sign in sign up
<<
Home , Carnian, Carnian Pluvial Event, Daxatina, Ladinian

Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 222–232

Contents lists available at ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology

journal homepage: www.elsevier.com/locate/palaeo

Cause of Upper crisis revealed by Re–Os geochemistry of Boreal black shales

Guangping Xu a,b,⁎, Judith L. Hannah a,b,c,HollyJ.Steina,b,c, Atle Mørk d,e, Jorunn Os Vigran d, Bernard Bingen b, Derek L. Schutt f, Bjørn A. Lundschien g a AIRIE Program, Colorado State University, Fort Collins, CO 80523-1482, USA b Geological Survey of Norway, 7491 Trondheim, Norway c CEED, University of Oslo, 0316 Oslo, Norway d SINTEF Petroleum Research, NO-7465 Trondheim, Norway e Norwegian University of Sciences and Technology, NO-7491 Trondheim, Norway f Department of Geosciences, Colorado State University, Fort Collins, CO 80523-1482, USA g Norwegian Petroleum Directorate, NO-4003 Stavanger, Norway article info abstract

Article history: The Triassic Period is bracketed by two of the ‘big five’ Phanerozoic mass . Though long viewed as a Received 2 April 2013 period of climatic stability, emerging data suggest multiple climatic swings and at least one severe ecological cri- Received in revised form 13 December 2013 sis. Linking these climatic instabilities with probable causes is hampered by poor age control within the Triassic Accepted 19 December 2013 time scale. Here we present new Re–Os ages for shale sections straddling Middle–Upper Triassic bound- Available online 29 December 2013 aries. Nominal Re–Os isochron ages of 236.6 and 239.3 Ma for the top and base of the (upper ) bring absolute time into the contentious Triassic time scale, and place the beginning of the Keywords: 187 188 Triassic about 12 m.y. earlier than previously assigned. A marked decrease in initial Os/ Os in Upper Ladinian shale Ladinian– boundary records input from Wrangellian flood — an instigator in the Carnian (Late Triassic) Pluvial Event and ac- Kong Karls Land companying radiation of key groups (e.g., and calcareous nanoplankton). An absolute time scale Svalis Dome is proposed for the –Ladinian–Carnian boundaries based on Re–Os geochronology. Svalbard © 2013 Elsevier B.V. All rights reserved. Re–Os geochemistry

1. Introduction a global climate and oceanographic perturbation (Furin et al., 2006; Dal Corso et al., 2012). Accurate age information is essential to interpret The Carnian Stage (Upper Triassic) is characterized by a major eco- processes underlying these biotic crises. logical crisis (Simms and Ruffell, 1989). The crisis began with a major The radiometric time scale for Late Triassic ties the magnetostratigraphy decrease in carbonate productivity in the early Carnian, followed by a of Tethyan marine sections to the cycle-scaled terrestrial Newark significant episode of very humid conditions known as the Carnian Plu- magnetic polarity chrons (Gradstein et al., 2004, 2012). Subsequent vial Event (CPE). The CPE spurred a marked turnover of faunal and floral intervals are scaled assuming equal duration of ammonoid subzones. assemblages in Late Julian (early Carnian). Subsequently, a major ma- Re–Os geochronology for black shales has been successful for rine at the Julian–Tuvalian (Late Carnian) boundary impacted documenting depositional ages and paleoenvironmental changes many ammonoid and taxa, even as some terrestrial groups in- (Cohen and Coe, 2002; Hannah et al., 2004). New Re–Os isochron ages creased in diversity with increasing humidity (Raup and Sepkoski, for stratigraphically well-constrained sedimentary rocks from Triassic 1982; Simms and Ruffell, 1989; Hornung et al., 2007; Dal Corso et al., sections in the Boreal realm are presented. The Ladinian–Carnian 2012). With the return to more arid conditions in the Late Carnian, stage boundary, as defined by and ammonoid , is many water-dependent terrestrial taxa died off, while other groups bracketed using the upper Ladinian Botneheia and lower Carnian (calcareous nanoplankton, scleractinian reef builders, and dinosaurs) Tschermakfjellet formations from east of Kong Karls Land in the north- expanded rapidly (Furin et al., 2006). The cause of the Carnian crisis is ern Barents Sea. These new ages, along with our recent age constraints not fully understood (Preto et al., 2010), but the prevailing explanation on the Anisian–Ladinian stage boundary (Xu et al., 2009), confirm re- is that volcanism related to the Wrangellian oceanic plateau, now ac- cent revisions proposed for the controversial Middle–Late Triassic creted to the Cordilleran margin of North America, could have triggered time scale (Ogg et al., 2008; Gradstein et al., 2012). In addition, the data define variations in the initial 187Os/188Os ratio, a tracer for global changes in magmatism, continental , and ⁎ Corresponding author at: AIRIE Program, Colorado State University, Fort Collins, CO 80523-1482, USA. Tel.: +1 970 491 3816; fax: +1 970 491 6307. atmospheric chemistry (Cohen and Coe, 2002; Cohen et al., 2004; 187 188 E-mail address: [email protected] (G. Xu). Hannah et al., 2004). A sharp drop in initial Os/ Os ratios in the

0031-0182/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.palaeo.2013.12.027 G. Xu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 222–232 223 early–mid Carnian may reflect atmospheric and oceanic input from may be within the upper part of drill core 7831/02-U-01 or lower part of Wrangellian flood basalts, affirming their role as a trigger for the 7831/02-U-01 (Fig. 2, Table S1 and Vigran et al., in press). and accompanying Carnian ecological crisis. The upper part of core 7831/02-U-01 (from 19 to 14 m depth) differs lithologically from core 7831/02-U-02; it belongs to Tschermakfjellet Formation which grades into the De Geerdalen Formation (Riis et al., 2. Description of samples and stratigraphy 2008)(Figs. 1b,c and 2). The Tschermakfjellet Formation consists of gray, non-bituminous shales in a coarsening-upwards sequence Shales from two shallow stratigraphic drillholes (7831/02-U-01 reflecting prodeltaic conditions. These shales have only moderate or- and 7831/02-U-02), cored by the Norwegian Petroleum Directorate ganic contents compared to the underlying core (Fig. 3) and are similar east of Kong Karls Land (Riis et al., 2008), were sampled for Re–Os to the Tschermakfjellet Formation as seen on Edgeøya (Lock et al., dating (Fig. 1). The 7831/02-U-02 core represents the Botneheia 1978), with sparse fossils and siderite beds, and abundant pyrite nod- Formation, which comprises organic-rich shale containing cherty hori- ules. The presence of the Aulisporites astigmosus Composite Assemblage zons and phosphate nodules. The units are markedly similar to expo- Zone in this interval indicates an early Carnian age (Table S1) (Vigran sures in eastern Svalbard (Krajewski, 2008; Riis et al., 2008) where the et al., in press). The gray shales in the lower part of the core 7831/02- Botneheia Formation is ca. 90–110 m thick and spans lower Anisian to U-01 (from 28 to 19 m depth) contain sparse phosphate nodules and upper Ladinian deposition. Its upper part, known as the Blanknuten cherty horizons, as observed more abundantly in the dark shales of Member, is a notably organic-rich, calcareous black shale with locally the Botneheia Formation in core 7831/02-U-02 and at Edgeøya abundant phosphate nodules. The formation was deposited in a re- (Krajewski, 2008). The lower part of core 7831/02-U-01 may represent gressive and mostly anoxic restricted shelf environment (Mørk et al., a gradual transition from the Botneheia to phosphate- and chert-free 1999; Krajewski, 2008). The lithology of core 7831/02-U-02 resembles Tschermakfjellet Formation. Palynological evidence at the base of core that of the Blanknuten Member on Edgeøya (Krajewski, 2008). Three 7831/02-U-01 as in the core below is equivocal, permitting either a 14–18 cm intervals of black shale from the Botneheia Formation were late Ladinian or early Carnian age (Table S1). The clear shift in sedimen- sampled from 7831/02-U-02 (Fig. 2). tary regime between the two cores resembles the stratigraphic succes- The recognition of the Echinitosporites iliacoides Composite sion on eastern Svalbard (Lock et al., 1978; Mørk et al., 1999; Krajewski, Assemblage Zone indicates that the lower part of core 7831/02-U-02, 2008). Two sampled intervals, 27 and 31 cm thick, in 7831/02-U-01 from 18.79 to 13.88 m, is Ladinian (Table S1) (Vigran et al., in press). were taken from the best laminated sections between 20.74 and The ammonoid ex gr. Nathorstites sp. Juv. was found in a phosphate 19.13 m (Fig. 2). nodule at 12.32–12.27 m of core 7831/02-U-02 (Fig. S1), which can be Riis et al. (2008) proposed that prodelta shales of the of late Ladinian (Nathorstites mclearni) or early Carnian age (Nathorstites Tschermakfjellet Formation prograded from SE to NW. Lithostratigra- lindstroemi and Nathorstites mcconnelli). Nathorstites specimens from phy and together (Weitschat and Dagys, 1989) show core 7831/02-U-02 have globular internal whorls which indicate a late that the base of the Tschermakfjellet Formation and its equivalents, Ladinian age (Wolfgang Weitschat, personal communication, 2007). the Skuld and Snadd formations, are time-transgressive (Fig. 1c). The deposits from the upper part of the 7831/02-U-02 core (11.43 to The Global Boundary Stratotype Sections and Point (GSSP) for 6.16 m) contain a palynological assemblage associated mainly with an the Ladinian–Carnian boundary is placed at Prati di Stuores, , early Carnian age (Table S1). Therefore, the Ladinian–Carnian boundary with the first appearance datum of the ammonoid canadensis

a 15°E 25°E b W 7831/02-U-01 sea floor E 80°N De Geerdalen Fm Svalbard Kong Karls Carnian Tschermakfjellet – Snadd Fm Land 7831/02-U-02 Edgeoya Proposed Ladinian – CarnianLadinian boundary Botneheia Fm

Barents Sea c Central Svalbard Edgeøya East of KKL Bjørnøya Svalis Dome

Norwegian - Greenland Sea De Geerdalen Fm De Geerdalen Fm ° De Geerdalen Fm Bjornoya 75 N Carnian Tschermakfjellet Fm Tschermakfjellet Fm Tschermakfjellet - Snadd Fm Skuld Fm Snadd Fm Svalis Dome

Ladinian N Botneheia Fm Botneheia Fm Botneheia Fm

Anisian Verdande Bed Steinkobbe Fm 200 km Norway

Fig. 1. (a) Map showing the sampling locations off Kong Karls Land (KKL), Svalbard and Svalis Dome (dots). (b) Schematic sketch showing relative positions of the two drillholes, ~200 m apart, immediately east of Kong Karls Land. The solid thick line between the Botneheia and Tschermakfjellet–Snadd formations represents the prominent seismic reflector interpreted as the formation boundary. The proposed Ladinian–Carnian boundary (thick dash line) is within the upper part of Botneheia Formation in core 7831/02-U-02 (details in text). The estimated stratigraphic gap between the two cores is from a few meters to 30 m (Riis et al., 2008). (c) Anisian to Carnian lithostratigraphic correlations for selected sections from the Barents Sea region (after Mørk et al. (1999)). 224 G. Xu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 222–232

Age (Ma) 230 235 240

15 7831/02-U-01 Claystone - shale Limestone

Dolomite

No core

Nodule

Lower Carnian ± 232 11 Ma Phosphate cement Calcite cement 20 236.5 ± 3.5 Ma Pyrite Chert Ammonoid Bivalve shrafelt-Tschermakfjellet Snadd Fm

25

28 (m) unsampled interval 6 7831/02-U-02 Upper Ladinian - Lower Carnian

236.6 ± 0.4 Ma 10 Fm ia e 237.6 ± 0.6 Ma eh Nathorstites otn B an i

in 15 241.5 Ma 237 Ma ad L

238.4 ± 1.5 Ma

19 (m) 239.3 ± 2.6 Ma

Svalis Dome (uppermost Steinkobbe Fm) 241.2 ± 2.1 Ma Svalbard

Anisian (Botneheia Fm)

Carnian Ladinian Anisian

Fig. 2. Stratigraphic columns and Re–Os ages for the two cored sections from east of Kong Karls Land. The small black rectangles pinpoint sampled intervals and the arrow for the location of ammonoid ex gr. Nathorstites sp. Juv. Uncertainties shown for ages are analytical and do not include the 0.31% decay constant uncertainty (Smoliar et al., 1996)sinceRe–Os ages are not compared to those from other isotopic systems. Stage boundary ages (dashed vertical lines) are from Gradstein et al. (2012).Re–Os ages for Svalbard and Svalis Dome are from Xu et al. (2009).

at the base of Carnian (Mietto et al., 2007, 2008, 2012). In the Riis et al., 2008); In the Barents Sea, D. canadensis occurs only on Boreal realm D. canadensis was traditionally assigned to the upper- Bjørnøya, 40 m below the top of the 140-m-thick Skuld Formation most Ladinian Stage (Weitschat and Dagys, 1989; Dagys and and very close to the Ladinian–Carnian stage boundary as defined by Weitschat, 1993; Hounslow and Nawrocki, 2008). In Svalbard, the Mørk et al. (1990). The palynological assemblages from the Ladinian–Carnian stage boundary is roughly equivalent to the D. canadensis Subzone at the GSSP contain Camerosporites secatus Botneheia–Tschermakfjellet Formation boundary (Mørk et al., 1999; (Mietto et al., 2007). Camerosporites secatus is also present at the G. Xu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 222–232 225

depth (m) TOC (wt%) Re (ppb) 192 Os (ppb) 187Re/ 188Os Oxygen Index V/Mo Mo/TOC Re/Mo Pyrite size (µm) Pyrite size std (µm) 51020 0 400 0.0 0.5 1.0 010002000 10 20 50 1000.02 1 100 10 100 10 20 51015 18

21

24 7831/02-U-01 27 unsampled interval Tschermakfjellet - Snadd Fm. 7

10

13 7831/02-U-02 Botneheia Fm. 16

19

Fig. 3. Vertical distribution of TOC, Re and 192Os concentrations, oxygen index, ratios of V/Mo, Mo/TOC and Re/Mo, and the average size (diameter) and its standard deviation of pyrite framboids in the cores. uppermost level (i.e., 6.16 m) of 7831/02-U-02 in the off Kong Karls underlying 31-cm interval (20.43–20.74 m) also define a scattered Land cores. The newly defined GSSP for the Ladinian–Carnian corre- isochron; however, data for a 22-cm sub-interval (20.46–20.68 m) sponds to a stage boundary placed within the uppermost Botneheia For- yield a Model 1 age of 236.5 ± 3.5/3.6 Ma (n = 7, MSWD = 0.1) mation in core 7831/02-U-02 (Fig. 2, Table S1). with initial 187Os/188Os ratio of 0.732 ± 0.041 (Fig. 4d). Restricting the stratigraphic interval of sampling can eliminate scatter caused by 3. Results rapid temporal variations in seawater 187Os/188Os or significant deposi- tional hiatuses not recognized in the core. The larger uncertainty in this Shales from the Botneheia and Tschermakfjellet Formations are isochron compared to those from the Botneheia Formation arises in part geochemically distinct (Table 1; Fig. 3). Botneheia shales have higher from a limited range in 187Re/188Os and 187Os/188Os ratios (Fig. 4), but total organic carbon (TOC), Re, Os, trace metals, Mo/TOC ratio, lower may also be attributed to minor oxidation (Fig. 3). V/Mo and Re/Mo ratio and oxygen index, and smaller pyrite framboids, The isochron ages overlap within uncertainties, but their age rela- pointing to dominantly anoxic pore waters (Fig. 3). Shales from the tionships determined by the stratigraphic locations are known; that is, uppermost Botneheia Formation (intervals 7831/02-U-02, 11.12– samples from stratigraphically higher locations are younger than sam- 11.30 m and 10.18–10.34 m) have markedly larger variations in TOC, ples from stratigraphically lower locations. This information constitutes Re, 187Re/188Os, and 187Os/188Os, suggesting a more condensed section a known prior distribution; therefore, Bayesian statistical analysis can (Table 1, Fig. 3). be applied to obtain a posterior distribution to refine the age uncer- Re–Os isochron errors are reported both with and without the un- tainties (Biasi and Weldon, 1994) (refer to Supplementary information certainty in the 187Re decay constant (Table 2 and Fig. 4). The first num- for more details). The age uncertainties after applying the Bayesian ber given is the propagated error including uncertainties from Re and Os analysis are no longer symmetric and the overlaps decrease significantly measurements, weighing, spikes, oxygen and blank corrections, and un- (Table 2). certainty from natural 187Re/185Re ratio (Gramlich et al., 1973). The sec- ond number (after “/”) incorporates the decay constant uncertainty (Smoliar et al., 1996). Excluding the decay constant uncertainty allows 4. Discussions detection of internal variations in a of samples, as that uncertainty would offset all results in the same direction. Including the decay con- 4.1. The paleo-redox conditions stant uncertainty permits comparison with ages determined by other methods. The transition from the Botneheia to Tschermakfjellet formations Regressions of Re–Os data for three intervals from the Botneheia shows a pronounced lithologic change reflecting the transition from Formation in drill core 7831/02-U-02 yield Model 1 ages of 238.41 ± deep shelf to prodeltaic conditions (Mørk et al., 1982, 1999). The transi- 1.47/1.65 Ma (18.18–18.32 m, n = 8), 237.55 ± 0.59/0.95 Ma (11.12– tion is accompanied by an abrupt decrease in trace metal and TOC con- 11.30 m, n = 7) and 236.61 ± 0.39/0.83 Ma (10.18–10.34 m, n = 6), tents and quality of source rock (Fig. 3). Increased V/Mo and Re/Mo and with initial 187Os/188Os ratios of 0.791 ± 0.008, 0.659 ± 0.006 and decreased Mo/TOC ratios from the Botneheia to Tschermakfjellet forma- 0.730 ± 0.003, respectively (Fig. 4a–c). All three isochrons have mean tions (Fig. 3) suggest a change in depositional conditions from domi- square of weighted deviates (MSWD) near 1. Each interval preserves a nantly anoxic to oxic (or suboxic) (Kendall et al., 2010; Xu et al., unique age, nearly indistinguishable within two-sigma (95%) uncertainty, 2012). It is possible — even probable — that pore waters in unconsoli- with nominal ages in the correct stratigraphic order; each interval pre- dated organic-rich muds are less oxic than the overlying water column. serves a statistically distinguishable initial 187Os/188Os ratio. These data most likely reflect conditions in pore waters prior to expul- Re–Os isotopic data from the uppermost interval of the sion upon compaction; that is, the geochemical data (and Re–Os ages) Tschermakfjellet Formation in drill core 7831/02-U-01 (19.13– reflect earliest diagenesis, or conditions several cm below the sedi- 19.40 m), show excess scatter and define a Model 3 age of 232 ± ment–water interface. 11 Ma (n = 12, MSWD = 94) with an initial 187Os/188Os ratio of Pyrite framboids in sediments of anoxic/euxinic conditions are on 0.72 ± 0.11 (Table 2 and Fig. S2). The scatter is not reduced by average smaller and less variable in size than those from sediments in restricting the sampling interval. Thus, the scatter most likely results suboxic or oxic conditions (Wilkin et al., 1996). From the Botneheia to from oxic conditions that mobilized Re and/or Os. Re–Os data for an the Tschermakfjellet formation, pyrite framboids on average increase 226 G. Xu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 222–232

a b

± ± ± ± i i

c d

± ± ± i i ±

Fig. 4. Re–Os Model 1 age regressions for black shales from the Botneheia and Tschermakfjellet formations using Isoplot 3.60 (Ludwig, 2003) and a 187Re decay constant of 1.666 × 10−11 (Smoliar et al., 1996). The 2σ error ellipses are encased in squares for clarity. Re–Os ages are reported with 2σ uncertainties with values after "/" including the 0.31% uncertainty in the 187Re decay constant.

in both size and size variations, suggesting more oxic conditions at or 2012 ICS-approved (Gradstein et al., 2012), and just below the sediment–water interface (Figs. 3 and 5). the recently published U–Pb zircon date of 237.77 ± 0.14 Ma from an ash bed that is one ammonite subzone below the Ladinian–Carnian 4.2. The Triassic time scale boundary (Mietto et al., 2012). Based on our Re–Os ages, the Ladinian Stage was brief, nominally 1.8–3.1 m.y. (from 239.3 to 236.2– The Ladinian–Carnian boundary is constrained by three isochron 237.5 Ma) (Table 2), confirming the latest version of ICS-approved Tri- ages from the uppermost Botneheia and Tschermakfjellet formations, assic time scale (Gradstein et al., 2012). 0:56 0:38 0:67 Many efforts have been made to constrain the Anisian–Ladinian 237:5 Æ Ma, 236:6 Æ Ma and 236:2 Æ Ma (Table 2, 0:59 0:39 3:00 boundary using 206Pb/238Uand40Ar/39Ar methods (Mundil et al., Fig. 2). These ages place the Ladinian–Carnian stage boundary between 1996, 2010; Pálfy et al., 2003; Brack et al., 2007; Brühwiler et al., 233.2 and 238.1 Ma with a nominal age of 236.2–237.5 Ma. This is con- 2007; Furrer et al., 2008). In a recent review, Mundil et al. (2010) pro- sistent with 237 Ma for the age of Ladinian–Carnian boundary in the posed 242.1 ± 0.6 Ma as the boundary age instead of 240.5 Ma (Ogg et al., 2008), arguing that the younger age, determined by zircon pre- Pyrite framboid treated by air abrasion, was incorrect. The 242.1 Ma zircon age is from Mean diameter (µm) a volcanic ash layer (MSG.09) in the uppermost Anisian at Monte San Boundary Giorgio, close to the GSSP at the base of the Nevadites secedensis Zone Wilkin et al.,1996 at Bagolino, Italy (Fig. 2 of Mundil et al., 1996)(Fig. 6). Alkali feldspars extracted from the non-magnetic fraction of residue from zircon separa- 8 tion yield a slightly younger 40Ar/39Ar age of 241.0 Ma (after adjusting Anoxic the original 239.5 ± 0.5 Ma to the revised age of 28.201 Ma for the 6 Fish Canyon sanidine standard; (Kuiper et al., 2008)) (Mundil et al., 2010). Brack et al. (2007) also obtained an age of 242.8 ± 0.2 Ma for 4 zircon in the Latemar (LAT30) from a stratigraphic Oxic 7831/02-U-01 level within the uppermost Reitziites reitzi Zone, which underlies the Euxinic Suboxic Tschermakfjellet Fm N. secedensis Zone (Fig. 6). In the newly revised Geological Time Scale, 2 7831/02-U-02 Gradstein et al. (2012) assigned 241.5 ± 1.5 Ma as the age for Botneheia Fm Anisian–Ladinian boundary by scaling radiometric U–Pb zircon ages 0 across the boundary. 02468101214 The 241.5 ± 1.5 Ma boundary age for Anisian–Ladinian is still with- Standard deviation (µm) in the upper limit of Re–Os isochron ages of 239.3 ± 2.7 Ma near the boundary and 241.2 ± 2.2 Ma at 10–22 m below the boundary (Xu Fig. 5. Pyrite framboid size distribution for shales from the Botneheia and Tschermakfjellet et al., 2009). The Re–Os isochron ages obtained from Late Anisian to formations east of Kong Karls Land. The boundary between the oxic–suboxic field and the anoxic–euxinic field is from Wilkin early Carnian are consistently young upwards (Table 2). Thus, the nom- et al. (1996). inal difference between Boreal Re–Os isochron ages and Tethyan U–Pb Table 1 Re–Os isotopic data, Rock-Eval pyrolysis data, V and Mo contents, and the size of pyrite framboids for black shales from drill cores east of Kong Kars Land.

AIRIE Sample name Re–Os isotopic data Rock-Eval data Trace metals Pyrite framboid size distribution run # 192 187 187 Re 2σ Os 2σ Os Re/ 2σ Os/ 2σ Rho Re Os S1 S2 S3 TOC Tmax Hydrogen Oxygen V Mo Average Standard Maximum Number (ppb) (ppb) (ppb) 188 Os 188 Os blank blank (mg (mg (mg (wt%) (C) index index (ppm) (ppm) diameter deviation diameter of pyrite c (%)d (%)d HC/g HC/g HC/g (μm) (μm) (μm) framboid rock) rock) rock)

7831/02-U-01 Tschermakfjellet Formation ORG- 7831/02-U-01, 12.643 0.040 0.1894 0.00095 0.061 414.9 1.4 2.350 0.003 0.140 0.81 0.21 1.09 2.47 0.15 1.75 433 141 9 29 0.29 16.40 11.8 51.3 172 481 19.130–19.165 m (A)a ORG- 7831/02-U-01, 21.198 0.057 0.2691 0.00248 0.083 505.6 1.5 2.670 0.005 0.243 0.27 0.76 661 19.130–19.165 m (B) ORG- 7831/02-U-01, 11.823 0.029 0.1957 0.00135 0.064 368.1 1.0 2.156 0.004 0.224 0.88 0.21 0.84 2.61 0.14 1.74 433 150 8 29 0.49 482 19.165–19.170 m 222 (2014) 395 Palaeoecology Palaeoclimatology, Palaeogeography, / al. et Xu G. ORG- 7831/02-U-01, 10.766 0.027 0.1823 0.00102 0.060 358.5 1.0 2.117 0.003 0.178 0.92 0.21 0.89 3.57 0.22 1.75 439 204 13 23 0.32 483 19.170–19.180 m ORG- 7831/02-U-01, 22.501 0.063 0.2560 0.00214 0.077 580.8 1.8 2.974 0.005 0.201 0.45 0.16 0.71 2.95 0.20 1.66 439 177 12 29 0.62 9.87 9.34 38.7 119 484 19.330–19.355 m (A) ORG- 7831/02-U-01, 18.670 0.047 0.2183 0.00129 0.066 563.4 1.5 2.940 0.004 0.178 0.34 1.04 662 19.330–19.355 m (B) ORG- 7831/02-U-01, 17.700 0.046 0.2174 0.00162 0.067 527.0 1.5 2.760 0.004 0.156 0.57 0.18 0.74 3.12 0.23 1.67 439 187 14 29 0.47 485 19.355–19.370 m (A) ORG- 7831/02-U-01, 26.195 0.070 0.2535 0.00184 0.072 720.9 2.1 3.559 0.005 0.191 0.21 0.79 663 19.355–19.370 m (B) ORG- 7831/02-U-01, 22.745 0.056 0.2467 0.00105 0.073 616.5 1.6 3.101 0.003 0.153 0.46 0.17 0.81 3.09 0.22 1.76 439 176 13 31 0.51 486 19.370–19.380 m ORG- 7831/02-U-01, 26.955 0.076 0.2492 0.00186 0.070 767.1 2.3 3.745 0.004 0.121 0.23 0.81 0.73 2.79 0.22 1.57 439 178 14 30 0.61 487 19.380–19.400 m (A) ORG- 7831/02-U-01, 26.851 0.071 0.2783 0.00130 0.082 649.2 1.8 3.167 0.003 0.155 0.38 0.16 0.73 2.79 0.22 1.57 439 178 14 30 0.61 5.25 5.60 45.1 270 665 19.380–19.400 m (A) (rep)b ORG- 7831/02-U-01, 25.765 0.064 0.2554 0.00220 0.074 690.8 1.9 3.354 0.005 0.213 0.25 0.90 664 19.380–19.400 m (B) ORG- 7831/02-U-01, 43.083 0.103 0.4567 0.00141 0.134 640.8 1.6 3.268 0.003 0.148 0.14 0.20 1.06 4.33 0.29 2.17 436 200 13 33 0.29 e 445 20.430–20.440 m – 232 ORG- 7831/02-U-01, 41.589 0.098 0.4593 0.00207 0.136 606.2 1.5 3.111 0.003 0.154 0.15 0.20 0.93 5.61 0.24 2.04 439 275 12 22 0.2 446 20.440–20.450 me ORG- 7831/02-U-01, 42.893 0.102 0.5176 0.00119 0.158 540.2 1.4 2.831 0.002 0.146 0.14 0.17 1.03 5.12 0.29 2.08 438 246 14 27 0.52 447 20.450–20.460 me ORG- 7831/02-U-01, 43.300 0.120 0.4752 0.00230 0.141 612.2 1.8 3.149 0.003 0.145 0.14 0.19 0.93 4.99 0.38 2.06 440 242 18 25 0.29 9.61 12.8 78.0 83 448 20.460–20.480 m ORG- 7831/02-U-01, 89.149 0.215 0.9029 0.00299 0.261 680.0 1.7 3.417 0.003 0.138 0.07 0.10 1.67 9.81 0.43 3.37 438 291 13 51 0.63 442 20.620–20.625 m ORG- 7831/02-U-01, 92.321 0.221 0.8852 0.00451 0.251 731.8 1.8 3.623 0.003 0.145 0.07 0.10 1.59 10.73 0.46 3.48 437 308 13 48 0.56 443 20.625–20.640 m ORG- 7831/02-U-01, 77.523 0.184 0.7812 0.00337 0.225 684.7 1.7 3.436 0.003 0.153 0.08 0.12 1.34 9.20 0.50 3.16 437 291 16 45 0.56 444 20.640–20.650 m ORG- 7831/02-U-01, 59.576 0.194 0.5517 0.00291 0.155 766.9 2.6 3.759 0.004 0.153 0.04 0.09 1.34 7.60 0.51 2.68 436 284 19 40 0.44 421 20.650–20.665 m (A) ORG- 7831/02-U-01, 57.275 0.188 0.5198 0.00369 0.145 788.4 2.7 3.844 0.004 0.141 0.18 0.70 666 20.650–20.665 m (B) 227 (continued on next page) 228

ORG- 7831/02-U-01, 42.775 0.124 0.4289 0.00304 0.124 689.0 2.1 3.452 0.005 0.217 0.05 0.12 1.10 5.65 0.52 2.26 437 250 23 34 0.28 6.98 3.49 14.1 142 422 20.665–20.680 m ORG- 7831/02-U-01, 42.345 0.119 0.4050 0.00261 0.115 734.7 2.2 3.641 0.005 0.207 0.05 0.13 1.21 5.81 0.51 2.23 435 261 23 37 0.33 423 20.680–20.700 me ORG- 7831/02-U-01, 45.289 0.130 0.4331 0.00220 0.123 735.1 2.2 3.645 0.004 0.165 0.05 0.12 1.16 5.35 0.32 2.29 434 234 14 39 0.42 6.16 3.68 28.8 96 424 20.700–20.720 me ORG- 7831/02-U-01, 42.293 0.118 0.4363 0.00217 0.126 665.3 2.0 3.378 0.004 0.186 0.05 0.12 1.15 5.67 0.33 2.20 435 258 15 30 0.3 425 20.720–20.730 me ORG- 7831/02-U-01, 47.065 0.145 0.4825 0.00347 0.140 669.6 2.2 3.382 0.005 0.208 0.05 0.11 1.16 6.33 0.44 2.34 436 271 19 35 0.28 426 20.730–20.740 me

7831/02-U-02 Botneheia Formation 7831/02-U-02, 6.03 4.44 53.0 249 6.48 m ORG- 7831/02-U-02, 334.511 0.830 1.8658 0.01909 0.396 1678.9 4.4 7.359 0.006 0.136 0.05 0.26 2.87 36.55 0.53 7.51 431 487 7 281 159 645 10.18–10.20 m ORG- 7831/02-U-02, 74.967 0.186 1.3294 0.01609 0.438 340.6 1.3 2.069 0.009 0.469 0.13 0.20 2.21 35.44 0.57 6.15 434 576 9 258 12.3 222 (2014) 395 Palaeoecology Palaeoclimatology, Palaeogeography, / al. et Xu G. 646 10.20–10.22 m ORG- 7831/02-U-02, 221.170 0.623 1.5940 0.01228 0.400 1101.0 3.2 5.085 0.005 0.133 0.07 0.26 3.06 48.63 0.53 8.26 432 589 6 290 26.3 647 10.22–10.25 m ORG- 7831/02-U-02, 250.075 0.661 1.7398 0.01264 0.428 1161.2 3.2 5.313 0.004 0.126 0.07 0.28 2.99 45.95 0.53 8.18 435 562 6 411 70.5 648 10.25–10.28 m ORG- 7831/02-U-02, 445.552 1.148 2.5238 0.02230 0.542 1634.1 4.4 7.187 0.006 0.136 0.04 0.20 3.94 54.65 0.59 9.75 432 561 6 540 124 5.29 2.90 36.5 561 649 10.28–10.30 m ORG- 7831/02-U-02, 40.394 0.097 1.1138 0.00116 0.388 207.2 0.5 1.549 0.001 0.134 0.24 0.23 1.78 22.69 0.39 4.44 434 511 9 158 5.95 650 10.30–10.34 m ORG- 7831/02-U-02, 54.344 0.127 0.9016 0.00321 0.295 369.2 0.9 2.126 0.002 0.153 0.19 0.18 1.39 18.84 0.33 3.81 437 494 9 131 17.0 629 11.12–11.13 m ORG- 7831/02-U-02, 187.191 0.596 1.0375 0.03155 0.220 1691.1 5.6 7.373 0.010 0.131 0.06 0.16 1.70 21.46 0.34 4.32 436 496 8 155 55.1 630 11.17–11.18 m ORG- 7831/02-U-02, 208.537 0.686 1.1503 0.02085 0.243 1704.2 5.8 7.416 0.007 0.098 0.05 0.14 1.55 24.35 0.36 4.73 436 515 8 157 67.3 4.89 2.58 26.8 571 631 11.18–11.19 m ORG- 7831/02-U-02, 204.342 0.687 1.1980 0.01529 0.266 1529.5 5.3 6.726 0.006 0.092 0.05 0.13 1.68 26.74 0.40 4.84 435 552 8 177 38.0 632 11.20–11.21 m ORG- 7831/02-U-02, 380.108 2.278 1.8188 0.07718 0.336 2249.7 13.7 9.586 0.016 0.107 0.03 0.08 2.26 31.05 0.47 6.15 434 505 8 145 173 633 11.22–11.23 m ORG- 7831/02-U-02, 78.235 0.199 1.0287 0.00334 0.322 483.3 1.3 2.571 0.002 0.116 0.12 0.14 1.57 23.19 0.43 4.14 437 561 10 117 11.3 634 11.26–11.28 m ORG- 7831/02-U-02, 85.179 0.236 1.0953 0.00595 0.341 497.2 1.5 2.631 0.004 0.186 0.12 0.15 1.66 23.71 0.33 4.17 436 568 8 103 11.3 635 11.28–11.30 m 7831/02-U-02, 4.82 1.99 14.9 413 12.83 m 7831/02-U-02, 5.78 3.03 32.5 415

16.42 m – 232 ORG- 7831/02-U-02, 110.901 0.278 2.4825 0.00947 0.838 263.4 0.7 1.839 0.003 0.198 0.11 0.13 3.69 47.65 0.37 8.24 438 578 4 127 7.87 5.39 2.62 22.7 446 651 18.180–18.200 m ORG- 7831/02-U-02, 120.470 0.366 2.6161 0.00883 0.879 272.6 1.0 1.879 0.005 0.372 0.06 0.02 3.90 49.03 0.46 9.17 436 535 5 112 6.23 466 18.200–18.205 m ORG- 7831/02-U-02, 71.905 0.173 1.1610 0.00448 0.373 383.7 1.0 2.317 0.002 0.131 0.10 0.05 2.14 26.95 0.49 5.43 437 497 9 85 2.96 467 18.205–18.225 m ORG- 7831/02-U-02, 81.294 0.204 1.2198 0.00552 0.386 418.7 1.1 2.456 0.002 0.126 0.09 0.04 2.24 25.73 0.48 5.58 436 461 9 87 3.24 468 18.225–18.230 m ORG- 7831/02-U-02, 82.041 0.218 1.1418 0.00623 0.356 459.0 1.3 2.622 0.004 0.190 0.21 0.54 667 18.230–18.240 m ORG- 7831/02-U-02, 64.425 0.167 1.1111 0.00319 0.361 355.0 1.0 2.201 0.002 0.119 0.11 0.05 2.33 22.67 0.47 4.85 437 467 10 75 10.3 8.18 9.12 50.3 420 469 18.240–18.280 m ORG- 7831/02-U-02, 62.847 0.148 1.2237 0.00386 0.405 308.5 0.8 2.017 0.002 0.147 0.11 0.04 2.66 29.05 0.45 5.75 437 506 8 76 4.54 470 18.280–18.300 m ORG- 7831/02-U-02, 85.758 0.207 1.5818 0.00408 0.519 328.5 0.8 2.100 0.002 0.161 0.09 0.03 3.54 39.37 0.42 7.48 438 526 6 74 4.61 471 18.300–18.320 m G. Xu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 222–232 229

Table 2 Radiometric isochron ages and results after Bayesian data analysis.

Sample intervals Sample location Original isochron ages Results after Bayesian analysis

Age 2σ Age 2σ 2σ (Ma) (Ma) (Ma) (Ma) (Ma)a

7831/02-U-01 Kong Karls Land 232.0 11.0 231.8 +4.8 +4.8 19.130–19.400 m (this study) −10.1 −10.1 7831/02-U-01 236.5 3.5 236.2 +0.67 +0.99 20.460–20.680 m −3.00 −3.11 7831/02-U-02 236.6 0.39 236.6 +0.38 +0.83 10.18–10.34 m −0.39 −0.84 7831/02-U-02 237.6 0.59 237.5 +0.56 +0.92 11.12–11.30 m −0.59 −0.93 7831/02-U-02 238.4 1.47 238.4 +1.26 +1.46 18.18–18.32 m −0.90 −1.17 7323/07-U-09 Svalis Dome 239.3 2.6 239.5 +1.83 +1.97 94.07–94.26 m (Xu et al., 2009) −1.50 −1.70 Outcrop samples over 0.9-meter interval Svalbard 241.2 2.13 241.3 +2.00 +2.13 (Xu et al., 2009) −1.80 −1.97

a Including 0.31% of 187Re decay uncertainty from Smoliar et al. (1996) which is used to compare ages from other methods.

zircon ages and 40Ar/39Ar ages may result from slightly different sam- input; (2) a global decrease in continental weathering rates; or (3) an pling levels. The 242.1 Ma age of Mundil et al. (2010) derives from a increased flux of mantle derived Os caused by volcanic eruptions. sample collected one ammonoid zone (i.e., Nevadites secedensis Zone) A transient excursion to lower 187Os/188Os ratios delivered to the below the Anisian–Ladinian boundary, whereas the 239.3 Ma age of Barents shelf area may have accompanied the arrival of the pro-deltaic Xu et al. (2009) was obtained within the Parafrechites sublaqueatus Tschermakfjellet sediments (Mørk et al., 1999) and/or change of source Zone, which is equivalent to upper N. secedensis Zone in the Tethyan provenance (Glørstad-Clark et al., 2010). However, shale from the realm (Ogg and Ogg, 2008), the uppermost ammonoid zones in the Tschermakfjellet has 187Os/188Os ratios as high as or similar to shale Anisian. It is possible that N. secedensis Zone lasted 2 m.y. (Fig. 6). We from the uppermost Botneheia Formation (Fig. 4). The recorded decline argue that based on our Re–Os isochron ages the Anisian–Ladinian in 187Os/188Os ratios from Anisian to early Carnian is contemporaneous boundary is best placed between 239 and 241 Ma, in agreement within with decreasing 87Sr/86Sr ratios in Tethys (Korte et al., 2003) and in- uncertainty with the newly revised Triassic time scale (Gradstein et al., creasing δ13C values in both Boreal and Tethys realms (Payne et al., 2012). 2004; Katz et al., 2011). Thus, it most likely reflects a global process and the sharp decrease of 187Os/188Os ratios is probably not due to the regional variation. 4.3. Mid-Triassic decline in seawater 187Os/188Os and Wrangellian The decrease in 187Os/188Os could reflect a decrease in delivery of ra- volcanism diogenic Os from continental weathering to the . If there were no transient excursion to lower 187Os/188Os ratios in average riverine input The initial 187Os/188Os ratio of an isochron captures the 187Os/188Os to seawater, then the decreased ratio in 187Os/188Os represents a ~24% ratio of global seawater at the time of deposition. As Os is a conservative decrease in the continental weathering rate using present-day isotopic element with short residence time in seawater (10–40 ka), this ratios for the riverine and seawater components (Ravizza et al., 2001; ratio records short-term paleo-environmental variations (Peucker- Cohen, 2004). Yet the change in 187Os/188Os closely precedes the CPE, Ehrenbrink and Ravizza, 2000). The 187Os/188Os ratios in seawater result which is interpreted as a significant increase in terrestrial weathering from mixing of continental radiogenic Os (i.e., riverine input) with generated by a warm, humid climate with strong . Therefore, unradiogenic Os from hydrothermal fluids, primitive volcanic input, a dramatic decrease in continental weathering rate within nominally and cosmic dust. The seawater 187Os/188Os ratios sharply decrease from 1 m.y. is unlikely. 0.79 to 0.66 in 7-meter-thick shale (within nominally 1 m.y.) during Amajorflux of mantle Os into seawater can result from growth of a the late Ladinian (Fig. 7). The 0.13 unit (or 16%) decrease in seawater , such as the Wrangellian oceanic plateau that 187Os/188Os ratio is not trivial and three possible scenarios could explain formed during the late Ladinian and Carnian stages. A ~13% increase the decrease: (1) a regional decrease in the 187Os/188Os ratio of riverine in volcanic flux caused by the eruption of Wrangellian flood can

Notes to Table 1 Re and Os concentrations and isotopic compositions were measured by isotope dilution-negative thermal ionization mass spectrometry (ID-NTIMS) on a Triton® at the AIRIE Program, Colorado State University. Analytical procedures are previously published in Xu et al. (2009).The2σ uncertainties for Re and Os abundances, 187Re/188Os and 187Os/188Os isotope ratios, and error correlations (rho) are determined by propagation of uncertainties from Re and Os measurements, weighing, spikes, oxygen and blank corrections, and natural 187Re/185Re ratio of 0.59738 ± 0.00039 (2σ)(Gramlich et al., 1973). Trace metal abundances were measured by inductively coupled plasma atomic emission spectrometry (ICP-AES) at SGS Laboratories in Toronto using an aqua regia digestion following the procedure of Xu et al. (2012) with reproducibility better than 10%. Rock-Eval pyrolysis was performed using a Rock-Eval 6 instrument at Geolab Nor (Fugro) in Trondheim, Norway. Framboidal pyrite diameter distributions were measured on polished thin sections with a scanning electron microscope operated in backscattered electron mode at 15 keV. For polyframboids and pyrite concretions, individual framboids were measured separately whenever possible. aA and B are different pieces of shales broken from the same sampled interval and the stratigraphic order between A and B is not known. bRep is full procedural replicate on the same homogenized powder. c192Os is nonradiogenic isotope of Os that does not change with time. dThe amount of Re or Os in blank compared to Re or Os in sample. The blank was measured in each batch with 6 to 7 samples. The blank during this period has 187Os/188Os ratios between 0.12 and 0.48. eSamples are not included in the Isoplot regression in Fig. 4d. 230 G. Xu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 222–232

Fig. 6. Summary of Re–Os and U–Pb radiometric ages across the Anisian–Ladinian boundary, calibrated against ammonoid zones. U–Pb ages determined on zircons pre-treated by chemical abrasion are shown. Two Re–Os isochron ages determined by Kong Karls Land shales which are from ca. 1 m and ca. 2 m above the diagnostic late Ladinian ammonoid ex gr. Nathorstites sp. Juv. are shown in bold. Uncertainties for Re–Os ages include both propagated analytical error and decay constant uncertainty. Modified from http://engineering.purdue.edu/stratigraphy/charts/educational.html.Re–Os isochron ages for late Anisian shales from Svalbard and Svalis Dome are from Xu et al. (2009). explain the sharp drop in 187Os/188Os if other end-members are un- In the uppermost Ladinian Botneheia Formation, 187Os/188Os ratios changed. The Os concentration in the shale is controlled not by influx recovered from 0.66 to 0.73 (11% increase) in less than 1 m of section. to seawater, but by local factors, such as the amount and type of organic CO2 flux to the ocean–atmosphere system from Wrangellian eruptions carbon, sulfide concentration, and redox conditions near the sediment– may have induced warming which enhanced continental weathering. water interface. Major biotic changes, both on land and in the , Thus, the increasing 187Os/188Os ratios reflect re-balancing of seawater started prior to the most prominent negative C-isotope excursion Os ratios as increased continental weathering accompanied onset of (Turgeon and Creaser, 2008; Lindström et al., 2012). This is also consis- the CPE. The higher 187Os/188Os ratio was maintained throughout the tent with the proposal that increased submarine volcanism in the Late early Carnian (Fig. 7). Ladinian initiated global change, but the pronounced CPE and carbon The upper Ladinian Botneheia Formation shales (intervals 10.18– isotope excursion, which are not yet directly dated, did not begin until 10.34 m and 11.12–11.30 m) show a marked increase in 187Re/188Os early Carnian. ratios (average 187Re/188Os ratio of 1127) compared to lower Botneheia G. Xu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 222–232 231

0.90 0.7084 Acknowledgments Carnian Anisian 0.85 0.7082

Ladinian This study is part of Petromaks Project 180015/S30 funded by the Norwegian Research Council, Statoil ASA and Eni Norge AS under an 0.80 0.7080 agreement between the Geological Survey of Norway, the AIRIE Program at Colorado State University, and SINTEF Petroleum Research. Os 0.75 0.7078 Sr The NPD is thanked for providing access to the cores and permission 86 188 to publish. Wolfgang Weitschat is thanked for helpful discussions. The Sr /

Os / 0.70 0.7076 manuscript is improved by the comments from two anonymous re- 87

187 viewers and Editor F. Surlyk. onset of carbon 0.65 isotope excursion 0.7074 and start of CPE Appendix A. Supplementary data 0.7072 0.60 biostratigraphically-bracketed age for Wrangellia eruption Supplementary data to this article can be found online at http://dx. 0.55 0.7070 doi.org/10.1016/j.palaeo.2013.12.027. 220 230 240 250 Age (Ma) References Fig. 7. Temporal variations of 87Sr/86Sr and 187Os/188Os in seawater from late to 187 188 Triassic. The Os/ Os ratio from Svalis Dome is not shown as it reflects local variations Biasi, G.P., Weldon II, R., 1994. Quantitative refinement of calibrated 14C distributions. in Os input characteristic of a seawater-restricted basin (Xu et al., 2009). Also shown are Quat. Res. 41, 1–18. the biostratigraphically constrained age span for Wrangellian flood basalts (Greene Brack, P., Rieber, H., Mundil, R., Blendinger, W., Maurer, F., 2007. Geometry and chronol- et al., 2010 and references therein), and the estimated onset of the carbon isotope excur- ogy of growth and drowning of Middle Triassic carbonate platforms (Cernera and sion and start of Carnian Pluvial Event (Dal Corso et al., 2012). Bivera/Clapsavon) in the Southern (northern Italy). Swiss J. Geosci. 100, 87 86 – Sr/ Sr curve data are taken from Korte et al. (2003) and is re-calibrated based on the 327 348. revised Triassic time scale of Gradstein et al. (2012). 187Os/188Os curve data are from this Brühwiler, T., Hochuli, P., Mundil, R., Schatz, W., Brack, P., 2007. Bio- and chronostratigraphy fl study (in solid squares) and Xu et al. (2009) and Georgiev et al. (2011) (in open squares). of the Middle Triassic Rei ing Formation of the westernmost Northern Calcareous Alps. Swiss J. Geosci. 100, 443–455. Cohen, A.S., 2004. The rhenium–osmium isotope system: applications to geochronological and palaeoenvironmental problems. J. Geol. Soc. 161, 729–734. Cohen, A.S., Coe, A.L., 2002. New geochemical evidence for the onset of volcanism in the Formation shales (18.18–18.32 m, average 187Re/188Os ratio of 349) Central Atlantic magmatic province and environmental change at the Triassic–Juras- – (Fig. 3). Extraordinarily high 187Re/188Os ratios in latest Permian black sic boundary. Geology 30, 267 270. Cohen, A.S., Coe, A.L., Harding, S.M., Schwark, L., 2004. Osmium isotope evidence for the shales are attributed to rising seawater temperature and acidity regulation of atmospheric CO2 by continental weathering. Geology 32, 157–160. (Georgiev et al., 2011). Though less extreme than the latest Permian Dagys, A.S., Weitschat, W., 1993. Correlation of the Boreal Triassic. Mitt. Geol.-Palaontol. – increase, increased 187Re/188Os ratios in the Ladinian may similarly re- Inst. Univ. Hamburg 75, 249 256. fl Dal Corso, J., Mietto, P., Newton, R.J., Pancost, R.D., Preto, N., Roghi, G., Wignall, P.B., 2012. ect rising seawater temperature and acidity induced by Wrangellian Discovery of a major negative δ13C spike in the Carnian (Late Triassic) linked to the volcanism. eruption of Wrangellia flood basalts. Geology 40, 79–82. Wrangellian flood basalts, initiated by a mantle plume (Richards Furin, S., Preto, N., Rigo, M., Roghi, G., Gianolla, P., Crowley, J.L., Bowring, S.A., 2006. High- precision U–Pb zircon age from the Triassic of Italy: implications for the Triassic time et al., 1991), erupted as early as middle Ladinian, based on underlying scale and the Carnian origin of calcareous nannoplankton and dinosaurs. Geology 34, fossils (Jones et al., 1977; Greene et al., 2010). The oldest radiometric 1009–1012. age published for Wrangellian volcanism is an average of 3 discordant Furrer, H., Schaltegger, U., Ovtcharova, M., Meister, P., 2008. U–Pb zircon age of 207 206 volcaniclastic layers in Middle Triassic platform carbonates of the Austroalpine Pb/ Pb dates of 232.2 ± 1 Ma from multigrain zircon fractions Silvretta nappe (Switzerland). Swiss J. Geosci. 101, 595–603. (Greene et al., 2010 and references therein), but radiometric age con- Georgiev, S., Stein, H.J., Hannah, J.L., Bingen, B., Weiss, H.M., Piasecki, S., 2011. Hot acidic straints are few for this broadly dispersed terrain. The abrupt negative Late Permian seas stifle life in record time. Earth Planet. Sci. Lett. 310, 389–400. 187 188 Glørstad-Clark, E., Faleide, J.I., Lundschien, B.A., Nystuen, J.P., 2010. Triassic seismic se- excursion in seawater Os/ Os and subsequent warming associated quence stratigraphy and paleogeography of the western Barents Sea area. Mar. Pet. with the CPE is consistent with the hypothesis that a major pulse of Geol. 27, 1448–1475. Wrangellian volcanism in the late Ladinian is the direct cause of the Gradstein, F., Ogg, J., Smith, A., 2004. A Geologic Time Scale 2004. Cambridge University Carnian biotic crisis. Press, Cambridge, UK 589 pp. Gradstein, F.M., Ogg, J.G., Schmitz, M.D., Ogg, G. (Eds.), 2012. The Geologic Time Scale 2012. Elsevier, p. 1144. 5. Conclusions Gramlich, J.W., Murphy, T.J., Garner, E.L., Shields, W.R., 1973. Absolute isotopic abundance ratio and atomic weight of a reference sample of rhenium. J. Res. Natl. Bur. Stand. Sect. A Phys. Chem. A 77, 691–698. Re–Os isochron ages from the Barents Sea place the Ladinian– Greene, A.R., Scoates, J.S., Weis, D., Katvala, E.C., Israel, S., Nixon, G.T., 2010. The architec- Carnian boundary at 236.6–237.5 Ma and the Anisian–Ladinian bound- ture of oceanic plateaus revealed by the volcanic stratigraphy of the accreted Wrangellia oceanic plateau. Geosphere 6, 47–73. ary at 239.3 Ma, respectively. These results are consistent within uncer- Hannah, J.L., Bekker, A., Stein, H.J., Markey, R.J., Holland, H.D., 2004. Primitive Os and tainties with the recently revised Triassic Time Scale (Gradstein et al., 2316 Ma age for marine shale: implications for Paleoproterozoic glacial events and 2012), and provide precise and consistent age results for a period in the rise of atmospheric oxygen. Earth Planet. Sci. Lett. 225, 43–52. Hornung, T., Brandner, R., Krystyn, L., Joachimski, M.M., Keim, L., 2007. Multistratigraphic Earth history with few geochronologic controls. Results confirm that constraints on the NW Tethyan “Carnian Crisis”.In:Lucas,S.G.,Spielmann,J.A.(Eds.), the Ladinian Stage was brief (~4 m.y.) (Gradstein et al., 2012), in con- The Global Triassic. New Mexico Museum of Natural History and Science Bulletin, trast to the previously assigned interval of ~12 m.y. (Gradstein et al., pp. 321–330. 2004). Hounslow, M.W., Nawrocki, J., 2008. Palaeomagnetism and magnetostratigraphy of the Permian and Triassic of Spitsbergen: a review of progress and challenges. Polar Res. 187 188 The Os/ Os ratio for seawater decreased markedly during 27, 502–522. the late Ladinian, synchronously with well-documented temporal Jones, D.L., Silberling, N.J., Hillhouse, J., 1977. Wrangellia — a displaced terrane in north- – variations in 87Sr/86Sr ratios. The 16% decrease in 187Os/188Os ratio with- western North America. Can. J. Earth Sci. 14, 2565 2577. Katz, M., Mørk, A., Milligan, A., 2011. Triassic Carbon Isotope Stratigraphy — Correlations in a nominal 1 m.y. is likely caused by a major pulse of Wrangellian vol- Between Panthalassic and Tethyan sections. AAPG Annual Convention & Exhibition, canism in the late Ladinian. The timing of these events also correlates Houston, TX. with onset of the Carnian Pluvial Event, the carbon isotope excursion, Kendall, B., Reinhard, C.T., Lyons, T.W., Kaufman, A.J., Poulton, S.W., Anbar, A.D., 2010. Per- vasive oxygenation along late Archaean ocean margins. Nat. Geosci. 3, 647–652. and the accompanying biotic crisis in the Carnian, corroborating the Korte, C., Kozur, H.W., Bruckschen, P., Veizer, J., 2003. Strontium isotope evolution of Late role of Wrangellian volcanism. Permian and Triassic seawater. Geochim. Cosmochim. Acta 67, 47–62. 232 G. Xu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 222–232

Krajewski, K.P., 2008. The Botneheia Formation (Middle Triassic) in Edgeøya and Ogg, J.G., Ogg, G., 2008. Permian/Triassic (236–273 Ma time-slice). https:// Barentsøya, Svalbard: lithostratigraphy, facies, phosphogenesis, paleoenvironment. engineering.purdue.edu/Stratigraphy/charts/Timeslices/8_Lower_Triassic.pdf Polar Res. 29, 319–364. (Last visited on 10/10/2013). Kuiper, K.F., Deino, A., Hilgen, F.J., Krijgsman, W., Renne, P.R., Wijbrans, J.R., 2008. Ogg, J.G., Ogg, G., Gradstein, F.M., 2008. The Concise Geological Time Scale. Cambridge Synchronizing rock clocks of earth history. Science 320, 500–504. University Press 177. Lindström, S., van de Schootbrugge, B., Dybkjær, K., Pedersen, G.K., Fiebig, J., Nielsen, L.H., Pálfy, J., Parrish, R.R., David, K., Vörös, A., 2003. Mid-Triassic integrated U–Pb geochronol- Richoz, S., 2012. No causal link between terrestrial change and methane ogy and ammonoid biochronology from the Balaton Highland (Hungary). J. Geol. Soc. release during the end-Triassic mass extinction. Geology 40, 531–534. 160, 271–284. Lock, B.E., Pickton, C.A.G., Smith, D.G., Batten, D.J., Harland, W.B., 1978. The geology of Payne, J.L., Lehrmann, D.J., Wei, J.Y., Orchard, M.J., Schrag, D.P., Knoll, A.H., 2004. Large per- Edgeøya and Barentsøya, Svalbard. Nor. Polarinst. Skr. 168, 1–64. turbations of the carbon cycle during recovery from the end-Permian extinction. Sci- Ludwig, K., 2003. Isoplot/Ex, Version 3.6.3: A Geochronological Toolkit for Microsoft ence 305, 506–509. Excel. Special Publication , 4. Berkeley Geochronology Center, Berkeley, Califor- Peucker-Ehrenbrink, B., Ravizza, G., 2000. The marine osmium isotope record. Terra Nova nia 74. 12, 205–219. Mietto, P., Andreetta, R., Broglio, L.C., Buratti, N., Cirilli, S., De Zanche, V., Furin, S., Gianolla, Preto, N., Kustatscher, E., Wignall, P.B., 2010. Triassic state of the art and perspec- P., Manfrin, S., Muttoni, G., Neri, C., Nicora, A., Posenato, R., Preto, N., Rigo, M., Roghi, tives. Palaeogeogr. Palaeoclimatol. Palaeoecol. 290, 1–10. G., Spötl, C., 2007. A candidate of the global boundary stratotype section and point for Raup, D.M., Sepkoski, J.J., 1982. Mass extinctions in the marine fossil record. Science 215, the base of the Carnian Stage (Upper Triassic): GSSP at the base of the canadensis 1501–1503. subzone (FAD of Daxatina) in the Prati di Stuores/Stuores Wiesen section (Southern Ravizza, G., Norris, R.N., Blusztajn, J., Aubry, M.P., 2001. An osmium isotope excursion as- Alps, NE Italy). Albertiana 36, 78. sociated with the Late thermal maximum: evidence of intensified chemical Mietto, P., Manfrin, S., Preto, N., Gianolla, P., 2008. Selected ammonoid fauna from Prati di weathering. Paleoceanography 16, 155–163. Stuores/Stuores Wiesen and related sections across the Ladinian–Carnian boundary Richards,M.A.,Jones,D.L.,Duncan,R.A.,Depaolo,D.J.,1991.A mantle plume initiation model (southern alps, Italy). Riv. Ital. Paleontol. Stratigr. 114, 377–429. for the Wrangellia flood basalt and other oceanic plateaus. Science 254, 263–267. Mietto, P., Manfrin, S., Preto, N., Rigo, M., Roghi, G., Furin, S., Gianolla, P., Posenato, R., Riis, F., Lundschien, B.A., Høy, T., Mørk, A., Mørk, M.B.E., 2008. Evolution of the Triassic Muttoni, G., Nicora, A., Buratti, N., Cirilli, S., Spötl, C., Ramezani, J., Bowring, S.A., shelf in the northern Barents Sea region. Polar Res. 27, 318–338. 2012. The Global Boundary Stratotype Section and Point (GSSP) of the Carnian Simms, M.J., Ruffell, A.H., 1989. Synchroneity of climatic change and extinctions in the Stage (Late Triassic) at Prati di Stuores/Stuores Wiesen Section (Southern Alps, NE Late Triassic. Geology 17, 265–268. Italy). Episodes 35, 414–430. Smoliar, M.I., Walker, R.J., Morgan, J.W., 1996. Re–Os ages of group IIA, IIIA, IVA, and IVB Mørk,A.,Dallmann,W.K.,Dypvik,H.,Johannessen,E.P.,Larssen,G.B.,Nagy,J., iron meteorites. Science 271, 1099–1102. Nøttvedt, A., Olaussen, S., Pčelina, T.M., Worsley, D., 1999. lithostratig- Turgeon, S.C., Creaser, R.A., 2008. oceanic anoxic event 2 triggered by a mas- raphy. In: Dallmann, W.K. (Ed.), Lithostratigraphic Lexicon of Svalbard: Review sive magmatic episode. Nature 454, 323–326. and Recommendations for Nomenclature Use. Norwegian Polar Institute, Vigran, J.O., Mangerud, G., Mørk, A., Worsley, D., Hochuli, P.A., 2013. Palynology and geol- Tromsø, pp. 127–305. ogy of the Triassic succession of Svalbard and the Barents sea. Geological Survey of Mørk, A., Knarud, R., Worsley, D., 1982. Depositional and diagenetic environments of the Norway Special Publication 14. (in press). Triassic and Lower succession of Svalbard. In: Embry, A.F., Balkwill, H.R. Weitschat, W., Dagys, A.S., 1989. Triassic biostratigraphy of Svalbard and a comparison (Eds.), Arctic Geology and Geophysics. Canadian Society of Petroleum Geologists, with NE-Siberia. Mitt. Geol.-Palaontol. Inst. Univ. Hamburg 68, 179–213. pp. 371–398. Wilkin, R.T., Barnes, H.L., Brantley, S.L., 1996. The size distribution of framboidal pyrite in modern Mørk, A., Vigran, J.O., Hochuli, P.A., 1990. Geology and palynology of the Triassic succes- sediments: an indicator of redox conditions. Geochim. Cosmochim. Acta 60, 3897–3912. sion of Bjørnøya. Polar Res. 8, 141–163. Xu, G., Hannah, J.L., Bingen, B., Georgiev, S., Stein, H.J., 2012. Digestion methods for trace Mundil, R., Brack, P., Meier, M., Rieber, H., Oberli, F., 1996. High resolution U–Pb dating of element measurements in shales: paleoredox proxies examined. Chem. Geol. Middle Triassic volcaniclastics: time-scale calibration and verification of tuning pa- 324–325, 132–147. rameters for carbonate sedimentation. Earth Planet. Sci. Lett. 141, 137–151. Xu, G., Hannah, J.L., Stein, H.J., Bingen, B., Yang, G., Zimmerman, A., Weitschat, W., Mørk, Mundil, R., Palfy, J., Renne, P.R., Brack, P., 2010. The Triassic timescale: new constraints A., Weiss, H.M., 2009. Re–Os geochronology of black shales in Arctic regions: evaluat- and a review of geochronological data. Geological Society, London, Special Publica- ing the Anisian–Ladinian boundary and global faunal correlations. Earth Planet. Sci. tions 334, 41–60. Lett.288,581–587.

Supplementary Information For:

Cause of Upper Triassic Climate Crisis Revealed by Re-Os

Geochemistry of Boreal Black Shales

By Xu et al. 2014

1. Palynostratigraphy and Ladinian – Carnian boundary

The palynological distribution in cores 7831/02-U-02 and 7831/02-U-01 is listed in Table

S1. Vigran et al. (in press) assigned two composite assemblage zones.

The Echinitosporites iliacoides Composite Assemblage Zone is defined by E. iliacoides. This pollen is usually not seen in Svalbard, but is common in the Barents Sea and arctic (Mørk et al., 1992). Pollen like Schizaeoisporites worsleyi,

Staurosaccites quadrifidus, Triadispora verrucata and Ovalipollis pseudoalatus all have their FADs (First Appearance Datum) in the lower part of this zone. Protodiploxypinus ornatus appears consistently and monosaccoid pollen (Cordaitina gunyalensis) is a characteristic feature.

This assemblage is present in the upper part of Botneheia Formation at Svalbard, lowest in Skuld Formation at Bjørnøya, and in the lower part of Snadd Formation of the Barents

Sea. According to the ammonite constraints at these places, the lower part of the core

7831/02-U-02, from 18.79 to 13.88 m, is Ladinian.

1 The Aulisporites astigmosus Composite Assemblage Zone is defined by appearance and abundance of Aulisporites astigmosus. This zone has FADs for Camerosporites secatus and Enzonalasporites vigens and at higher stratigraphic levels, Ricciisporites tuberculatus, Camarozonosporites rudis and Uvaesporites argentaeformis. There are intervals with abundant A. astigmosus and Leschikisporis aduncus within the zone. The pollen Illinites chitonoides, Angustisulcites klausii, E. iliacoides and Triadispora verrucata have their LADs (Last Appearance Datum) within the zone.

At Bjørnøya, the layers with Daxatina canadensis in Skuld Formation are overlain by layers with E. iliacoides, which shows that the pollen grain ranges into the Early Carnian

(Vigran et al. in press). E. iliacoides reached maximum abundance during the latest

Ladinian / Carnian transition in the southeastern Barents Shelf (Mørk et al., 1992). The

Aulisporites astigmosus Composite Assemblage Zone in the upper part of core 7831/02-

U-02 (6.16 to 11.43 m) and lower part core 7831/02-U-01 is thus inferred to be present in upper Ladinian to lower Carnian beds. Therefore, the Ladinian – Carnian boundary according to palynology may be within the upper part of drillcore 7831/02-U-02, 6.16 –

11.43 m (Fig. 2) up to 19.66 m of core 7831/02-U-01.

2 Table S1. Selected taxa from drill cores east of Kong Karls Land Locality 7831/2- U-2 U-1 Formation Botneheia Snadd Age Ladinian l. Lad / e. Carn L/C e. C 18.79 17.87 14.89 14.32 13.88 11.43 10.28 26.61 24.36 19.66 14.83 11.85 9.85 7.70 6.83 6.21 6.16 Taxa / Depth below sea level

Bisaccate pollen (abundance) DDDDDDDDDDDDDDDDD Deltoidospora minor xC xxCCAxDAA ACAAA Granasporites magnus xxx xxCxxxCx CCxxA Illinites chitonoides # CxxxCxCxCxxx CxCCC Podosporites amicus xC xC xxCAAC ACAAC * Schizaeoisporites worsleyi x x x x x x x x x x x * Staurosaccites quadrifidus xCCxxxxAACCA CACxx Striatoabieites balmei AxACAAAAACCC ACADx Striatoabieites multistriatus x xx xxxxx x xxxxx * Angustisulcites klausii x x C x C x C C x * Angustisulcites spp. x C x x x x Protodiploxypinus spp. xCCx ACxxx x x A Protodiploxypinus macroverrucosus x x x x x Triadispora spp. # x A x C C A C A C x Kyrtomisporis laevigatus x x x x x x x x x x * Leschikisporis aduncus x xCxxCxxC xCCAC Pinuspollenites spp. x x x x x x x x x Lunatisporites noviaulensis xx Cx CxCxxCAx Lunatisporites spp. x x x x x Protodiploxypinus decus xxxxxxxx xx xx Protodiploxypinus doubingeri x x x x Protodiploxypinus ornatus xxxx xxxxx CADxx Striatoabieites aytugii x x x x C A A x x x * Triadispora verrucata # x C C A C x A x C x Baculatisporites/Osmundacidites x x xxxxxx xxCC Anapiculatisporites spp. x x x x x x x x x * Heliosaccus dimorphus x x x x x x x x * Aulisporites astigmosus x x C C C x C A D D Thomsonisporites toralis x x Gordonispora fossulata x x x x x Infernopollenites spp. x x x Protodiploxypinus minor x x * Echinitosporites iliacoides # x x x x * Ovalipollis pseudoalatus x x x x x x x * Camarozonosporites laevigatus x x x * Camarozonosporites rudis x x x x x x x x x Conbaculatisporites spp. xxCxCCx xxxC Todisporites minor x C x x Dictyophyllidites spp. x x x x x x C x C Semiretisporis spp. x x x x x x Verrucosisporites spp. x x C Kraeuselisporites spp. x x x Punctatosporites walcomii x x x Chasmatosporites spp. C x x x x x C C A Calamospora tener x x Conbaculatisporites sp.1 CC Nevesisporites limatulus x x x x x x Kraeuselisporites cooksonae x x x A Retusotriletes mesozoicus x x x x Uvaesporites spp. x x x x x Deltoidaspora toralis x x x x Aratrisporites spp. x x x x x x * Camerosporites secatus x * Retisulcites perforatus C Succinctisporites grandior LEGEND x * Corollina spp. D >16 dominant x Cingulate zonate spores A 7-15 abundant x Stereisporites perforatus C 3-6 common x Decisporis reticulatus x 1-2 present x & varia Leiosphere DDDDDDxDxDDD AADAA Crassosphaera spp. ACAAxCxx xxC DDCC Leiosphere (large) x x x C x x x Micrhystridium spp. A ACxxxxxxxx Dxxx Pterospermella spp. x x x x x x x * Spherical body (large) CCD * Botryococcus spp. x C C x x x Tasmanites spp. C x x Acanthomorph acritarch D x * Veryhachium spp. C x C C x Cymatiosphaera spp. x x x C Composite Assemblage Zone E. Iliacoides Aulisporites astigmosus * Denotes a stratigraphically important taxon with first appearance datum (FAD) in the late Ladinian # Detnotes taxon with LAD in the Carnian, missing representation in Core 7831/2-U-1, but present at higher levels 3 2. Bayesian data analysis

To tighten the uncertainties for our isochron ages, we take advantage of prior knowledge

that stratigraphically higher samples are younger than stratigraphically lower samples. By

imposing stratigraphic constraints on the radiometric ages using Bayesian statistics, the

uncertainties can be quantitatively limited (Biasi and Weldon II, 1994).

Consider a set of k+1 samples with ages of n0, n1, …, nk calculated from Re-Os isochrons, where with n0 > n1 > n2 > … > nk. If the original probability density function for the age n isP() t , then the new constrained probability density function at t = t 0 n0 0 given n0 > n1 is:

tP |( n older than ∝ > ntPtPn )()() n0 0 0 1 0 nn 1 100 (1) where

t ∝> 0 )()( dttPntP n1 10 ∫ n1 0 (2)

The proportionality symbol in equation (1) is required because the total probability must sum to 1; hence there is a proportionality constant associated with each equation.

Given another sample n2 farther up section, such that n0 > n1 > n2, then the probability density function at t = t0 is:

()()() older than and older than |( ntP older than and nn older than ∝ ()()() >> ntPntPtPn ) n0 00 1 1 2 n0 n1 n2 20100 (3)

The above equations (1) to (3) are based on stratigraphically up-section constraints; that is, stratigraphically lower samples are older than stratigraphically higher samples. We can construct similar equations based on down-section constraints; that is, stratigraphically higher samples are younger than stratigraphically lower samples. Consider a set of j+1 samples with ages of n0, n-1, …, n-j, where with n0 < n-1 < n-2 < … < n-j.

4 For n-1 that is older than n0, we simply flip the greater-than sign in equation (1):

younger th |( ntP younger th − ∝ ()()an tPtPn < n− ) n0 0 0 1 n0 100

In practice, these constraints lead to multiple nested integrals. If there are j samples that are progressively older (i.e., stratigraphically lower) than sample set n0 , and k samples that are progressively younger than sample set n0, the probability for sample n0 at time t=t0 is:

1 tP all|( )sconstraint = tP )( × n0 0 C n0 0

t0  t0  t0  t0   ()()( )  )( dttPdttPdttPdttP  × (up-section constraints) ∫n1 ∫n2 ∫n3 L ∫ nk 0  0  0  0 

∞  ∞  ∞  ∞   )()()(  )( dttPdttPdttPdttP   (down-section constraints) ∫n−1 ∫n−2  ∫n−3 L ∫ n− j  t0 t 0 t 0  t0 

Here, C is a normalization constant that is adjusted so that the total probability sums to 1.

New age uncertainties for our data are calculated using the above equations and listed in

Table 2.

3. Supplementary references

Mørk, A., Vigran, J.O., Korchinskaya, M.V., Pchelina, T.M., Fefilova, L.A., Vavilov, M.N., Weitschat, W., 1992. Triassicrocks in Svalbard, the Arctic Soviet islands and the Barents Shelf: bearing on their correlations. In: Vorren, T.O. et al. (Eds.), Arctic Geology and Petroleum Potential, NPF Special Publication. Elsevier Sci. Publ., pp. 457-479.

5 4. Supplementary figures

Fig. S1. Ammonoid ex gr. Nathorstites sp. Juv. from 7831/02-U-02, 12.27 - 12.32 m.

6 4.0 Tschermakfjellet Formation off Kong Karls Land 7831/02-U-01, 19.13-19.40 m 3.6 11 samples from 27 cm interval

s 3.2

O

8

8

1

/

s 2.8

O

7

8

1 2.4

Model 3 Age = 232 ± 11 Ma 2.0 Initial 187Os/188Os = 0.72 ± 0.11 MSWD=94,n = 12 1.6 250 350 450 550 650 750 850 187 188 Re/ Os

Fig. S2. Re-Os isochron for black shales from top Tschermakfjellet Formation,

constructed using Isoplot 3.60 (Ludwig, 2003) and a 187Re decay constant of 1.666 x

10-11 (Smoliar et al., 1996). Re-Os ages are reported with 2σ uncertainties.

7