of the remaining fluids within the coal andhost rocks facilitate the transport of heat laterally through the coal andeasily distributes the remaining excess heat of the intrusion over a largearea. These processes reduce the temperature of the intrusionsignificantly, and would result in very minor amounts of energy leftfor coalvaporization. It seemsveryunlikely that an intrusion into amoistcoal at depth will be able to vaporize significant parts of the coal. Thissupports the earlier conclusions by Snyman and Barclay (1989), whosuggest that themagmatic heatwas consumed bymoisture evaporationin the lignite and associated sedimentary rocks.

All of the above evidences, including the absence of a significant shifttowards heavier δ13C in the contact aureoles, call into questionwhetherthe major negative δ13C shift recorded from T-OAE marine sedimentsand terrestrial components (seeHesselbo et al., 2000; Kemp et al., 2005;Hesselbo et al., 2007; Suan et al., 2008a) can be attributed to themassiverelease of isotopically light thermogenic methane associated with thedolerite intrusions of the Karoo-Ferrar LIP, as purported by McElwainet al. (2005) and Svensen et al. (2007).

6. Alternative explanations for the Toarcian OAE negativeδ13C excursion

Based on the results of this study we conclude that the negativeδ13C excursion associated with the T-OAE was not the result ofthermogenic methane generation from dyke intrusions into coalcaused by the Karoo-Ferrar LIP. Other organic geochemical data (e.g.,Summons et al., 2008), sedimentological and stratigraphic evidence(Anhaeusser and Maske, 1986) and thermal energy models (e.g.,Snyman and Barclay, 1989) would suggest that the Karoo-Ferrar LIPhad limited impact on the generation of thermogenic methane togenerate the rapid T-OAE carbon-isotope excursion. Although thisstudy does not dismiss one mechanism for another, several mechan-isms and scenarios can be used to generate the T-OAE negative δ13Cexcursion:

1. Degassing CO2 from LIPs (e.g., Kerr, 1998; Larson and Erba, 1999).2. Release of terrestrial and/or continental margin methane clathrates

(e.g., Dickens et al., 1995; Hesselbo et al., 2000; Nisbet, 2002).3. Biomass burning of terrestrial vegetation (e.g., Kurtz et al., 2003;

Finkelstein et al., 2006).4. Increased weathering of organic-rich sediments and soils (e.g.,

Higgins and Schrag, 2006).

At present all of the above are likely candidates, especially acombination of several or all of them. The above have been purposelyordered to generate a scenario in which LIP intrusions degas CO2 thatwarms the ocean–atmosphere system leading to the release oftrapped methane (continental or marine), and accentuating thegreenhouse effect. This has the potential to increase storm activity(Hesselbo et al., 2007), and the prevalence of terrestrial wildfires,which ultimately expose organic-rich sediments and soils to be

211D.R. Gröcke et al. / Earth and Planetary Science Letters 277 (2009) 204–212

eroded and oxidized through increased weathering (Waltham andGröcke, 2006). Although knowledge about the cause of the T-OAEnegative δ13C excursion is critical, a greater understanding of thebackground state of the Earth system prior to and after the T-OAEshould be pursued.

Acknowledgements

SASOL is acknowledged for allowing sampling at their MiddelbultColliery and for granting permission to publish this article. Karin vander Merwe (SASOL Divisional Manager Operational Geology) andPartick Ndlovu (mine geologist) assisted BC and HT during visits andsampling underground. This project was supported, in part, by grantsfrom Natural Sciences and Engineering Research Council of CanadaDiscovery Grant to DRG (No. 288321) and the University of Kentuckyto LEY (Graduate School and Brown-McFarlan Fund). We thankMargaret Grider, Gerald Thomas, and Henry Francis of the Universityof Kentucky for their assistancewith coal analyses. Two useful reviewswere provided by anonymous academics.

References

Anhaeusser, C.R., Maske, S., 1986. Mineral deposits of Southern Africa, II. GeologicalSociety of South Africa. 2376 pp.

Auret, S., Buck, P., 1992. Coalbed methane – an important Southern African energyresource. Geobulletin 35, 12–18.

Barker, C.E., Pawlewicz, M.J., 1994. Calculation of vitrinite reflectance from thermalhistories and peak temperatures: a comparison ofmethods. In: Mukhopadhyay, P.K.,Dow, W.G. (Eds.), Vitrinite Reflectance as a Maturity Parameter: Applications andLimitations, ACS Symposium Series No. 570, pp. 216–229.

Barker, C.E., Bone, Y., Lewan, M., 1998. Fluid inclusion and vitrinite-reflectancegeothermometry compared to heat-flow models of maximum paleotemperaturenext to dikes, western onshore Gippsland Basin, Australia. Int. J. Coal Geol. 37, 73–111.

Bell, F.G., Germy, C.A., 2000. The geotechnical character of some South African dolerites,especially their strength and durability. Q. J. Eng. Geol. Hydrogeol. 33, 59–76.

Bostick, N.H., Pawlewicz, M.J., 1984. Paleotemperatures based on vitrinite reflectance ofshales and limestones in igneous dike aureoles in the Upper Cretaceous PierreShale, Walsenburg, Colorado. In: Woodward, J.G., Meissner, F.F., Clayton, J.L. (Eds.),Hydrocarbon Source Rocks of the Greater Rocky Mountain Region. Rocky MountainAssociation of Geologists, pp. 387–392.

Bostick, N.H., Cashman, S.M., McCullogh, T.H., Waddel, C.T., 1979. Gradients of vitrinitereflectance and present temperature in the Los Angeles and Ventura Basins,California. In: Oltz, D.F. (Ed.), Low Temperature Metamorphism of Kerogen and ClayMinerals. Pacific Section. Soc. Econ. Paleontol. Mineral., Los Angeles, pp. 65–96.

Boudou, J.-P., Pelet, R., Letolle, R., 1984. A model of the diagenetic evolution of coalsedimentary organic matter. Geochim. Cosmochim. Acta 48, 1357–1392.

Cadle, A.B., Cairncross, B., Christie, A.D.M., Roberts, D.L.,1993. The Karoo Basin of South Africa:type-basin for coal-bearing deposits of Southern Africa. Int. J. Coal Geol. 23, 117–157.

Cairncross, B., 1989. Paleodepositional environments and tectono-sedimentary controls ofthe postglacial Permian coals, Karoo Basin, South Africa. Int. J. Coal Geol. 12, 365–380.

Cairncross, B., 2001. An overview of the Permian (Karoo) coals of Southern Africa. J. Afr.Earth Sci. 33, 529–562.

Carslaw, H.S., Jaeger, J.C., 1959. Conduction of heat in solids. Oxford University Press,New York, NY. 510 pp.

Catuneanu, O., Wopfner, H., Eriksson, P.G., Cairncross, B., Rubidge, B., Smith, R., Hancox,J., 2005. The Karoo Basins of South-Central Africa. J. Afr. Earth Sci. 43, 211–253.

Clayton, J.L., Bostick, N.H., 1986. Temperature effects on kerogen and on molecular andisotopic composition of organic matter in Pierre Shale near an igneous dike. Org.Geochem. 10, 135–143.

Cohen, A.S., Coe, A.L., Harding, S.L., Schwark, L., 2004. Osmium isotope evidence for theregulation of atmospheric CO2 by continental weathering. Geology 32, 157–160.

Cooper, J.R., Crelling, J.C., Rimmer, S.M., Whittington, A.G., 2007. Coal metamorphism byigneous intrusion in the Raton Basin, CO and NM: implications for generation ofvolatiles. Int. J. Coal Geol. 71, 15–27.

Cramer, B., Krooss, B.M., Littke, R., 1998. Modelling isotope fractionation during primarycracking of natural gas: a reaction kinetic approach. Chem. Geol. 149, 235–250.

Davies, A.W., Isaac, A.K., Cook, P.M., 2000. Investigation of a coal-mine explosion andrelevanceof risk assessment. Trans. Inst.Min.Metall. Sect. A,Mining Industry109,61–69.

De Oliveira, D.P.S., Cawthorn, R.G., 1999. Dolerite intrusionmorphology at Majuba colliery,northeast Karoo Basin, Republic of South Africa. Int. J. Coal Geol. 41, 333–349.

Dickens, G.R., O'Neil, J.R., Rea, D.K., Owen, R.M., 1995. Dissociation of oceanic methanehydrate as a cause of the carbon-isotope excursion at the end of the Paleocene.Paleoceanography 10, 965–971.

Duncan, A.R., Marsh, J.S., 2006. The Karoo igneous province. In: Johnson, M.R.,Anhaeusser, C.R., Thomas, R.J. (Eds.), The Geology of South Africa. Geological Societyof South Africa. Johannesburg/Council for Geoscience, Pretoria, pp. 501–520.

Falcon, R.M.S., Snyman, C.P., 1986. An introduction to coal petrography: atlas ofpetrographic constituents in the bituminous coals of Southern Africa. Geol. Soc. S.Africa, Review Paper No. 2. 27 pp.

Finkelstein, D.B., Pratt, L.M., Brassell, S.C., 2006. Can biomass burning produce a globallysignificant carbon-isotope excursion in the sedimentary record? Earth Planet. Sci.Lett. 250, 501–510.

Hagelskamp, H.H.B., 1987. The Influence of Depositional Environment and DoleriteIntrusions on the Quality of Coal: A Study of the Twistdraai Mine Area near Secunda,Southeastern Transvaal. Ph.D. Dissertation (Unpublished). University of Pretoria,Pretoria, South Africa, 272 pp.

Hagelskamp, H.H.B., Eriksson, P.G., Snyman, C.P., 1988. The effect of depositionalenvironment on coal distribution and quality parameters in a portion of theHighveld Coalfield, South Africa. Int. J. Coal Geol. 10, 51–77.

Hermoso, M., Le Callonnec, L., Minoletti, F., Renard, M., Hesselbo, S.P., 2009. Expressionof the Early Toarcian negative carbon-isotope excursion in separated carbonatemicrofractions (Jurassic, Paris Basin). Earth Planet. Sci. Lett. 277, 194–203.doi:10.1016/j.epsl.2008.10.013.

Hesselbo, S.P., Gröcke, D.R., Jenkyns, H.J., Bjerrum, C.J., Farrimond, P., Morgans Bell, H.S.,Green, O.G., 2000. Massive dissociation of gas hydrate during a Jurassic oceanicanoxic event. Nature 406, 392–395.

Hesselbo, S.P., Jenkyns, H.C., Duarte, L.V., Oliveira, L.C.V., 2007. Carbon isotope record ofthe Early Jurassic (Toarcian) Oceanic Anoxic Event from fossil wood and marinecarbonate (Lusitanian Basin, Portugal). Earth Planet. Sci. Lett. 253, 455–470.

Hessley, R.K., Reasoner, J.W., Riley, J.T., 1986. Coal Science: An Introduction to Chemistry,Technology, and Utilization. John Wiley and Sons, New York, NY. 269 pp.

Higgins, J.A., Schrag, D.P., 2006. Beyond methane: towards a theory for the Paleocene–Eocene Thermal Maximum. Earth Planet. Sci. Lett. 245, 523–537.

Hobday, D.K., 1978. Palaeoenvironmental models in the eastern Karoo Basin. Palaeontol.Afr. 21, 1–13.

Hoefs, J., Frey, M., 1976. The isotopic composition of carbonaceous matter in ametamorphic profile from the Swiss Alps. Geochim. Cosmochim. Acta 40, 945–951.

Hunt, J.M., 1996. Petroleum Geochemistry and Geology, Second Edition. W.H. Freemanand Company, New York, NY. 743 pp.

Johnson, M.R., van Vuuren, C.J., Visser, J.N.J., Cole, D.J., de Wickens, V.H., Christie, A.D.M.,Roberts,D.L., Brandl,G., 2006. Sedimentary rocks of theKarooSupergroup. In: Johnson,M.R., Anhaeusser, C.R., Thomas, R.J. (Eds.), The Geology of South Africa. GeologicalSociety of South Africa. Johannesburg/Council for Geoscience, Pretoria, pp. 461–499.

Kemp, D.B., Coe, A.L., Cohen, A.S., Schwark, L., 2005. Astronomical pacing of methanerelease in the Early Jurassic period. Nature 437, 396–399.

Kerr, A.C., 1998. Oceanic plateau formation: a cause of mass extinction and black shaledeposition around the Cenomanian–Turonian boundary? J. Geol. Soc. (Lond.) 155,619–626.

Kurtz, A.C., Kump, L.R., Arthur, M.A., Zachos, J.C., Paytan, A., 2003. Early Cenozoicdecoupling of the global carbon and sulfur cycles. Paleoceanography 18, 1090.doi:10.1029/2003PA000908.

Larson, R.L., Erba, E., 1999. Onset of the mid-Cretaceous greenhouse in the Barremian-Aptian: igneous events and the biological, sedimentary and geochemical responses.Paleoceanography 14, 663–678.

Lloyd, P., Cook, A., 2004. Methane release from South African coal mines. J. South Afr.Inst. Min. Metall. 105, 483–490.

Mastalerz, M., Schimmelmann, A., 2002. Isotopically exchangeable organic hydrogen incoal relates to thermal maturity and maceral composition. Org. Geochem. 33,921–931.

McElwain, J.C., Wade-Murphy, J., Hesselbo, S.P., 2005. Changes in carbon dioxide during anoceanic anoxic event linked to intrusion into Gondwana coals. Nature 435, 479–482.

McKirdy, D.M., Powell, T.G., 1974. Metamorphic alteration of carbon isotopic composi-tion in ancient sedimentary organic matter. New evidence from Australia and SouthAfrica. Geology 2, 591–595.

Meyers, P.A., Simoneit, B.R.T., 1999. Effects of extreme heating on the elemental andisotopic compositions of an Upper Cretaceous Coal. Org. Geochem. 30, 299–305.

Nisbet, E.G., 2002. Have sudden large releases of methane from geological reservoirsoccurred since the Last Glacial Maximum, and could such releases occur again?Philos. Trans. – Royal Soc., Math. Phys. Eng. Sci. 360, 581–607.

Rightmire, C.T., 1984. Coalbed methane resources. In: Rightmire, C.T., Eddy, G.E., Kirr, J.M.(Eds.), Coalbed Methane Resources of the United States. AAPG Studies in GeologySeries, vol. 17, pp. 1–13.

Rimmer, S.M., Rowe, H.D., Taulbee, D.N., Hower, J.C., 2006. Influence of maceral contenton δ13C and δ15N in a Middle Pennsylvanian coal. Chem. Geol. 225, 77–90.

Sackett, W.M., 1978. Carbon and hydrogen isotope effects during the thermocatalyticproduction of hydrocarbons in laboratory simulation experiments. Geochim.Cosmochim. Acta 42, 571–580.

Sackett, W.M., Nakaparksin, S., Dalrymple, D., 1970. Carbon isotope effects in methaneproduction by thermal cracking. In: Hobson, G.D., Speers, G.C. (Eds.), Advances inOrganic Geochemistry. Pergamon Press, Oxford, pp. 37–53.

Saghafi, A., Pinetown, P.G., Grobler, P.G., van Heerden, J.H.P., 2008. CO2 storage potentialof South African coals and gas entrapment enhancement due to igneous intrusions.Int. J. Coal Geol. 73, 74–87.

Saxby, J.D., Stephenson, L.C., 1987. Effect of an igneous intrusion on oil shale at Rundle(Australia). Chem. Geol. 63, 1–16.

Scott, A.R., 2002. Hydrogeologic factors affecting gas content distribution in coal beds.Int. J. Coal Geol. 50, 363–387.

Simoneit, B.R.T., Brenner, S., Peters, K.E., Kaplan, I.R., 1981. Thermal alteration ofCretaceous black shales by diabase intrusions in the Eastern Atlantic – II. Effects onbitumen and kerogen. Geochim. Cosmochim. Acta 45, 1581–1602.

Snyman, C.P.,1998. Coal. In: Wilson, M.G.C., Anhaeusser, C.R. (Eds.), The Mineral Resourcesof South Africa. Handbook, vol. 16. Council for Geoscience, Pretoria, pp. 136–205.

Snyman, C.P., Barclay, J., 1989. The coalification of South African coal. Int. J. Coal Geol. 13,375–390.

Snyman, C.P., Botha, W.J., 1993. Coal in South Africa. J. Afr. Earth Sci. 16, 171–180.

212 D.R. Gröcke et al. / Earth and Planetary Science Letters 277 (2009) 204–212

Stewart, A.K., Massey, M., Padgett, P.L., Rimmer, S.M., Hower, J.C., 2005. Influence of abasin intrusion on vitrinite reflectance and chemistry of the Springfield (no. 5) coal,Harrisburg, Illinois. Int. J. Coal Geol. 63, 58–67.

Suan, G., Mattioli, E., Pittet, B., Mailloiot, S., Lécuyer, C., 2008a. Evidence for majorenvironmental perturbation prior to and during the Toarcian (Early Jurassic)oceanic anoxic event from the Lusitanian Basin, Portugal. Paleoceanography 23,PA1202. doi:10.1029/2007PA001459.

Suan, G., Pittet, B., Bour, I., Mattioli, E., Duarte, S.V., Mailloiot, S., 2008b. Duration of theEarly Toarcian carbon isotope excursion deduced from spectral analysis: conse-quence for its possible causes. Earth Planet. Sci. Lett. 267, 666–679.

Summons, R.E., Hope, J.M., Swart, R., Walter, M.R., 2008. Origin of Nama Basin bitumenseeps: petroleum derived from a Permian lacustrine source rock traversingsouthwestern Gondwana. Org. Geochem. 39, 589–607.

Svensen, H., Planke, S., Chevallier, L., Malthe-Sørenssen, A., Corfu, F., Jamtveit, B., 2007.Hydrothermal venting of greenhouse gases triggering Early Jurassic globalwarming. Earth Planet. Sci. Lett. 256, 554–566.

Taylor, G.H., Teichmüller, M., Davis, A., Diessel, C.F.K., Littke, R., Robert, P., 1998. OrganicPetrology. Gebrüder Bornträger, Berlin. 704 pp.

Turcotte, D.L., Schubert, G., 2002. Geodynamics, 2nd ed. Cambridge University Press.456 pp.

van Zijl, J.S.V., 2006. Physical characteristics of the Karoo sediments and mode ofemplacement of the dolerites. S. Afr. J. Geol. 109, 329–334.

van de Schootbrugge, B., McArthur, J.M., Bailey, T.R., Rosenthal, Y., Wright, J.D., Miller, K.G.,2005. Toarcianoceanic anoxic event: an assessmentof global causesusing belemnite Cisotope records. Paleoceanography 20, PA3008. doi:10.1029/2004PA001102.

Waltham, D., Gröcke, D.R., 2006. Non-uniqueness and interpretation of the seawater87Sr/86Sr curve. Geochim. Cosmochim. Acta 70, 384–394.

Whiticar, M.J., 1996. Stable isotope geochemistry of coals, humic kerogens and relatednatural gases. Int. J. Coal Geol. 32, 191–215.