Marine osmium isotope record across the Triassic-Jurassic boundary from a Pacific pelagic site

Marine osmium isotope record across the Triassic-Jurassic boundary from a Pacific pelagic site Junichiro Kuroda1,2*, Rie S. Hori3, Katsuhiko Suzuki1, Darren R. Gröcke4, and Naohiko Ohkouchi1 1

Japan Agency for Marine-Earth Science and Technology, Yokosuka 237-0061, Japan National Oceanography Centre, Southampton, University of Southampton, Southampton SO14 3ZH, UK 3 Graduate School of Science, Ehime University, Matsuyama 790-8577, Japan 4 Department of Earth Sciences, Durham University, Durham DH1 3LE, UK 2

ABSTRACT The Triassic-Jurassic (T-J) boundary ca. 200 Ma represents one of the major mass extinction events of the Phanerozoic; however, the cause of this event remains controversial because of a paucity of geological evidence. In this study we present an isotopic record of osmium extracted from a bedded chert succession across the T-J boundary in the Kurusu section of Japan, deposited within a PaleoPacific (Panthalassa) deep basin. The data show a gradual decrease in seawater 187Os/188Os values during the Rhaetian, followed by a sharp increase in the latest Rhaetian, and a subsequent stable phase across the T-J boundary. The decreasing trend of 187Os/188Os values during the Rhaetian indicates a gradual increase in the relative supply rate of unradiogenic Os from the mantle associated with emplacement of the Central Atlantic Magmatic Province. The subsequent shift toward radiogenic values reflects an increased supply of radiogenic Os due to enhanced continental weathering. This interval marks more negative isotopic values of organic carbon, the onset of radiolarian faunal turnover, and conodont extinctions, indicating that the rapid increase in continental weathering rate was closely linked to the perturbation of the carbon cycle and the T-J biotic crisis. INTRODUCTION Much research has focused on the relationship between massive igneous eruptions and major environmental perturbations, including those associated with mass extinctions (e.g., Courtillot et al., 1996; Wignall, 2001). The Triassic-Jurassic (T-J) boundary ca. 200 Ma (Pálfy et al., 2000; Ogg et al., 2008) marks one of the five largest mass extinction events in the Phanerozoic, when many marine and terrestrial species became extinct (Raup and Sepkowski, 1982). It is also characterized by the occurrence of extensive magmatic activity associated with the breakup of Pangaea and the initial stages of rifting in the Central Atlantic Magmatic Province (CAMP) (e.g., McHone, 2000; Nomade et al., 2007). These magmatic activities have been implicated as a possible forcing mechanism for the climatic and biotic changes recorded at the T-J boundary (Hesselbo et al., 2002; Marzoli et al., 2004; McElwain et al., 2007; van de Schootbrugge et al., 2009; Deenen et al., 2010). However, the evidence in this regard remains circumstantial because of the difficulties encountered in correlating the timing of CAMP volcanism with the environmental and biotic events, and in estimating the environmental impact of this type of igneous activity. Seawater 187Os/188Os values reflect the balance between average input from continental crust (187Os/188Os = ~1.3), and mantle and extraterrestrial inputs (~0.13) to the global ocean (e.g., Shirey and Walker, 1998; PeuckerEhrenbrink and Ravizza, 2000). Based on the idea that the emplacement of young, mantle-derived basaltic provinces releases large amounts of unradiogenic Os, the Os isotopic record in marine sediments may provide evidence of the occurrence of flood basalt eruptions (e.g., Cohen and Coe, 2002; Ravizza and Peucker-Ehrenbrink, 2003; Turgeon and Creaser, 2008; Tejada et al., 2009). The relatively short residence time of seawater *E-mail: [email protected].

Os means that whole-ocean shifts in isotopic composition may occur on the order of 10 k.y.; consequently, the marine Os isotope record serves as a clear stratigraphic marker of the initiation of major flood basalt volcanism. Cohen and Coe (2002, 2007) reported the Os isotopic record across the T-J boundary in southern England; however, no data have been reported from the Paleo-Pacific (Panthalassa) pelagic basin, which covered approximately half of the Earth’s surface. This study investigates the link between the CAMP eruptive event and the T-J biotic event by assessing the changing trends of the Os isotope composition of seawater, as recorded in pelagic deep-sea sediment (bedded chert) of the Panthalassa (Fig. 1). SAMPLES AND METHODS Samples were collected from the Kurusu section, Inuyama area, Japan (Fig. 1). The Kurusu section is composed of greenish-gray chert and thin intercalations of shale, representing a continuous sequence across the T-J boundary (Hori et al., 2007). Chert is mostly composed of hydrogenous material (i.e., biogenic silica) with a very minor terrigenous fraction. Hori et al. (2000) demonstrated that redistribution of elements during the formation of bedded chert is insignificant for the stratigraphic interval investigated in this study. They also indicated that chert is well suited as an archive of elements, because of its strongly consolidated, impermeable nature. A detailed description of the biostratigraphy in this section is given in Appendix DR1 (see the GSA Data Repository1). The highest occurrences of Rhaetian conodonts Misikella hernsteini and M. posthernstaini are at ~0.05 m and 0.28 m, respectively. No conodont fossils were obtained above 0.30 m level. Pantanellium tanuense, which is one of the characteristic radiolarians of the Hettangian (Carter et al., 1998), occurs at 0.28 m level. The co-occurrence of Bipedis elizabethae and B. hannai in the horizon at 0.48 m characterizes the lower Hettangian, based on radiolarian data from Queen Charlotte Islands, Canada, documented by Carter et al. (1998). Accordingly, the T-J boundary is placed within the interval between 0.28 and 0.48 m (referred to herein as the T-J boundary zone). The mean sedimentation rates during the Rhaetian and Hettangian were ~0.7 and 0.4 m m.y.–1, respectively (Fig. 1). Each sample was crushed to coarse fragments, from which we handpicked fresh fragments without veins or nodules, and prewashed before pulverizing. After spiking, Re and Os in the hydrogenous fraction of chert were extracted by inverse aqua regia digestion. Abundances of Re and Os and isotopic compositions of Os were analyzed by negative thermal ionization mass spectrometry. Initial Os isotopic compositions (187Os/188Osi) were calculated for the time of deposition based on the measured 187Os/188Os and 187Re/188Os values, the age of sediment (200 Ma), and the 187Re decay constant. Detailed analytical methods are given in Appendix DR1. 1 GSA Data Repository item 2010298, Appendix DR1 (analytical method and biostratigraphy), Figure DR1 (data around the T-J boundary), and Table DR1 (data), is available online at www.geosociety.org/pubs/ft2010.htm, or on request from [email protected] or Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301, USA.

© 2010 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or [email protected]. GEOLOGY, December 2010 Geology, December 2010; v. 38; no. 12; p. 1095–1098; doi: 10.1130/G31223.1; 1 figure; Data Repository item 2010298.

1095

B

Kurusu section, Inuyama Area, Japan

Stable phase Stable phase

-2

-3 205 0

5

10

Os (pg g-1)

15 0

100

200

300

Re (pg g-1)

0.2 187

0.4

0.6

Os/188Os

0.8 -28 -26 -24 -22

δ13Corg (‰)

δ13Corg (‰) -30

-28

-26

-24

MNE

200 Rhaetian

~0.7 m/m.y.

-1

CAMP volcanism

1st rise

KU+6U KU-4

Sinemurian Hettangian

KU+23

195

2nd rise

PR INE

TRIASSIC

Decrease

St. Audrie’s Bay, England

Age

JURASSIC

C

2

1

TRIASSIC

Rhaetian Norian

Haeckelicyrtium breviora

Canoptum triassicum

Betraccium deweveri

200 Ma Ma

~0.4 m/m.y.

Sinemurian

JURASSIC

Hettangian

Parahsuum simplum Pantanellium tanuense

3

0

Ku

CAMP

4

G. tozeri

En

Initial ratio (200 Ma) Measured value

m 5

Gradual decrease

Radiolarian Age zones

Norian

A

0

0.2 0.4 0.6 0.8 187

Os/188Os

Figure 1. A: Depth profiles of Re and Os concentrations, and isotopic compositions of Os (187Os/188Os) and total organic carbon (δ13Corg) in Triassic-Jurassic (T-J) bedded chert succession from Kurusu, Japan. Filled squares and open gray squares in the Os isotopic profile are agecorrected isotopic ratios (187Os/188Osi, t = 200 Ma) and measured isotopic ratios, respectively. Horizontal gray interval indicates T-J boundary zone. G. tozeri—Globolaxtorum tozeri subzone. B: Late Triassic paleogeographic map with locations of Kurusu (Ku; Hori et al., 2007), St. Audrie’s Bay section (En), and Central Atlantic Magmatic Province (CAMP) after Hesselbo et al. (2002). C: Depth profiles of 187Os/188Os (Cohen and Coe, 2002, 2007) and δ13Corg (Hesselbo et al., 2002) from St. Audrie’s Bay section. INE—initial negative excursion of δ13Corg, PR— positive recovery, MNE—main negative excursion. Broken line connecting panels A and C indicates boundary between INE and PR.

RESULTS AND DISCUSSION Stable isotopic ratios of total organic carbon (δ13Corg) show relatively negative values of –27‰ to –26‰ around the top of the Rhaetian interval (0.1–0.3 m), followed by a sharp positive excursion to –21‰ within the radiolarian T-J boundary zone (Fig. 1). This structure is similar to the initial negative excursion and subsequent positive recovery, respectively, which have been already documented in records from Europe and North America (e.g., Hesselbo et al., 2002; Guex et al., 2004; Williford et al., 2007; Ruhl et al., 2009). The positive shift of δ13Corg occurs within the radiolarian T-J boundary zone in the Kurusu section, while the positive recovery of δ13Corg precedes the T-J boundary in the St. Audrie’s Bay section (Somerset, England; Fig. 1). We consider that the base of radiolarian T-J boundary zone in the Kurusu section would be correlated to the Langport Member in St. Audrie’s Bay, which is characterized by the transition from the initial negative excursion to the positive δ13Corg values (dashed line in Fig. 1). The age-corrected Os isotopic record (187Os/188Osi) from the Kurusu section shows a minimum in the late Rhaetian and a maximum at the Hettangian-Sinemurian boundary (Fig. 1). The hydrogenous 187Os/188Osi record shows a gradual decrease from 0.6 in the early Rhaetian to 0.2 in the late Rhaetian; the lowest value is observed at sample KU-4, ~0.7 m below the radiolarian T-J boundary zone. This gradual decrease is overlain by a 50-cm-thick interval (from KU-4 to KU+23) that records a sharp increase in 187Os/188Osi values up to ~0.52 during the latest Rhaetian (1st rise, Fig. 1). The sample with the highest Os concentration (14.5 pg g–1; KU+6U) is located in the middle of the 1st rise, showing no distinctive isotopic anomaly. The 187Os/188Osi profile shows a stable phase from the end of the Rhaetian through the early Hettangian, with minor fluctuations

1096

between values of 0.4 and 0.5. An isotopic excursion toward radiogenic values (~0.65) is observed around the Hettangian-Sinemurian boundary (2nd rise, Fig. 1). Subsequently, 187Os/188Osi values become stable at ~0.4– 0.5 in the Sinemurian interval. The gradual decrease in seawater 187Os/188Osi values in the Late Triassic (Fig. 1) indicates a gradual increase in the relative supply rate of unradiogenic Os. Such a gradual shift cannot be explained by extraterrestrial impact, as this would have produced a sharp decrease in the 187 Os/188Osi values followed by a rapid recovery, as observed in sediments from the Cretaceous-Paleogene boundary (Ravizza and Peucker-Ehrenbrink, 2003; Robinson et al., 2009). A decrease in the flux of radiogenic Os from continental crust is also unsatisfactory as an explanation of the observed decrease in seawater 187Os/188Osi values to 0.2, as this scenario requires an unreasonably large decrease in the global rate of continental weathering (to nearly zero). Therefore, an increase in the flux of unradiogenic Os from the mantle is necessary to explain the observed decrease in seawater 187Os/188Osi values. This shift would have started from the early Rhaetian (ca. 204–203 Ma; Ogg et al., 2008), predating the T-J boundary by ~3 m.y. (Fig. 1). Although most of basaltic rocks from the CAMP show ages of ca. 201–196 Ma (e.g., Marzoli et al., 2004; Schoene et al., 2006), the initiation of the CAMP formation is considered to be ca. 203 Ma, in particular the north African CAMP (Nomade et al., 2007). Therefore, we attribute the gradual decrease to an increase in the supply rate of unradiogenic Os associated with CAMP volcanism (Cohen and Coe, 2002, 2007). The Os isotopic data from the Kurusu section mark the onset of the CAMP, indicating that released Os would have diffused into Panthalassa within the residence time of Os.

GEOLOGY, December 2010

Similar gradual decreases in Os isotopic composition have been observed for the early Aptian and the latest Maastrichtian, coincident with emplacement of the Ontong Java Plateau (Tejada et al., 2009) and Deccan Traps (Robinson et al., 2009), respectively. However, the rate of decrease in seawater 187Os/188Osi value is