Correlating the End-Triassic Mass Extinction and Flood Basalt
Total Page：16
File Type：pdf, Size：1020Kb
Correlating the end-Triassic mass extinction and fl ood basalt volcanism at the 100 ka level Blair Schoene1*†, Jean Guex2†, Annachiara Bartolini3†, Urs Schaltegger1†, and Terrence J. Blackburn4† 1Earth Sciences, University of Geneva, Rue des Maraîchers 13, CH-1205 Geneva, Switzerland 2Geology and Paleontology, University of Lausanne, l’Anthropole, Lausanne, Switzerland 3Muséum National d’Histoire Naturelle, Histoire de la Terre, CP 38 CR2P UMR 7207 du CNRS, 8 rue Buffon, Paris, France 4Earth, Atmosphere and Planetary Sciences, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, Massachusetts 02142-1479, USA ABSTRACT Proxies for rising atmospheric CO2 have been New high-precision U/Pb geochronology from volcanic ashes shows that the Triassic-Juras- reported from terrestrial fossil plants straddling sic boundary and end-Triassic biological crisis from two independent marine stratigraphic the Triassic-Jurassic boundary (McElwain et al., sections correlate with the onset of terrestrial fl ood volcanism in the Central Atlantic Mag- 1999; Retallack, 2001), though the effects of matic Province to <150 ka. This narrows the correlation between volcanism and mass extinc- other gases such as SO2 on such proxies may tion by an order of magnitude for any such catastrophe in Earth history. We also show that also be important (Guex et al., 2004; Tanner et a concomitant drop and rise in sea level and negative δ13C spike in the very latest Triassic al., 2007). Terrestrial correlatives to the marine occurred locally in <290 ka. Such rapid sea-level fl uctuations on a global scale require that extinction are debated (Lucas and Tanner, 2007; global cooling and glaciation were closely associated with the end-Triassic extinction and Tanner et al., 2004). An apparent palynological potentially driven by Central Atlantic Magmatic Province volcanism. event <1 m below the lowest CAMP basalt in the Newark and Fundy Basins in North America INTRODUCTION Mountain Basalt, the lowest CAMP basalt from was proposed as correlative of the Triassic- Mass extinctions refl ect important interac- the Fundy Basin, Nova Scotia (Greenough and Jurassic boundary (Whiteside et al., 2007); this tions between biology, geology, geochemical Dostal, 1992). Both the ash beds and the North has been challenged on the basis of biostrati- cycles, and climate. The end-Triassic mass Mountain Basalt were dated using chemical graphic and magnetostratigraphic work from extinction is one of the fi ve largest extinctions in abrasion–isotope dilution–thermal ionization North America and Morocco (Marzoli et al., Earth history, though considerable uncertainty mass spectrometry (CA-ID-TIMS; Mattinson, 2004, 2008). Others argue that vertebrate and remains in terms of its duration, causes, and 2005) U-Pb zircon geochronology employing a palynological biostratigraphy in the Newark and effects. Many workers suggest that the extinc- new well-calibrated 202Pb-205Pb-233U-235U tracer Fundy Basins, respectively, place the Triassic- tion was related directly or indirectly to adverse solution, which removes random uncertainty in Jurassic boundary in sedimentary slivers above climate following the onset of the Central Atlan- mass fractionation during mass spectrometry. the North Mountain Basalt (Lucas and Tanner, tic Magmatic Province (CAMP), which erupted Data with this solution are as much as 70% more 2007; Cirilli et al., 2009). >2.5 × 106 km3 of basalt, possibly in <1 Ma, precise compared to single-Pb/single-U tracers, An age for the marine Triassic-Jurassic making it perhaps the most voluminous fl ood revealing complexity in tuff zircon populations boundary comes from the Pucara basin in basalt sequence of the Phanerozoic (Marzoli et that require new data interpretation strategies. northern Peru, where Schaltegger et al. (2008) al., 2004; McHone, 2003; Nomade et al., 2007; reported a weighted-mean 206Pb/238U date of Whiteside et al., 2007). However, there remains TRIASSIC-JURASSIC BOUNDARY 201.58 ± 0.18/0.38 Ma (2σ; without/with a need for precise and accurate geochronol- Recent consensus places the Triassic-Jurassic decay constant uncertainties). Abundant CAMP ogy to correlate the onset of CAMP volcanism, boundary at the fi rst occurrence of the oldest 40Ar/39Ar data cluster at 199 Ma (e.g., Nomade recorded uniquely in terrestrial sections, with the Jurassic ammonite Psiloceras spelae, which et al., 2007), but uncertainties of 1–2 Ma on indi- well-documented marine extinction event (Mar- marks the beginning of postextinction biodi- vidual dates in addition to the well-documented zoli et al., 2008; Tanner et al., 2004; Whiteside versity recovery (Guex et al., 2004; Morton ~0.7%–1% bias between the 40Ar/39Ar and U-Pb et al., 2007). Also lacking are time constraints and Hesselbo, 2008). Pinpointing the extinction dating methods (Kuiper et al., 2008; Schoene for the rates of the Triassic-Jurassic bound- interval is more complicated, but coincides with et al., 2006) make this correlation imprecise. A ary extinction and associated geochemical and a sharp negative spike in δ13C at the end-Triassic, 206Pb/238U date from the North Mountain Basalt paleoenvironmental fl uctuations. We sampled when there were steep declines in the bio- of 201.27 ± 0.06/0.30 Ma (Schoene et al., 2006) three volcanic ash beds bracketing the Triassic- diversity of ammonites, bivalves, radiolarians, would suggest that the CAMP postdates the Tri- Jurassic boundary from the Pucara basin, north- corals, and conodonts (Morton and Hesselbo, assic-Jurassic boundary, precluding a causative ern Peru (Fig. 1A; Schaltegger et al., 2008), and 2008). This initial negative excursion is fol- relationship. However, those two U-Pb dates also the fi rst discovered ash bed from the New lowed by a gradual positive recovery (Fig. 1B), were measured using different tracer solutions, York Canyon, Nevada, which has been proposed which precedes a slow negative excursion in allowing for systematic bias and preventing as the Global Boundary and Stratotype Sec- the Early Jurassic (beginning at bend N13 in high-precision comparison. tion and Point for the Triassic-Jurassic bound- Fig. 1B; Guex et al., 2004; Hesselbo et al., 2004; ary (Guex et al., 2004). We also provide new Kuerschner et al., 2007; Ward et al., 2001). The LOCALITIES AND U-Pb U/Pb zircon data from two labs for the North end-Triassic negative δ13C excursion is recorded GEOCHRONOLOGY in marine organic and carbonate carbon and We sampled three volcanic ash beds brack- *Current address: Geosciences, Guyot Hall, Princ- continent-derived wood material, illustrating eting the Triassic-Jurassic boundary in the eton University, Princeton, New Jersey 08544, USA. †E-mails: [email protected]; Jean.Guex@ that the anomaly resulted from a global carbon Pucara basin (samples LM4–86, LM4–90, and unil.ch; [email protected]; [email protected]; cycle perturbation (Galli et al., 2005; Hesselbo LM4–100/101; Figs. 1A, 1B), which is well [email protected]. et al., 2004; Pálfy et al., 2001). calibrated biostratigraphically (Schaltegger et © 2010 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or [email protected]. GEOLOGY,Geology, May May 2010; 2010 v. 38; no. 5; p. 387–390; doi: 10.1130/G30683.1; 1 fi gure; Data Repository item 2010109. 387 Downloaded from http://pubs.geoscienceworld.org/gsa/geology/article-pdf/38/5/387/3537956/387.pdf by Princeton University user on 19 October 2020 Figure 1. A: Location map of three sections A 2. Pucara basin (N. Peru) 3. New York Canyon (Nevada, USA) studied, with ca. 200 Ma + Sea level paleogeography. Num- M2 bers correspond to stratigraphic sections in FO Nevadaphyllites N13 B. Red area outlines ap- N11 proximate extent of Cen- tral Atlantic Magmatic B 1. Fundy Basin Hettangian (J) Province (e.g., McHone, LM4-100/101 2003) B: Stratigraphic (Nova Scotia, Canada) NYC-N10 N9 columns for sections Hettangian (J) studied; scale bars at J McCoy Brook Fm. TJB bottom. J—Jurassic; FO P. spelae Tr—Triassic. Fundy FO P. spelae Basin section is after LM4-90 Whiteside et al. (2007). North Mtn. N6c Pucara basin biostra- Basalt N6b tigraphy is detailed in LM4-86 Schaltegger et al. (2008). N3 extinction intervalextinction <<290,000 years New York Canyon stra- LO C. crickmayi intervalextinction tigraphy, biostratig- raphy, bed numbers, Blomidon Formation AC5 LO C. crickmayi Cooling and glaciation Global warming carbon isotopes, and Tr 50 m sea-level curve are from Rhaetian 1 m Guex et al. (2004, 2008). Thick bed limestone δ13 Age of TJB Green curve in C plot Thin bed limestone is running mean of red Duration of extinction, δ13C spike, -31 -30 -29 -28 -27 -26 -25 Carb. siltstone Ash bed data points. FO—fi rst regression/transgression couplet: Rhaetian (Tr) Siltstone Black shale δ 13C occurrence. LO—last 1 m org occurrence; org— 204 organic. Stars indicate C North Mtn. Basalt ash beds sampled, with NMB-03-1 interpreted 206Pb/238U 203 deposition age. All Zircon used as eruption age uncertainties are 2σ. 202 TJB—Triassic-Jurassic boundary, defi ned as North U date (Ma) 201 Mtn. Basalt fi rst occurrence of Psilo- UNIGE 206 238 238 ceras spelae. C: Pb/ U Pb-loss Pre-eruptive growth of zircon? dates for single-grain Pb/ 200 zircons, color-coded to LM4-86 (Schaltegger et al., 2008) Antecrysts? sample locations in B. 206 Xenocrysts? Date for LM4–86 from Schaltegger et al. (2008) includes tracer calibration uncertainty. Data for North Mountain Basalt are from Massachusetts Institute of Tech- nology (MIT) and University of Geneva (UNIGE); all ash-bed data are from University of Geneva. al., 2008). In this section, the disappearance dia plots and U-Pb data are presented in Fig- between 201.3 and 201.9 Ma; a weighted-mean of the latest Triassic ammonite Choristoceras ure DR1 and Table DR1, respectively. yields an MSWD of >2. Two younger grains crickmayi immediately precedes the peak Ash beds yielded 20–100 zircons between 50 give 206Pb/238U dates younger than 201 Ma and extinction rate (Guex et al., 2004).