Earth and Planetary Science Letters 277 (2009) 490–500

Contents lists available at ScienceDirect

Earth and Planetary Science Letters

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

Siberian gas venting and the end-Permian environmental crisis

Henrik Svensen a,⁎, Sverre Planke a,b, Alexander G. Polozov a,c, Norbert Schmidbauer d, Fernando Corfu e,Yuri Y. Podladchikov a, Bjørn Jamtveit a

a Physics of Geological Processes (PGP), University of Oslo, PO Box 1048 Blindern, 0316 Oslo, Norwayb Volcanic Basin Petroleum Research (VBPR), Oslo Research Park, 0349 Oslo, Norwayc Institute of Geology of Ore Deposits, Petrography, Mineralogy and Geochemistry, Russian Academy of Sciences, 119017 Moscow, Russiad Norwegian Institute for Air Research, Kjeller, Norwaye Dept. of Geosciences, University of Oslo, PO Box 1047 Blindern, 0316 Oslo, Norway

gas venting

⁎ Corresponding author.E-mail address: hensven@fys.uio.no (H. Svensen).

0012-821X/$ – see front matter © 2008 Elsevier B.V. Adoi:10.1016/j.epsl.2008.11.015

a b s t r a c t

a r t i c l e i n f o

Article history:

The end of the Permianperio Received 8 July 2008Received in revised form 7 November 2008Accepted 14 November 2008

Editor: Dr. R.W. Carlson

Keywords:end-PermianSiberian TrapsTunguska Basin

d ismarked byglobalwarming and the biggest knownmass extinction on Earth. Thecrisis is commonly attributed to the formation of the Siberian Traps Large Igneous Province although the causalmechanisms remain disputed. We show that heating of Tunguska Basin sediments by the ascending magmaplayed a key role in triggering the crisis. Our conclusions are based on extensive field work in Siberia in 2004 and2006.Heating of organic-rich shale andpetroleumbearing evaporites around sill intrusions led to greenhouse gasandhalocarbongeneration in sufficient volumes to causeglobalwarmingand atmospheric ozonedepletion. Basinscale gas production potential estimates show that metamorphism of organic matter and petroleum could havegenerated N100,000 Gt CO2. The gases were released to the end-Permian atmosphere partly through spectacularpipe structures with kilometre-sized craters. Dating of a sill intrusion by the U–Pb method shows that the gasrelease occurred at 252.0±0.4 million years ago, overlapping in time with the end-Permian global warming andmass extinction. Heating experiments to 275 °C onpetroleum-bearing rock salt fromSiberia suggests thatmethylchloride and methyl bromide were significant components of the erupted gases. The results indicate that globalwarming and ozone depletion were the two main drivers for the end-Permian environmental crisis. Wedemonstrate that the composition of the heated sedimentary rocks below the flood basalts is themost importantfactor in controlling whether a Large Igneous Provinces causes an environmental crisis or not. We propose that asimilar mechanism could have been responsible for the Triassic-Jurassic (~200 Ma) global warming and massextinction, based on the presence of thick sill intrusions in the evaporite deposits of the Amazon Basin in Brazil.

© 2008 Elsevier B.V. All rights reserved.

1. Introduction

The end-Permian crisis is the biggest known marine and terrestrialextinction event (Wignall, 2001; Erwin et al., 2002). Themain stage of theextinction occurred during a relatively short time span of about200,000 years contemporaneous with the Siberian Traps volcanism(Bowring et al., 1998; Mundil et al., 2004; Kamo et al., 2006). A globalcarbon cycle perturbation is recorded in marine and terrestrial strata (i.e.,Retallack andKrull, 2006)with globalwarming extending 4–5m.y. into tothe Triassic (Payne et al., 2004; Payne and Kump, 2007). It has recentlybeen shown that the main carbon isotope excursion occur before theextinction, showing that the changes in the global carbon cyclewere not aside effect or consequenceof the extinction (e.g., Retallack andKrull, 2006;Xie et al., 2007). This link provides an important constraint of the type ofgeological processes potentially causing the extinction. Amongst thehypotheses for the end-Permian event are H2S degassing from a stratifiedocean followed by ozone layer collapse (Grice et al., 2005; Kump et al.,2005; Meyer and Kump, 2008), degassing of CO2 from the Siberian Traps

ll rights reserved.

Large Igneous Province (LIP) (e.g., Wignall, 2001; Reichow et al., 2002;Grard et al., 2005; Xie et al., 2007) possibly also releasing methane fromgas hydrates (Wignall, 2001; Berner, 2002), and contactmetamorphismofcoal and other carbonaceous sediments in the Tunguska Basin in EasternSiberia generating carbon gases and possibly halocarbons (Svensen et al.,2004; Visscher et al., 2004; Retallack andKrull, 2006; Beerling et al., 2007;Payne and Kump, 2007; Retallack and Jahren, 2008). Siberian Traps sillintrusionswere emplaced into the vast evaporite deposits of the TunguskaBasin, but so far the effects of contact metamorphism of salt and sulphatesequences on the end-Permian environment have not been fullyaddressed. The aimof this study is todevelop anewgeologicalmechanismto explain the end-Permian globalwarming andmass extinction, by usinga combination of field work in the Tunguska Basin (followed by corestudies, petrography, U–Pb dating), and laboratory gas generation ex-periments on salt samples collected during a 2004 field campaign.

2. The Tunguska Basin

The vast Tunguska Basin in Eastern Siberia contains the oldest pe-troleum system in theworld, formed duringmaturation of 1–8 km thickdeposits of Neo-Proterozioc shale and carbonate (e.g., Meyerhoff, 1980;

491H. Svensen et al. / Earth and Planetary Science Letters 277 (2009) 490–500

Surkov et al., 1991; Sokolov et al., 1992). The Tunguska basin (Fig. 1) ispetroleum-bearing, with numerous reservoirs of oil and gas (Fig. 2).Carbonates and minor sandstone and shale horizons dominate theCryogenian and Tonian (formerly Riphean) source rock sequenceswhichare overlain by the carbonate and evaporate facies of the Ediacaran(formerly Vendian). Furthermore, enormous volumes of Cambrianevaporites are present in the basin, with up to 2.5 km thick sequencesof halite-rich strata, anhydrite, and carbonates (Fig. 2) (Zharkov, 1984;Petrychenko et al., 2005). Five major phases of salt deposition occurredin the Cambrian, the most extensive being the 2 million km2 EarlyCambrian Usolye salt basinwith an average of 200 m of halite (Zharkov,1984). Note that the “Tunguska Basin” in the literature is frequentlyincluded in the terms “SiberianPlatform” and “SiberianCraton”, and thatthe TunguskaBasin is often consideredasoneofmanybasins situatedonthe platform/craton. We use the term to encompass all the post Neo-Proterozoic sedimentary rocks on the platform/craton.

The total thickness of the basin stratigraphy commonly variesbetween 3 km and 12.5 km (Meyerhoff, 1980; Kontorovich et al., 1997),however the Neo-Proterozoic rocks are locally present as 7–10 kmthick rift segment deposits (Sokolov et al., 1992; Kuznetsov, 1997;Drobot et al., 2004). Post-Cambrian rocks comprise carbonates, marls,

Fig. 1. Geological map of the Tunguska Basin in Eastern Siberia, Russia. Note the high abundancefilled pipes north of 68°. Our main study area during a 2004 field campaign is indicated by themodified fromMalich et al (1974), and the positions of the pipes were compiled from various sFlaass and Naumov, 1995; Ryabov et al., 2005; Ryabov, 2006). The outline of the Cambrian eva

sandstones, and coal (Fig. 2), and the sedimentation terminated in thelatest Permian with the onset of Siberian Traps volcanism.

The Tunguska Basin sediments were intruded by the sub-volcanicpart of the Siberian Traps. Sills and dykes are abundant throughout thebasin, and form sheets up to 350m thick, locally comprising up to 65% ofthe stratigraphy (Meyerhoff, 1980; Fedorenko and Czamanske, 1997;Ulmishek, 2001). The maximum accumulated sill thickness in theCambrian to Permian strata is 1200 m (Kontorovich et al., 1997). Thethickness of sill intrusions in the Neo-Proterozoic rocks is uncertain dueto a limited number of deep boreholes in the bulk part of the basin(Fig. 2). However, thick sills are commonly present at the base of theCambrian evaporate sequence (Kontorovich et al., 1997; Ulmishek,2001). The present day area with outcropping sill intrusions is at least1.6 million km2 (Fig. 1). The sill emplacement led to widespread contactmetamorphism of the host sediments (e.g., Kontorovich et al., 1997) andto enhanced maturation of organic matter and the formation ofmethane-rich petroleum accumulations (Sokolov et al., 1992). Themost profound results of the magma-sediment interaction are specta-cular magnetite-rich breccia pipes rooted in the Cambrian evaporites orpossibly deeper. These pipes are numerous in the southern parts of thebasin, where they are filled with up to 700 m deep and 1.6 km wide

of phreatomagmatic pipes withmagnetite south of latitude 64, and the numerous basalt-star symbol. The aerial extent of evaporite is from Zharkov (1984). The geological map is

ources (Malich et al., 1974; Nikulin and Von-der-Flaass, 1985; Pukhnarevich, 1986; Von derporite is from Petrychenko et al. (2005), comprising a total area of 2 million km2.

Fig. 2. Schematic stratigraphy with compiled information about the occurrences ofpetroleum reservoirs in the Tunguska Basin (Meyerhoff, 1980). The stratigraphy isconsidered representative for the basin segment with pipes. The thicknesses of the pre-Cambrian strata vary considerably (1–10 km) but has not been drilled in the centralparts of the basin. Sill intrusions are present throughout the basin stratigraphy(Kontorovich et al., 1997). The pipes are rooted in the Cambrian evaporite or possiblydeeper, but have only been drilled to Upper Cambrian levels, as at Nepa (see Fig. 3). Theinferred source region of the pipes, where intense magma-sediment interactions tookplace, is likely within the lower Cambrian strata. Craters are filled with up to 750 m ofcrater-lake sediments (Von der Flaass and Naumov, 1995; Von der Flaass, 1997).

492 H. Svensen et al. / Earth and Planetary Science Letters 277 (2009) 490–500

crater lake deposits (Von der Flaass and Naumov, 1995; Von der Flaass,1997).About250mineralizedpipeswithmagnetitematrix are identifiedin the basin based on aero-magnetic surveys. Several of these arecurrently mined for magnetite. Many more pipes, especially withoutmagnetite, are likely present due to the poor rock exposures on theSiberian taiga. For instance, pipes without magnetite mineralizationhave been discovered by coincidence during mining. In addition, morethan500 basaltic diatreme-like pipes are known in the northernparts ofthe Tunguska Basin (Fig. 1).

Siberian Traps lava flows are covering parts of the Tunguska Basin,and comprise up to 6500 compositemeters in the north (e.g., Federenko

andCzamanske,1997;Kamoet al., 2006). Lavaflows are not outcroppingsouth of ~60°, and the absence of lava within the crater-lake depositssuggests that lavaflowsnever formed an extensive cover in the southernparts of the basin. However, pyroclastic rocks are abundant in interval of68 to 62° (Fig. 1) (e.g., Federenko and Czamanske, 1997; Kamo et al.,2006), but their formation and source remains poorly understood.

The Cambrian evaporite strata in the Nepa area contain sills, contactaureoles, and volcanic breccia pipes of SiberianTraps age (Fig. 3). A 2004field campaign to Nepa (N 59° and E 107°; Fig. 1) was targeted at coresdrilled at Nepa in the 1970's and 80's for potassium salt prospecting. Thesite is unique as the drilled salt is preserved, the cores are largely intact,and it was possible to identify core tray labels during fieldwork. The on-site work included mapping, borehole logging, and sampling of end-Permian pipe structures. A 200 m thick mafic sill is located in the upperCambrian strata, and a second level of sillswas discovered in the section.The upper sill has a maximum thickness of 326 m in the Nepa area(Zamaraev et al., 1985). The sills and pipes are associated with extensivehydrothermal alteration zones and contact aureoles. The northernmostpipe structure at the Nepa locality, the Scholokhovskoie pipe, ischaracterized by five main breccia lithologies with varying degrees ofmagmatic and sedimentary fragments (Fig. 4).

3. Methods

3.1. Heating experiments

Gas extraction and heating experiments were conducted at theNorwegian Institute for Air Research on natural rock salt samples fromthe 194 borehole at Nepa (Fig. 3). Eight samples were analysed, allcontaining petroleum-bearing fluid inclusions. Salt from the interior ofthe coreswere selected for analyses tominimize possible contaminationduring drilling, storage, and sampling. Contamination is accordinglyregarded as minimal, also considering that compounds formed fromheating of wood fragments or plastic bags would easily be identifiedduring the heating runs. Between 2 and 4 g of sample were gentlycrushed to 3–10 mm pieces and split in two fractions. Gas was flushedthroughPerkin Elmer stainless steel adsorption tubes prior to desorptionat 250 °C for 7.5 min, followed by GC-MS analyses on a Hewlett Packard1800Awith a detector temperature of 225 °C, and a 1 µm DB-170132 mseparation column. The analyses are semi-quantitative and measuredrelative to a toluene standard. The degree of adsorption in the tubes is10–50% for light gases, thus significantly underestimating the actualconcentrations. Furthermore, CO2 andCH4 are not adsorbed in the tubes.However, the analytical setup is efficient in terms of analyzinghalocarbons, which are the targeted gas compounds in this study.

The first set of samples was crushed in a 20 ml sealed steelchamber in an argon atmosphere at room temperature, and a knownvolume of gas was sucked through the absorption tubes by a pump.

The second set of samples was used in heating experiment toinvestigate the potential for gas synthesis. The crushed samples wereplaced directly in the adsorption tubes and heated to 275 °Cwithin thethermodesorption oven in a helium atmosphere. The released gaseswere continuously trapped at −60 °C before transfer to the GC-MS.

3.2. Gas production potential

We have applied the following relationship to calculate the gasproduction potential Wc=V⁎ρ⁎S during contact metamorphism oforganic carbon, where V is the aureole volume (sill area times aureolethickness), ρ is the rock density, and S is the aureole shape parameterdefined as the TOC profile through the aureole (Svensen et al., 2007).Weassume a linear relationship in carbon loss profiles away from the sill-sediment contacts (S=0.5), implying that half of the organic matter inthe aureole was converted to carbon gas during metamorphism.Generally, contact aureoles have the same thicknesses as their respectivesills, both below and above intrusions in low permeability sediments

Fig. 3. Composite cross section from the Nepa locality. The location map shows four of the Nepa boreholes that we have studied in detail and three breccia pipes. The Cambrianevaporite strata are overlain by Ordovician clastic sediments. Core data from the 6G hole in the Scholokhovskoie pipe and from the 194 hole in the lower sill intrusion are presentedhere. The topmost stratigraphic unit is Ordovician (O), and the rest of the drilled sequences are from the Late Cambrian (Verkholensk Suite), Middle Cambrian (Litvintsev Suite,abbreviated L-S), and the Lower Cambrian (Angara Suite and Bulay Suite, abbreviated B-S). The Upper sill is sometimes referred to as the Usol'e sill (Zamaraev et al., 1985).

493H. Svensen et al. / Earth and Planetary Science Letters 277 (2009) 490–500

(e.g., Raymond and Murchison, 1989, 1991; Galushkin, 1997; Fjeldskaaret al., 2008). This implies that the total volume of sediments affected bycontact metamorphism is equal to twice the sill volume. Note that thegas will be produced in the aureole independent of the specific type oforganic material undergoingmetamorphism (dispersed organic matter,coal beds, or petroleum). Themass conversion factors for calculating gasequivalents from carbon are 1.34 and 3.66 for methane and carbondioxide, respectively.

3.3. Dating

U–Pb analyses on two dolerite samples from the 194 borehole(sampled at 860.8 m and 868.7 m) (Fig. 2) were carried out using ID–TIMS (isotope dilution thermal ionization) and a Finnigan MAT262instrument at the Department of Geosciences in Oslo. The 20 to 30zircon grains found in each of the two samples occurred largely asfragments, locally with some preserved euhedral faces. Most grainsshowed some local turbidity, fractures or inclusions of other minerals.Baddeleyite was only observed as an inclusion in one zircon. Abrasiongenerally removed most of the turbid parts of the fragments. The bestgrains were selected for analysis, some were perfectly clear but othersstill contained some imperfections. Zircon grains were abraded beforeanalysis, then dissolved and transferred directly to the mass spectro-meter for measurement, except for one larger fraction processedthrough anion exchange resin. A mixed 235U–205Pb–202Pb spike wasused for internal normalization of the fractionation of Pb. See Corfu(2004) for analytical procedures. The datawere calculated using decayconstants from Jaffey et al. (1971). The uncertainties are 2σ.

4. Results

4.1. The breccia

The magmatic fragments of the Scholokhovskoie pipe are rich inglass (Fig. 4), demonstrating rapid melt quenching in the pipe, and thepipe formation was accordingly contemporaneous with the sill em-placement. Hydrothermal minerals include calcite, dolomite, halite,garnet, epidote, and chlorite, either as pore filling minerals or alter-ation products from igneous fragments.

4.2. Gas generation experiments

Heating experiments on evaporite samples from the 194 boreholewere done to determine the type of gas generated during contactmetamorphism in the Tunguska Basin. We use natural rock samplesequivalent to those that were heated by sills during the end-Permian. The samples were collected at depths between 807 and949 m, including two reference samples from the contact aureole ofthe lower sill (Table 1). The samples consist of coarse grained halitewith minor sylvine, anhydrite, and pyrite. Trails of liquid and gasinclusions are abundant in the salt (Grishina et al., 1998), releasingaromatic and sulphurous gases when crushed. Table 1 show thatbutane, benzene, and sulphur-bearing gases are the most abundantof the analysed petroleum compounds at room temperature.Sulphur dioxide is identified in most samples, whereas only twosamples contain dimethyl sulphide. Note that no halocarbons weredetected.

Fig. 4. Fieldwork and sampling in eastern Siberia. A) The 3000 m wide Zheleznogorsk pit is located 350 km south-west of Nepa. The pipe is mined for magnetite. Thisphreatomagmatic pipe is of similar type as the one studied at Nepa. The pipe itself is outlined on the figure (stippled) and cuts the surrounding Ordovician sediments (carbonates andsandstones). It is one of 5–6 magnetite-bearing pipes that are currently mined in the Tunguska Basin. B) Parts of the remaining core storage at Nepa, now exposed to weather. Corelogging and sampling was conducted on site in 2004. C) Altered volcanic breccia from the 6G borehole. The matrix minerals include chlorite, epidote, magnetite, and halite. D)Phreatomagmatic breccia with altered sedimentary fragments (red) in a fine grained volcanic matrix (black and grey). Halite is identified as cement in the breccia. E) Doleritefragment dominated by glass (black) and plagioclase phenocrysts (grey).

494 H. Svensen et al. / Earth and Planetary Science Letters 277 (2009) 490–500

When the samples were heated to 275 °C for 7.5 min in an inertatmosphere the concentrations of sulphur dioxide increased signifi-cantly (up to 130 times), and the concentration of other sulphur gasesand hydrocarbons decreased (up to 30 times less butane). It isimportant that halocarbons like methyl chloride and methyl bromidewere identified in all heating runs, as these compounds were notpresent at room temperature. Maximum concentrations of methylchloride was 161 ng/g rock for sample 194/4. The mass ratio of methyl

chloride tomethyl bromide is between 6.6 and 58.6 with an average of17.3 for the 8 experiments. We use the highest value (i.e., 60) tocalculate the methyl bromide generation during contact metamorph-ism, as bromide is present in significantly lesser amounts in thesedimentary rocks compared to chloride. Halogenated butane,especially 1-chloro-butane, is present in considerable amounts (upto 51 ng/g rock) in all runs except one (sample 194/3). Whenextrapolated, only 1–350 mg of methyl chloride was generated per m3

Table 1Gas extraction and generation from hole 194 halite

Sample 194/5 194/2 194/3 194/4 194/9⁎ 194/21⁎ 194/30 194/32 194/5 194/2 194/3 194/4 194/9⁎ 194/21⁎ 194/30 194/32Depth 830.2 807.2 819.4 824.4 840.9 890.1 937.9 949.3 830.2 807.2 819.4 824.4 840.9 890.1 937.9 949.3

Method Room temperature, ng gas/g rock Heating and extraction at 275 °C, ng gas/g rock

Sulfur dioxide 0.1 0.03 0.3 1.3 0 1.1 0.2 1.4 2.5 3.9 0.2 4.6 97.2 109.9 11.5 7.0Methyl tioetane 157.0 13.4Thiobis ethane 35.7 6.5Dimethyl sulfide 115.7 114.2 11.0Butane 8.1 4.4 6.1 13.3 2.0 2.0 0 3.9 2.3 3.0 0.2 4.8 1.7 10.6 3.7 1.5Benzene 20.6 36.2 7.0 3.7Methyl chloride 5.8 17.7 0.5 161 5.9 12.1 9.1 3.5Methyl bromide 0.9 2.1 0.01 15.6 0.7 1.3 1.0 0.131-chloro-butane 8.1 7.3 0 19.3 29.8 50.9 24.1 11.11-bromo-butane 0 7.4 0.03 6.7 0 12.6 5.5 0Ch3Cl/CH3Br 6.6 8.4 58.6 10.3 8.4 9.3 9.4 27.2

⁎Sample affected by contact metamorphism.

Table 2Gas production potential in the Tunguska Basin (rounded numbers)

Method Volume Organic carbon C CO2 CH4 CH3Cla CH3Brb

km3 wt.% Gt Gt Gt Gt Gt

Pipes (1) 5 2 0.23 1.13 0.31 0.11 0.0019Pipes (1) 5 5 0.58 2.10 0.77 0.28 0.0047Pipes (6400) 32,000 2 1500 5400 2000 720 12Pipes (6400) 32,000 5 3700 13,500 4900 1800 30Aureoles 800,000 0.5 9200 33,700 12,300 4500 75Aureoles 800,000 1.5 27,600 101,000 37,000 13,500 225Total prod. pot. (Low) 10,700 39,000 14,300 5200 87Total prod. pot. (High) 31,200 114,000 41,900 15,300 255Lava degassing 5400 ~20,000

a CH3Cl/TOC=0.49 (Beerling et al., 2007).b CH3Cl/CH3Br=60 (this study).

495H. Svensen et al. / Earth and Planetary Science Letters 277 (2009) 490–500

heated evaporite. This is due to the small volume of fluid inclusionspresent, and that the fluid inclusions are the only hosts of reactivehydrocarbons in the samples. Thus, we cannot use the experimentsdirectly for calculating the mass of halocarbons generated duringcontact metamorphism.

4.3. Gas production potential

The gas production potential (Svensen et al., 2004) during meta-morphism of the Tunguska Basin sediments has been estimated basedon basin scale metamorphism of organic carbon. We have used twoapproaches to estimate the carbon gas production potential duringcontact metamorphism of the Tunguska basin sediments during sillemplacement and pipe formation. 1) Pipe source region. Assumes a5 km3pipe source region, a 2–5wt.% conversion of organicmatter to gas,and a rock density of 2300 kg/m3. The pipe source region size iscorroborated by geophysical data (Von der Flaass, 1997) and alterationzones from core data. 2) Basin scale gas generation in contact aureoles(Svensen et al., 2004). Here we assume the emplacement of one sill of200 m thickness (e.g., Vasil'ev, 1999) in an area of 2 million km2 withpipes, shale, evaporite, and petroleum (Fig. 1), leading to a sill volume of400,000 km3 and hence an aureole volume of 800,000 km3. Finally, weassume that between 0.5 and 1.5 wt.% of organic matter reacted to gas,and that this comprised half of the bulk rock TOC in the aureole with adensity of 2.3 g/cm3. Note that this approach underestimates the gasproduction because there are multiple sill emplacement levels in thebasin (cf. Kontorovich et al., 1997), and the metamorphism may locallyhave affected layers with higher TOC contents higher than 3 wt.% (cf.Kuznetsov, 1997). Furthermore, CO2 generation from metamorphism ofdolostone is not included in the estimates, but represents a potentiallysignificant source of isotopically heavy carbon (δ13C~−4 to +2‰ VPDB).We assume a CH3Cl/TOC of 0.49 (Beerling et al., 2007), and a CH3Cl/CH3Br of 60 (this study).

The resulting basin scale carbon gas production potential is between9200 and 27,600 Gt C (Table 2). The results from the calculations basedon the pipes show that the production potential per pipe is 0.23–0.58 GtC.Mapping of pipes has beenmost extensive in the Bratsk regionwhereabout 200 pipes are known in a 250×250 km (62,500 km2) large area(i.e., one pipe per 312 km2). We can extend this coverage to the poorlyexposed2,000,000km2 large basinwith evaporites, resulting in a total of6400 pipes in the basin. The results in a potential for generating 1500–3700 Gt C in the pipe source regions (Table 2).

4.4. Age of sill emplacement

The analytical data for zircon in the lower dolerite sill of hole 194define a slightly elongated cluster along the Concordia curve with206Pb/238U dates ranging from 248.0 to 252.4 Ma (Fig. 5 and Table 3). Acheck for a potential bias from disequilibrium-related excess 206Pb

shows that the effect is insignificant. The analyzed zircon grains arevery rich in U (N700 ppm–21,800 ppm) and also have high Th/U ratios(1.5–3.5). The spread in ages is therefore most likely a result of partialPb loss. The uppermost analyses, however, appear to have escaped dis-turbances because the weighted average of their 206Pb/238U age of252.0±0.4Ma. This is corroborated by the upper intercept age of 252.4±1.2Maof the regression line throughall 11datapoints. The ageof 252.0±0.4 Ma is thus considered the best estimate for crystallization of the sillintrusion.

5. Discussion

5.1. Volatile generation

The source rocks of the Tunguska Basin reached peak maturityduring the Carboniferous (Meyerhoff, 1980; Ulmishek, 2001), butpetroleumgeneration started as earlyas in theNeo-Proterozoic (Sokolovet al., 1992). This means that the basin was petroleum-bearing prior tothe end-Permian sill emplacement. The TOC values in rocks affected bycontact metamorphism, which form the basis for estimating the carbongasproductionpotential, are uncertain due to fewboreholespenetratingthe source rocks in the deep central parts of the basin. However,assuming that the available data from the basin margins are represen-tative, a 1–3 wt.% TOC is supported (Sokolov et al., 1992; Kuznetsov,1997;Bartleyet al.,1998;Ulmishek, 2001). Organic carbonvalues ashighas 8–22 wt.% are present locally in source rocks (Sokolov et al., 1992;Kuznetsov, 1997; Drobot et al., 2004). In addition, sills have intrudedpetroleum-bearing Ediacaran and Cambrian levels, where 8–10 vol.%petroleum fill is present in carbonate reservoirs (e.g., Meyerhoff, 1980;Kuznetsov, 1997). The metamorphism also led to enhanced petroleummigration (e.g., Polyanskii and Reverdatto, 2002).

From the composition of the sediments undergoing contact meta-morphism, gases like CH4 and CO2 (from carbonates, organic matter,and petroleum), SO2 (from evaporites) were released, in addition to

Fig. 5. Concordia diagram showing the results of the zircon dating of the lower sill in the194 Nepa borehole and the weighted average 206Pb/238U age of 252.0±0.4 Ma.

496 H. Svensen et al. / Earth and Planetary Science Letters 277 (2009) 490–500

minor volumes of magmatic volatiles (CO2, HCl, SO2). This is supportedby gas geochemistry from the Tunguska Basin, and methane-gen-eration during contact metamorphism is considered the final phase ofpetroleum formation in the basin (Sokolov et al., 1992). Moreover, highconcentrations of organic sulphur-compounds (mercaptans) have beenidentified in oil and condensates, attributed to the contact metamorph-ism (Kontorovich et al., 1997). Note that the same compounds wereidentified during gas extraction from the contact aureole salt samples(Table 1).

The basin scale gas production potential estimates shows thatcontact aureoles in the Southern Tunguska Basin could have generatedbetween9200 and 27,600GtC (equivalent to 33,700–101,000Gt CO2). Inaddition, the pipes released 1500–3700 Gt C (equivalent to 5400–13,500 Gt CO2). This estimated number of pipes is high (6400), but is stillconsidered realistic as it is within the same order of magnitude as thenumber of hydrothermal vent complexes in the Karoo Basin in SouthAfrica (Svensen et al., 2006, 2007) and in the Vøring Basin offshoreNorway (Svensen et al., 2004; Planke et al., 2005). In the Vøring Basin,there are about 2500 hydrothermal vent complexes in an 85,000 km2

Table 3U–Pb data for zircon

Characteristicsa Weightb Ub Th/Uc Pbcd 206Pb/204Pbe

207Pb/235Uf

±2 206Pb/238Ud

[g] [ppm] [pg] [abs]

194/15–868.7 m, mafic sillZ fr [1] b1 N920 1.98 1.6 1487 0.2821 0.0018 0.03984Z f rim [1] b1 N2890 2.03 1.2 5897 0.2814 0.0011 0.03981Z fr [1] b1 N2670 2.29 0.6 11435 0.27996 0.00078 0.03965Z fr [1] b1 N930 2.21 0.5 4335 0.2793 0.0011 0.03951Z fr im [1] b1 N740 1.72 1.6 1146 0.2775 0.0020 0.03923

194/35 – 860.8 m, mafic sillZ fr [1] 2 5095 2.25 1.4 13357 0.28202 0.00096 0.03993Z fr [1] b1 N2830 1.54 0.5 14651 0.28129 0.00077 0.03987Z fr im [1] 1 21800 3.45 1.2 46947 0.28102 0.00090 0.03975Z fr im [1] b1 N3440 1.73 0.5 16733 0.28081 0.00078 0.03972Z fr eu im [6] 2 3131 2.00 1.9 6141 0.28001 0.00090 0.03964Z fr eu im [5] 7 4232 1.66 2.2 33506 0.27882 0.00064 0.03943Z fr [1] b1 N280 2.63 0.9 1747 0.7063 0.0055 0.08726

a Z=zircon; fr=broken pieces of crystals, locally with euhedral faces (=eu); im=slight imb Weight and concentrations are known to better than 10%, except for those near the ca.c Th/U model ratio inferred from 208/206 ratio and age of sample.d Pbc=total common Pb in sample (initial+blank).e Raw data corrected for fractionation.f Corrected for fractionation, spike, blank and initial common Pb; error calculated by pro

large basin segment (i.e., one pipe per 34 km2). Assuming that each pipein the Tunguska Basin has a source region of 5 km3, the resulting volumeof high temperature sediment metamorphism for the 6400 pipes is32,000 km3. This leads to an aureole-volume to pipe-volume ratio of 25.Thepipes likely formed in regionswithahigher gas productionpotentialat the base of the Cambrian evaporite sequence, thus we use aconversion of 2–5 wt.% organic matter to gas.

5.2. Gas experiment

The heating-experiments, although being semi-quantitative, suggestthat a series of sulphur gases and halocarbons were generated duringmetamorphism in addition to the carbongases. To ourknowledge, this isthefirst experimental demonstration that natural rock samples generatehalocarbons during heating. We attribute the halocarbon generation toreactions between the petroleum bearing fluid inclusions and the hostsalt on the prograde temperature path to 275 °C. A requirement formethyl chloride generation from organic matter is the presence ofchloride ions and methyl groups (Conesa et al., 1997; Keppler et al.,2000; Hamilton et al., 2003). In the salt samples, the chloride is presentin the inclusion fluid water, and methyl groups are produced from thepetroleumduringheating. An equivalent process is likely responsible forthe formation of 1-chloro-butane, methyl bromide and 1-bromo-butane. Assuming that the gas speciation is not too pressure sensitive,the experiments show that halocarbons are natural products of contactmetamorphism of the Tunguska Basin evaporites.

5.3. Pipe processes

The pipes in the Tunguska Basinwere formed as a consequence of thepressure build-up during heating of the Precambrian and Cambrian se-quences (Von der Flaass and Naumov, 1995; Von der Flaass, 1997).Pressure build-up mechanisms contributing to pipe formation includeheating of pore fluids and metamorphic devolatilization of carbonate,shale, liquid hydrocarbon and brine reservoirs, and halite-anhydriterocks, generating gases like CH4, CO2, and SO2 (cf. Jamtveit et al., 2004;Svensen et al., 2007). Cracking of liquid petroleum to gas could have beenof major importance for pipe formation, as cracking at atmosphericconditions results in a 700 fold volume increase (Barker, 1990). Even athigher pressures, this conversionwould lead to accelerated hydrofractur-ing and vertical fluid expulsion. In the Tunguska Basin, reactionswith theevaporites likely generated sulphurous gases and halocarbons, as shown

±2 Rho 207Pb/206Pbf

±2 206Pb/238Uf

±2 207Pb/235Uf

±2 207Pb/206Pbf

±2

[abs] [abs] [age in Ma]

0.00014 0.60 0.05134 0.00026 251.9 0.8 252.3 1.4 256 120.00014 0.85 0.05125 0.00011 251.7 0.9 251.7 0.9 252.2 4.90.00011 0.85 0.05121 0.00008 250.7 0.7 250.6 0.6 250.3 3.40.00013 0.78 0.05126 0.00013 249.8 0.8 250.1 0.9 252.7 5.60.00012 0.53 0.05131 0.00032 248.0 0.7 248.7 1.6 255 14

0.00013 0.95 0.05123 0.00005 252.4 0.8 252.3 0.8 251.0 2.40.00010 0.89 0.05117 0.00006 252.0 0.6 251.7 0.6 248.6 2.90.00012 0.97 0.05127 0.00004 251.3 0.7 251.5 0.7 253.2 1.80.00010 0.92 0.05127 0.00005 251.1 0.6 251.3 0.6 253.1 2.40.00011 0.88 0.05124 0.00008 250.6 0.7 250.7 0.7 251.4 3.50.00008 0.96 0.05129 0.00003 249.3 0.5 249.7 0.5 253.9 1.50.00060 0.82 0.05871 0.00026 539.3 3.5 542.6 3.2 556.3 9.6

perfections (local turbidity, rust or inclusions) [N]=number of grains in fraction.1 g limit of resolution of the balance.

pagating the main source of uncertainty.

Fig. 6. Schematic evolution of the Tunguska Basin pipes and the venting of carbon gases and halocarbons to the atmosphere. The pipe evolution is partly based Von der Flaass andNaumov (1995) and Von der Flaass (1997). 1) Emplacement of sills into organic rich sediments and evaporites with petroleum accumulations (P). 2) Contact metamorphism of shale,evaporite, and petroleum, leading to gas generation and overpressure (shown as stippled lines). Melt is accumulating within evaporite sequences in the source region of the pipe. 3)Pipe formation and eruption. Glass in the breccias show that the magma was disrupted and fragmented in the source region before vertical transport and phreatomagmatism.Powerful eruptions led to wide craters and subsidence. Gases generated in contact aureoles are now released to the atmosphere. 4) Continued degassing from both magma andsediments through the pipe and the crater-lake. Contact metamorphism of shallow organic-rich sequences (coal) along dikes, and appearance of the first lava flows further to thenorth in the basin. The inferred gas composition is shown in the frame, alongside the estimated carbon gas and halocarbon production potential for the pipe degassing alone.

497H. Svensen et al. / Earth and Planetary Science Letters 277 (2009) 490–500

by our experiments. Gas generation from sediment metamorphism isknown both from the Karoo Basin in South Africa and offshore Norway,resulting in vertical piercement structures (Jamtveit et al., 2004; Svensenet al., 2004; Svensen et al., 2007). The main differences in the pipeforming mechanisms between these two settings and the one in Siberia,is the presence of evaporites and petroleum in the source region, andextensive magma-sediment interactions within the pipes. The pipes arerooted at 2–4 km depth in magma-sediment mixing zones, probably

Fig. 7. Compilation of zircon U–Pb ages of key end-Permian events. The ages of the P–T bolabelled “B”), and from a revised age by Mundil et al. (2001) (in grey, labelled “M”). The agerelease of carbon gases and halocarbons to the atmosphere. The main phase of the extinction(labelled “B”) and 252.6±0.2 Ma by Mundil et al. (2004) (labelled “M”).

close to the base of the evaporite stratigraphyalthough the precise natureof the roots remains unknown. Fig. 6 shows the schematic evolution ofthe pipe structures. The sizes of the pipe craters in the Tunguska Basinsuggest powerful eruptions, with gases and ash likely reaching highatmospheric levels. The presence of glass in the Scholokhovskoie brecciapipe suggests that the pipe was formed as a phreatomagmatic event,where the partly molten magma cooled rapidly in the pipe duringeruption. Parts of the wall-rock collapsed into the pipe, mixing with the

undary are from (Gradstein et al., 2004) deduced from Bowring et al. (1998) (stippled,of the sill emplacement at Nepa (this study) is synchronous with pipe formation and

is dated to have occurred between 252.3±0.3 and 251.4±0.3 Ma by Bowring et al. (1998)

498 H. Svensen et al. / Earth and Planetary Science Letters 277 (2009) 490–500

brecciated sediments from the root zone and the magmatic material.Degassing of metamorphic volatiles continued along with circulation ofbrines and metamorphic reactions between fluids and breccia, assuggested by the hydrothermal mineralogy. The world-class magnetitereserves of the Tunguska Basinpipes likely originated fromhydrothermalleaching of iron by hypersaline brines from the source region followed byprecipitation in the breccia (e.g., Mazurov et al., 2007).

5.4. Timing, timescales, and fluxes

The sill dating demonstrates that sill emplacement at Nepa oc-curred 252.0±0.4 Ma, overlapping the U–Pb age of lavas in thenorthern parts of the basin, and the U–Pb age of 252.6±0.2 Ma for themain phase of the end-Permian extinction (Fig. 7) (Mundil et al.,2004). This supports a causal link between the pipe degassing and theend-Permian environmental crisis.

If the contact metamorphic carbon gas wholly or partially vented tothe atmosphere, the result would be global warming even for a 100 kydegassing scenario (c.f., Berner, 2002; Retallack and Jahren, 2008). Thereleased gaswould have a δ13C isotopic composition close to that of bulkoil and organic matter (~−30 to −25‰) (Hunt, 1979). The widely usedδ13C compositionof volcanic CO2 is about−6‰, andwould in that casebetoo isotopically heavy to have caused the end-Permian negative carbonisotope excursion (c.f. Berner, 2002). The isotopic composition of thereleased CO2 from the Siberian Traps flood basalts is not known, but isassumed to be close to −6‰. Sill emplacement and pipe formation isconsidered a rapid process (Jamtveit et al., 2004; Svensen et al., 2004),andwe assume that onepipe formedper year in the Tunguska Basin.Westress that this is a hypothetical assumption in order to evaluate thecarbon gas fluxes. The assumption is nevertheless realistic, as only alimited volume ofmelt is required to formeven themostwidespread sillintrusions, and contact metamorphism and venting occurs on time-scales of 10–100 years (cf. Jamtveit et al., 2004). Still, we cannot excludethat the pipe eruptions could have been clustered in time or formedduring a period of a few tens of thousands of years. The chosen modelscenario implies a total duration for the pipe degassing of 6400 years,and an integrated CO2 equivalent flux of 0.8–2.1 Gt CO2/y. Seepage fromcontact aureoles started after the pipe degassing, releasing 0.7–2.0 GtCO2/y over a 50 ky period. Note that this is less than the 2000–2005average anthropogenic emissions of about 25–28 Gt CO2/y (IPCC, 2007).The CO2flux from the lavas is based upon the SiberianTraps lava volumeestimate by Reichow et al. (2002), a 1 million year emplacement period,and the CO2 basaltic lava degassing estimate by McCartney et al. (1990)and Self et al. (2005) (giving 0.024 and 0.017 Gt C/y, respectively). Notethat a few thick lava flows or sill intrusions would have led toperturbations in the carbon degassing, with major consequences forthe atmospheric composition.

5.5. Implications for the end-Permian ozone layer

Generation and venting of chlorinated and brominated halocarbons,both of which are synthesized in the evaporites, could have environ-mental consequences due to the involvement of halocarbons instratospheric ozone depletion (Beerling et al., 2007). Based on massbalance calculations, between4500 and13,500Gtmethyl chloride couldhave been produced during the basin-scale metamorphism (Table 2),and additional 720–1800 Gt during pipe degassing. The fluxes of methylchloride according to the model scenario are 0.11–0.28 Gt/y for the pipestage and 0.09–0.27 Gt/y for the seep stage. Recent atmospheric model-ling shows that N1000 Gt of methyl chloride are required to cause sig-nificant stratospheric ozone depletion over a 100,000 year time scale,with a resulting flux of ~0.1 Gt/y (Beerling et al., 2007). Methyl bromidewill escalate this process, and our results show that 2–5 Tg methylbromide could have been released during the pipe degassing stage peryear. Even if the halocarbon degassing plume was restricted to loweratmospheric levels, coeval emissions of methane would enable the

halocarbons to pass the troposphere without reacting with OH (seeBeerling et al., 2007).

To conclude, the metamorphism and venting in the Tunguska Basincan successfully beapplied as amechanism toexplain several of the end-Permian extinction hypotheses: a) extinction from global warming andhypercapnia (e.g., Knoll et al., 2007), b) extinction from ozone layercollapse and high levels of surface radiation (e.g., Visscher et al., 2004;Foster and Afonin, 2005; Kump et al., 2005; Beerling et al., 2007; La-marque et al., 2007).

5.6. The relevance for the Triassic–Jurassic mass extinction

Carbon isotope excursions and global warming were consequencesof contact metamorphism and venting during the initial stages of theformation of theNorthAtlantic Volcanic Province and theKaroo LIP (e.g.,Svensen et al., 2004, 2007). The main mechanism was contactmetamorphism of carbon-rich sediments and release of carbon gasesto the atmosphere. Even though LIPs andmassextinctionare statisticallycorrelated (e.g., Stothers, 1993), these LIPs are not associated withmajorextinctions. The key question becomes why some LIPs are associatedwith extinctions, whereas others had apparently minor effects on thebiosphere. If the sediment composition around sill intrusions is themostimportant parameter in causing environmental changes as suggested bySvensenet al. (2004, 2007),wewouldnowpredict that extinctions occurwhen sedimentary basins with major evaporite and organic matterdeposits are intruded by thick sill intrusions.

The secondbiggest knownextinction occurred at the Triassic–Jurassicboundary at about 200 Ma, contemporaneous with the formation of theCentral AtlanticMagmatic Province in the easternUS, SouthAmerica, andwestern Africa (e.g., Palfy and Smith, 2000;Wignall, 2001;Hesselbo et al.,2007). Thick sills are common in this province, emplaced in thickPermianevaporite deposits in the Amazon Basin in Brazil (e.g., Szatmari et al.,1979).We propose that evaporitemetamorphism caused generation andventing of carbon gases and halocarbons to the atmosphere, leading toglobal warming and atmospheric ozone depletion. Although theAmazonBasin needs to be investigated in detail to lend support to this hypothesis,the fact that the two biggestmass extinctions the last 252Ma occurred atthe same time as two major sill emplacement events into evaporitebasins, seems to be too much of a coincidence.

6. Conclusions

We conclude that the pipe structures in the Tunguska Basin providedirect evidence for end-Permian carbon gas and halocarbon release tothe atmosphere synchronous with sill emplacement. The end-Permiancrisis can be attributed to element mobilization from the evaporites(producing halocarbons) and organic-rich deposits (producing green-house gases like CH4 and CO2) in Siberia. For the first time, we show thathalocarbonsare generatedwhennatural evaporite samples areheated to275 °C, simulating the conditions during contact metamorphism. Basinscale gas production potential duringmetamorphism shows that carbongaseswere generated in sufficient quantities to explain the end-Permiannegative carbon isotope excursion and global warming. Moreover, ourresults explain how theSiberianTraps intrusions led to amass extinctionwhereas the formation of many other Large Igneous Provinces did not:two kilometres of intruded evaporite made the difference.

Acknowledgments

This study was supported by a Centre of Excellence grant from theNorwegian Research Council to PGP, by a YoungOutstanding Researchergrant and a PetroMaks grant to H. Svensen. We thank Morten Schjol-dager for zircon separation, and Paul Meakin, Claus Nielsen, FrodeStordal, Stephane Polteau, and Grigorii G. Akhmanov for comments tothe manuscript. Constructive reviews by Andrew D. Saunders and ananonymous referee further improved the manuscript.

499H. Svensen et al. / Earth and Planetary Science Letters 277 (2009) 490–500

References

Barker, C., 1990. Calculated volume and pressure changes during the thermal-crackingof oil to gas in reservoirs. Aapg Bull.-Am. Assoc. Pet. Geol. 74, 1254–1261.

Bartley, J.K., Pope, M., Knoll, A.H., Semikhatov, M.A., Petrov, P.Y.U., 1998. A Vendian–Cambrian boundary succession from the northwestern margin of the SiberianPlatform: stratigraphy, palaeontology, chemostratigraphy and correlation. Geol. Mag.135, 473–494.

Beerling, D.J., Harfoot, M., Lomax, B., Pyle, J.A., 2007. The stability of the stratosphericozone layer during the end-Permain eruption of the Siberian Traps. Philos. Trans.-Royal Soc., Math. Phys. Eng. Sci. 365, 1843–1866.

Berner, R.A., 2002. Examination of hypotheses for the Permo-Triassic boundary ex-tinction by carbon cycle modeling. Proc. Natl. Acad. Sci. U. S. A. 99, 4172–4177.

Bowring, S.A., Erwin, D.H., Jin, Y.G., Martin, M.W., Davidek, K., Wang, W., 1998. U/Pbzircon geochronology and tempo of the end-Permian mass extinction. Science 280,1039–1045.

Conesa, J.A., Marcilla, A., Caballero, J.A., 1997. Evolution of gases from the pyrolysis of modi-fied almond shells: effect of impregnationwith CoCl2. J. Anal. Appl. Pyrolysis 43, 59–69.

Corfu, F., 2004. U–Pb age, setting and tectonic significance of the anorthosite–mangerite–charnockite–granite suite, Lofoten–Vesteralen, Norway. J. Petrol. 45, 1799–1819.

Drobot, D.I., Pak, V.A., Devyatilov, N.M., Khokhlov, G.A., Karpyshev, A.V., Berdnikov, I.N.,2004. Petroleum potential of Precambrian deposits of the Siberian platform andprospects for studying their hydrocarbon content. Russ. Geol. Geophys. 45,110–120.

Erwin, D.H., Bowring, S.A., Yugan, J., 2002. End-Permian mass extinction: a review. In:Koeberl, C., MacLeod, K.G. (Eds.), Catastrophic events andmass extinctions: Impactsand beyond. Geological Society of America. Special Paper 356, Boulder, Colorado.

Federenko, V., Czamanske, G.K., 1997. Results of new field and geochemical studies ofthe volcanic and intrusive rocks of the Maymecha–Kotuy area, Siberian Flood-Basalt Province, Russia. Int. Geol. Rev. 39, 479–531.

Fjeldskaar, W., Helset, H.M., Johansen, H., Grunnaleiten, I., Horstad, I., 2008. Thermalmodelling of magmatic intrusions in the Gjallar Ridge, Norwegian Sea: implicationsfor vitrinite reflectance and hydrocarbon maturation. Basin Res. 20, 143–159.

Foster, C.B., Afonin, S.A., 2005. Abnormal pollen grains: an outcome of deteriorating at-mospheric conditions around the Permian–Triassic boundary. J. Geol. Soc.162, 653–659.

Galushkin, Y.I., 1997. Thermal effects of igneous intrusions on maturity of organicmatter: a possible mechanism of intrusion. Org. Geochem. 26, 645–658.

Gradstein, F.M., Ogg, J.G., Smith, A.G., Bleeker, W., Lourens, L.J., 2004. A new geologic timescale, with special reference to Precambrian and Neogene. Episodes 27, 83–100.

Grard, A., Francois, L.M., Dessert, C., Dupre, B., Godderis, Y., 2005. Basaltic volcanism andmass extinction at the Permo-Triassic boundary: environmental impact and model-ing of the global carbon cycle. Earth Planet. Sci. Lett. 234, 207–221.

Grice, K., Cao, C.Q., Love, G.D., Bottcher, M.E., Twitchett, R.J., Grosjean, E., et al., 2005. Photiczone euxinia during the Permian–Triassic superanoxic event. Science 307, 706–709.

Grishina, S., Pironon, J., Mazurov, M., Goryainov, S., Pustilnikov, A., Fon-Der-Flaas, G., et al.,1998. Organic inclusions in salt. Part 3. Oil and gas inclusions in Cambrian evaporitedeposit from East Siberia. A contribution to the understanding of nitrogen generationin evaporites. Org. Geochem. 28, 297–310.

Hamilton, J.T.G., McRoberts, W.C., Keppler, F., Kalin, R.M., Harper, D.B., 2003. Chloridemethylation by plant pectin: an efficient environmentally significant process. Science301, 206–209.

Hesselbo, S.P., McRoberts, C.A., Palfy, J., 2007. Triassic–Jurassic boundary events: problems,progress, possibilities. Palaeogeogr. Palaeoclimatol. Palaeoecol. 244, 1–10.

Hunt, J.M., 1979. Petroleum geochemistry and geology, San Francisco, California.IPCC, 2007. Climate change 2007. Fourth Assessment Report (AR4). Intergovernmental

Panel on Climate Change, Geneva, 2007.Jaffey, A.H., Flynn, K.F., Glendeni.Le, Bentley, W.C., Essling, A.M., 1971. Precision measure-

ments of half-lives and specific activities of U-235 and U-238. Physical Review C 4,1889–1906.

Jamtveit, B., Svensen, H., Podladchikov, Y., Planke, S., 2004. Hydrothermal vent complexesassociated with sill intrusions in sedimentary Basins. Physical geology of high-levelmagmatic systems. Geological Society Special Publication, London.

Kamo, S.L., Crowley, J., Bowring, S.A., 2006. The Permian–Triassic boundary event anderuption of the Siberian flood basalts: an inter-laboratory U–Pb dating study.Geochim. Cosmochim. Acta 70, A303-A303.

Keppler, F., Eiden, R., Niedan, V., Pracht, J., Scholer, H.F., 2000. Halocarbons produced bynatural oxidation processes during degradation of organic matter. Nature 403,298–301.

Knoll, A.H., Bambach, R.K., Payne, J.L., Pruss, S., Fischer, W.W., 2007. Paleophysiology andend-Permian mass extinction. Earth Planet. Sci. Lett. 256, 295–313.

Kontorovich, A.E., Khomenko, A.V., Burshtein, L.M., Likhanov, Pavlov, A.L., Staroseltsev,V.S., et al., 1997. Intense basic magmatism in the Tunguska petroleum basin, easternSiberia, Russia. Petrol. Geosci. 3, 359–369.

Kump, L.R., Pavlov, A., Arthur,M.A., 2005.Massive release of hydrogen sulfide to the surfaceocean and atmosphere during intervals of oceanic anoxia. Geology 33, 397–400.

Kuznetsov, V.G., 1997. Riphean hydrocarbon reservoirs of the Yurubchen–Tokhom Zone,Lena-Tunguska Province, NE Russia. J. Petrol. Geol. 20, 459–474.

Lamarque, J.F., Kiehl, J.T., Orlando, J.J., 2007. Role of hydrogen sulfide in a Permian–Triassic boundary ozone collapse. Geophys. Res. Lett. 34.

Malich, N.S., Tazihin, N.N., Tuganova, E.V., Bunzen, E.A., Kulikova, N.G., Safonova, I. V. a.p. o. A., V.B., et al., 1974. Map of geological formations of the Siberian platform cover(1:1500 000). In: Malich, N.S. (Ed.), All-Union Research Geologic Institute (VSEGEI).Leningrad.

Mazurov, M.P., Grishina, S.N., Istomin, V.E., Titov, A.T., 2007. Metasomatism and oreformation at contacts of dolerite with saliferous rocks in the sedimentary cover ofthe southern Siberian platform. Geol. Ore Depos. 49, 271–284.

McCartney, K., Huffman, A.R., Tredoux, M., 1990. A paradigm for endogenous causationof mass extinctions. In: Sharpton, V.L., Ward, P.D. (Eds.), Global Catastrophes inEarth History, 247. Geological Society of America, pp. 125–138. Special Paper.

Meyer, K.M., Kump, L.R., 2008. Biogeochemical controls on photic-zone euxinia duringthe end-Permain mass extinction. Geology 36, 747–750.

Meyerhoff, A.A., 1980. Geology and petroleum fields in Proterozoic and Lower CambrianStrata, Lena-Tunguska Petroleum Province, Easter Siberia, USSR. In: Halbouty, M.T.(Ed.), Giant Oil and Gas Fields of the Decade 1968–1978. American Association ofPetroleum Geologists.

Mundil, R., Metcalfe, I., Ludwig, K.R., Renne, P.R., Oberli, F., Nicoll, R.S., 2001. Timing ofthe Permian–Triassic biotic crisis: implications from new zircon U/Pb age data (andtheir limitations). Earth. Planet. Sci. Lett. 187, 131–145.

Mundil, R., Ludwig, K.R., Metcalfe, I., Renne, P.R., 2004. Age and timing of the Permianmass extinctions: U/Pb dating of closed-system zircons. Science 305, 1760–1763.

Nikulin, V. I. and Von-der-Flaass, G. S., 1985. Prognosis map of Baikal-Amur Railway areaon iron ores (1:500 000). Explanation Note. Irkutsk, East-Siberian Research Ins-titute of Geology, Geophysics and Mineral Resources (VostSibNIIGGiMS).

Palfy, J., Smith, P.L., 2000. Synchrony between Early Jurassic extinction, oceanic anoxicevent, and the Karoo-Ferrar flood basalt volcanism. Geology 28, 747–750.

Payne, J.L., Kump, L.R., 2007. Evidence for recurrent Early Triassic massive volcanismfrom quantitative interpretation of carbon isotope fluctuations. Earth. Planet. Sci.Lett. 256, 264–277.

Payne, J.L., Lehrmann, D.J., Wei, J.Y., Orchard, M.J., Schrag, D.P., Knoll, A.H., 2004. Largeperturbations of the carbon cycle during recovery from the end-Permian extinction.Science 305, 506–509.

Petrychenko, O.Y., Peryt, T.M., Chechel, E.I., 2005. Early Cambrian seawater chemistryfrom fluid inclusions in halite from Siberian evaporites. Chem. Geol. 219, 149–161.

Planke, S., Rassmussen, T., Rey, S.S., Myklebust, R., 2005. Seismic characteristics anddistribution of volcanic intrusions andhydrothermalvent complexes in theVøringandMøre basins. In: Dore, T., Vining, B. (Eds.), PetroleumGeology: North-West Europe andGlobal Perspectives - Proceedings of the 6th Petroleum Geology Conference. Geo-logical Society Publishing House, London.

Polyanskii, O.P., Reverdatto, V.V., 2002. Fluid convection in sediment-hosted reservoirsdue to thermal action of dikes and sills. Geol. Geofiz. 43, 27–41.

Pukhnarevich, M.M., 1986. Conditions and formation peculiarities of endogenous iron oredeposits of the Siberian platform South. Irlutsk, Irkutsk State University Publishing.

Raymond,A.C.,Murchison, D.G.,1989.Organicmaturation and its timing in a carboniferoussequence in the Central Midland Valley of Scotland: comparisons with NorthernEngland. Fuel 68, 328–334.

Raymond, A.C., Murchison, D.G., 1991. The relationship between organic maturation, thewidths of thermal aureoles and the thicknesses of sills in the Midland Valley ofScotland and Northern England. J. Geol. Soc. 148, 215–218.

Reichow,M.K., Saunders, A.D.,White, R.V., Pringle,M.S.,Al'Mukhamedov,A.I.,Medvedev,A.I.,et al., 2002. Ar-40/Ar-39 dates from the West Siberian Basin: Siberian flood basaltprovince doubled. Science 296, 1846–1849.

Retallack, G., Jahren, A.H., 2008. Methane release from igneous intrusion of coal duringLate Permian extinction events. J. Geol. 116, 1–20.

Retallack, G.J., Krull, E.S., 2006. Carbon isotopic evidence for terminal-Permian methaneoutbursts and their role in extinctions of animals, plants, coral reefs, and peatswamps. In: Greb, S.F., DiMichele, W.A. (Eds.), Wetlands through time. GeologicalSociety of America. Special Paper 399.

Ryabov, V.V., 2006. Petrologic features of paleovolcanic apparatus at tuffs and lavas unit ofthe Siberian craton. Proceedings of 3d All-Russian Symposium on Volcanology andPalaeovolcanology. Buryat Science Center of Siberian Branch of Russian Academy ofScencies, Ulan-Ude, Buryat Republic, Russia.

Ryabov, V.V., Shevko, A.Y., Zateeva, S.N., 2005. Abnormal formations” in the Siberianplatform traps — indicators of geodynamic conditions of plateau basalts forming.Lithospheres 4, 165–177.

Self, S., Thordarson, T., Widdowson, M., 2005. Gas fluxes from flood basalt eruptions.Elements 1, 283–287.

Sokolov, B.A., Egorov, V.A., Nakaryakov, V.D., Bitner, A.K., Zkukovin, Y.A., Kuznetzov, L.L.,et al.,1992. Geological and geophysical conditions of formation of oil and gas bearingdeposits in the ancient rocks of Eastern Siberia. Petroconsultants Australasia, Syd-ney, Australia.

Stothers, R.B. (Ed.), 1993. Flood basalts and extinction events. Geophysical ResearchLetters, vol. 20, pp. 1399–1402.

Surkov, V.S., Grishin, M.P., Larichev, A.I., Lotyshev, V.I., Melnikov, N.V., Kontorovich, A.E.,et al., 1991. The Riphean sedimentary basins of the Eastern Siberia Province andtheir petroleum potential. Precambrian Res. 54, 37–44.

Svensen, H., Planke, S., Malthe-Sorenssen, A., Jamtveit, B., Myklebust, R., RasmussenEidem, T., et al., 2004. Release of methane from a volcanic basin as a mechanism forinitial Eocene global warming. Nature 429, 542–545.

Svensen, H., Jamtveit, B., Planke, S., Chevallier, L., 2006. Structure and evolution of hy-drothermal vent complexes in the Karoo Basin, South Africa. J. Geol. Soc. 163, 671–682.

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

Szatmari, P., Carvalho, R.S., Simoes, I.A.,1979. Comparison of evaporite facies in the Late Paleo-zoic Amazon and theMiddle Cretaceous South-Atlantic salt basins. Econ. Geol. 74, 432-&.

Ulmishek, G.F., 2001. PetroleumGeology and Resources of the Baykit High Province, EastSiberia, Russia. U.S. Geological Survey Bulletin. 2201-F.

Vasil'ev, Y.R., 1999. A quantitative estimation of large-volume occurrences of Permian–Triassic trap magmatism, Siberian Platform. Dokl. Earth Sci. 367, 754–757.

Visscher, H., Looy, C.V., Collinson, M.E., Brinkhuis, H., Cittert, J., Kurschner, W.M., et al.,2004. Environmental mutagenesis during the end-Permian ecological crisis. Proc.Natl. Acad. Sci. U. S. A. 101, 12952–12956.

500 H. Svensen et al. / Earth and Planetary Science Letters 277 (2009) 490–500

Von der Flaass, G.S., 1997. Structural and genetic model of an ore field of the Angaro-Ilimtype (Siberian Platform). Geol. Ore Depos. 39, 461–473.

Von der Flaass, G.S., Naumov, V.A., 1995. Cup-shaped structures of iron ore deposits inthe South of the Siberian Platform (Russia). Geol. Ore Depos. 37, 340–350.

Wignall, P.B., 2001. Large igneous provinces and mass extinctions. Earth-Sci. Rev. 53,1–33.

Xie, S.C., Pancost, R.D., Huang, J.H., Wignall, P.B., Yu, J.X., Tang, X.Y., et al., 2007. Changesin the global carbon cycle occurred as two episodes during the Permian–Triassiccrisis. Geology 35, 1083–1086.

Zamaraev, S.M., Geletii, N.K., Malykh, A.V., 1985. Temperature effects of traps onpotassium salts. Sov. Geol. Geophys. 26, 37–42.

Zharkov, M.A., 1984. Paleozoic Salt Bearing Formations of the World. Springer-Verlag,Berlin, Heidelberg, New York & Tokyo.