Available online at www.sciencedirect.

com

Earth and Planetary Science Letters 267 (2008) 266 – 275

www.elsevier.com/locate/epsl

Precise U–Pb age constraints for end-Triassic mass extinction, its correlation
to volcanism and Hettangian post-extinction recovery
Urs Schaltegger a , Jean Guex b,⁎, Annachiara Bartolini c , Blair Schoene a , Maria Ovtcharova a
a
Department of Earth Sciences, University of Geneva, rue des Maraîchers 13, 1205 Geneva, Switzerland
b
Department of Geology and Paleontology, University of Lausanne, l'Anthropole, Lausanne, Switzerland
c
Université “Pierre et Marie Curie” - Paris VI, CNRS-UMR 5143 “Paléobiodiversité et Paléoenvironnements”, 4, place Jussieu, Paris, France
Received 24 June 2007; received in revised form 30 October 2007; accepted 25 November 2007
Available online 4 December 2007
Editor: C.P. Jaupart

Abstract

New precise zircon U–Pb ages are proposed for the Triassic–Jurassic (Rhetian–Hettangian) and the Hettangian–Sinemurian boundaries.
The ages were obtained by ID-TIMS dating of single chemical-abraded zircons from volcanic ash layers within the Pucara Group, Aramachay
Formation in the Utcubamba valley, northern Peru. Ash layers situated between last and first occurrences of boundary-defining ammonites
yielded 206Pb/238U ages of 201.58 ± 0.17/0.28 Ma (95% c.l., uncertainties without/with decay constant errors, respectively) for the Triassic–
Jurassic and of 199.53 ± 0.19 / 0.29 Ma for the Hettangian–Sinemurian boundaries. The former is established on a tuff located 1 m above the
last local occurrence of the topmost Triassic genus Choristoceras, and 5 m below the Hettangian genus Psiloceras. The latter sample was
obtained from a tuff collected within the Badouxia canadensis beds. Our new ages document total duration of the Hettagian of no more than
c. 2 m.y., which has fundamental implications for the interpretation and significance of the ammonite recovery after the topmost Triassic
extinction.
The U–Pb age is about 0.8 ± 0.5% older than 40Ar–39Ar dates determined on flood basalts of the Central Atlantic Magmatic Province (CAMP).
Given the widely accepted hypothesis that inaccuracies in the 40K decay constants or physical constants create a similar bias between the two
dating methods, our new U–Pb zircon age determination for the T/J boundary corroborates the hypothesis that the CAMP was emplaced at the
same time and may be responsible for a major climatic turnover and mass extinction. The zircon 206Pb/238U age for the T/J boundary is marginally
older than the North Mountain Basalt (Newark Supergroup, Nova Scotia, Canada), which has been dated at 201.27 ± 0.06 Ma [Schoene et al.,
2006. Geochim. Cosmochim. Acta 70, 426–445]. It will be important to look for older eruptions of the CAMP and date them precisely by U–Pb
techniques while addressing all sources of systematic uncertainty to further test the hypothesis of volcanic induced climate change leading to
extinction. Such high-precision, high-accuracy data will be instrumental for constraining the contemporaneity of geological events at a 100 kyr
level.
© 2007 Elsevier B.V. All rights reserved.

Keywords: U–Pb dating; volcanic ash beds; Triassic–Jurassic boundary; Peru; mass extinction; post extinction recovery

1. Introduction volving the disappearance of about 80% of the known species

(Sepkoski, 1994; Hallam and Wignall, 1999). Widespread
The end-Triassic extinction is considered to be one of the magmatic activity of the Central Atlantic Magmatic Province
most important biotic crises our planet has experienced, in- (CAMP) has repeatedly been invoked to have caused this ca-
tastrophic event. A detailed model taking into account all of the
environmental perturbations known to have occurred during
⁎ Corresponding author.
this time was proposed recently by Guex et al. (2004). In this
E-mail addresses: urs.schaltegger@terre.unige.ch (U. Schaltegger),
jean.guex@unil.ch (J. Guex), Bartolini@mnhn.fr (A. Bartolini), paper we suggested that the main stresses on the environment
blair.schoene@terre.unige.ch (B. Schoene), maria.ovtcharova@terre.unige.ch may have been generated by repeated release of SO2 gas, heavy
(M. Ovtcharova). metals emissions, darkening and subsequent cooling causing an
0012-821X/$ - see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.epsl.2007.11.031
 U. Schaltegger et al. / Earth and Planetary Science Letters 267 (2008) 266–275 267

Fig. 1. A) Location of Chachapoyas in N-Peru. B) Geological sketch map of the distribution of the Pucara Formation in north-central Peru. Inset: Area of Chachapoyas,
showing the location of the studied Triassic–Jurassic boundary section in the Utcubamba Valley.

important regressive event. This phase was followed by a ma- characterizes the period of the post-extinction recovery of the
jor long term CO2 accumulation during the Early Hettangian ammonoids.
with development of nutrient rich marine waters favouring the Detailed biostratigraphic research has been carried out
recovery of productivity and, elsewhere, deposition of black during the last few years in the Utcubamba Valley (N Peru;
shales. The relationship between the extinction and its probable Fig. 1) along a new and fresh road section from Levanto to
volcanic cause can only be established by demonstrating the Maino. This new section preserves a complete deep marine
synchrony of the two events. This requires accurate and precise sedimentary sequence overlapping the upper Rhaetian to the
ages for both the TJB strata in a perfectly calibrated marine Early Sinemurian, complementing the detailed stratigraphy of
section and of volcanic rocks of the CAMP. The Hettangian Hillebrandt (2000).

Fig. 2. Detailed stratigraphic section with ammonoid biostratigraphy. TJB = Triassic Jurassic Boundary = Rhaetian–Hettangian boundary. HSB = Hettangian–
Sinemurian Boundary.
268 U. Schaltegger et al. / Earth and Planetary Science Letters 267 (2008) 266–275

The map in Fig. 1 gives the location of the section and a In addition to detailed biostratigraphic results, this paper pre-
measured stratigraphic column is provided in Fig. 2. The position sents new ID-TIMS U–Pb ages for zircons from ash beds close
of the Triassic–Jurassic boundary (TJB) in that section has been to the Triassic–Jurassic boundary (TJB) and the Hettangian–
precisely located by use of the ammonites indicated in Fig. 2. Ash Sinemurian boundary (HSB). These ages therefore directly brack-
beds containing zircon have been found interspersed throughout et the duration of the Early Jurassic biotic recovery and also test the
the section, allowing temporal calibration of the biostratigraphy possible causal relationship between the TJB extinction and the
with precise and accurate U–Pb zircon age determinations. eruption of the Central Atlantic Magmatic Province (CAMP).

Fig. 3. All specimens from the Levanto section (see Fig. 2 for the stratigraphic position of the taxa). A) Vandaites saximontanus, D = 29 mm; B) Choristoceras
crickmayi, D = 16 mm; C) Psiloceras spelae, D = 17 mm; D) Psiloceras tilmanni, D = 43 mm; E) Nevadaphyllites sp, D = 21 mm; F) Kammerkarites sp, D = 43 mm;
G) Angulaticeras sp, D = 30 mm; H) Badouxia canadensis D = 52 mm.
 U. Schaltegger et al. / Earth and Planetary Science Letters 267 (2008) 266–275 269

2. Regional geology, stratigraphy and ammonoid age control

The detailed paleogeography of the Pucara basin in northern

Peru is not well established. However, there is a major stra-
tigraphic trend in the Utcubamba Valley such that marine strata
from the uppermost Triassic and Lower Jurassic represent much
shallower water depths in the south near Chilingote, 20 km
south of Suta (see Fig. 2) where we find limy sequences rich in
three-dimensional ammonites (see also Hillebrandt, 2000). The
stratigraphic sequence near Suta is intermediate and the thickest
sequence is located at Levanto where we collected our tuff beds
(Fig. 2). These findings indicate a S–N deepening trend of the
basin around the TJB. The original paleogeographic scheme of
Rosas (1994) indicates a Toarcian volcanic arc to the west of the
Pucara basin, which is bordered by Gondwanan Paleozoic rocks
at the margin of the Amazonian craton to the east. Our new data
may allow the speculation that the Pucara basin may have been
situated behind a volcanic arc, which was active during the
Upper Triassic (latest Norian) to Lower Jurassic further to the
west and closer to the continental margin. The oldest Mesozoic
volcanism along the Peruvian segment of the pre-Andean con-
tinental margin known until now was located in the Toarcian and
little was known about the occurrence of Triassic volcanism.
Our new section (Fig. 2) is a thick monotonous sequence of
siltstones alternating with slightly more calcareous silty beds.
More than 20 fossiliferous beds have been excavated, allowing a
very precise correlation with the standard ammonoid zonations
used in the Upper Rhaetian and Lower Jurassic. The most im-
portant ammonites, which our correlations are based on, are
illustrated in Figs. 3 and 4. The field numbers of the beds used
below correspond to black dots in the stratigraphic column of
Fig. 2. Bed 29 contains a rich fauna of Vandaites saximontanus
(Middle Rhaetian; Fig. 3A), occurring a few tenth of meters
above the Upper Norian Monotis beds. Unfortunately the Norian-
Rhaetian contact is faulted and hidden. Beds 53 to 82 give quite
frequent Choristoceras ex gr marshi. The last Choristoceras bed
is 84b where we got a single specimen of C. crickmayi (Fig. 3B). Fig. 4. All specimens from the Levanto section (see Fig. 2 for the stratigraphic
The next ammonitiferous bed, 93b, gave us one specimen of position of the taxa). A) Metophioceras sp, D = ca 110 mm; B) Vermiceras
densicostatum, D = 46 mm; C) Coroniceras sp, H = 62 mm.
Psiloceras spelae (Fig. 3C), providing an excellent correlation
with the equivalent beds in Nevada (see Guex et al., 2004).
Bed 104 contains a few crushed specimens of Nevadaphyllites? the first occurrence of Badouxia canadensis and below Coroni-
(Fig. 3E) and beds 114 to 129 delivered a relatively abundant ceras. These two genera bracket 12 m of stratigraphic column,
fauna of Psiloceras tilmanni (Fig. 3D). Kammerkarites spp which contains the Hettangian - Sinemurian boundary. Note that
(Fig. 3F) have been found in beds 170 to 177 and bed 196 gave the correlation of the Canadensis zone has been the subject of
poorly preserved Schlotheimia and Alsatites, indicating a Mid- controversies in the past but now it is generally accepted that the
Hettangian age. The next 10 m of sediments are extensively Hettangian–Sinemurian boundary lies within the North-Amer-
fractured and covered by talus, and the next fauna consists in ican Canadensis zone, i.e. above the first occurrence of B.
finely costate, evolute, Angulaticeras (bed 203 LM 3a; Fig. 3G). canadensis (see Taylor and Guex, 2002).
A few meters above we collected a well-preserved fauna of Ba-
douxia canadensis (bed 213 LM-3 b; Fig. 3H), followed 10 m 3. U–Pb age determinations
above by a Coroniceras (240 LM 3 C; Fig. 4C). The last in-
vestigated fossiliferous Sinemurian levels provided Vermiceras 3.1. Analytical techniques
densicostatum Taylor (bed 246 LM 3 D; Fig. 4B) and Meto-
phioceras (beds 246-257) (Fig. 4A). The first ash layer analysed It has been repeatedly shown that high-precision ID-TIMS
in the present paper comes very near the Triassic–Jurassic (Isotope dilution — thermal ionization mass spectrometry) U–
boundary, bracketed between the last C.crickmayi and Psiloceras Pb analyses of single zircon crystals provide the most precise
spelae (see Fig. 2). The second ash bed has been collected above and accurate age data (e.g., Mundil et al., 2004; Ovtcharova
270 U. Schaltegger et al. / Earth and Planetary Science Letters 267 (2008) 266–275

et al., 2006). U–Pb in zircon is the most reliable chrono- system (other analyses). This multiplier was exchanged by an-
meter, because zircon has the lowest diffusion coefficients for other multiplier, called MasCom-2, in the course of this project.
Pb (Cherniak and Watson, 2001), and is resistant to post- The linearity of the ETP multiplier was calibrated using the
crystallization disturbance. Nevertheless, complications arise SRM982 standard, the MasCom multipliers were calibrated
due to two effects: (1) Post-crystallization loss of radiogenic using U500, Sr SRM987, and Pb NBS982 and SRM983 so-
lead due to elevated temperatures or during fluid percolation, lutions. The mass fractionation of Pb was controlled by repeated
which is enhanced according to the degree of radioactive decay SRM981 measurements (0.09 ± 0.05%/amu on ETP, 0.11 ±
enduced damage of the crystalline structure, and (2) incorpora- 0.05%/amu on MasCom-1, and 0.08 ± 0.05%/amu on MasCom-
tion of old cores acting as nucleii during crystallization, such as 2 multipliers, respectively). Both lead and uranium were loaded
inherited xenocrystic cores or “antecrystic zircons” crystallized with 1 μl of silica gel–phosphoric acid mixture (Gerstenberger
earlier in the magma chamber but showing a distinctly older age and Haase, 1997) on outgassed single Re-filaments, and Pb
(e.g., Bachmann et al., 2007). Pb-loss is at least partly com- as well as U (as UO2) isotopes measured sequentially on the
pensated for by treating the zircon with annealing-leaching electron multiplier. Total procedural Pb blank was estimated
(« chemical abrasion ») techniques prior to analysis, in order to at 0.5 ± 0.2 pg and corrected with the following isotopic
remove lattice domains that are severely disturbed by decay composition: 206Pb/204Pb: 19.67 ± 0.09, 207Pb/204Pb: 16.63 ±
damage (Mattinson, 2003; Mundil et al., 2004; Mattinson, 0.23, 208Pb/204Pb: 39.29 ± 0.15 (all 2 sigma). Common lead in
2005). excess of blank was corrected for using the model of Stacey
The techniques of sample preparation, zircon annealing and and Kramers (1975) for an age of 200 Ma. The uncertainty of
leaching, dissolution and chemical separation of Pb and U used the concentration of U and Pb in the spike solution (±0.1%) was
in this study are identical to those described in Ovtcharova et al. taken into account and propagated to each individual analysis.
(2006). Isotopic analysis was partly performed at ETH Zürich Calculation of concordant ages and averages was done with
on a MAT262 mass spectrometer equipped with a ETP electron the Isoplot/Ex v.3 program of Ludwig (2005). All uncertainties
multiplier (analyses 13–18; analytical details see Ovtcharova reported are at the 2 sigma level. The accuracy of the data was
et al., 2006), partly at University of Geneva on a TRITON mass assessed by repeated analysis of the international R33 standard
spectrometer from Thermo Fisher equipped with a modified zircon (Black et al., 2004), which was pretreated by annealing–
MasCom-1 electron multiplier backed by a digital ion counting leaching and measured at an average 206Pb/238U age of 419.19 ±

Table 1
Results of U–Pb age determinations
Number Weight Concentrations Atomic ratios Error Apparent ages
[mg] corr.
U Pb rad. Pb nonrad. Th/U 206/204 207/206 Error 207/235 Error 206/238 Error 206/238 207/235 207/206
2 s [%] 2 s [%] 2 s [%]
[ppm] [ppm] [pg] b) c) d) e) d) d) e) e) e)
Sample 86
1 0.0020 395.0 13.37 1.40 0.59 1160 0.05016 0.13 0.2194 0.27 0.03172 0.23 0.90 201.33 201.38 202.36
2 0.0014 189.9 6.70 0.54 0.75 1047 0.05014 0.29 0.2199 0.39 0.03181 0.24 0.67 201.89 201.85 201.45
3 0.0016 64.4 2.46 0.60 1.09 375 0.04997 0.79 0.2205 0.87 0.03200 0.26 0.44 203.07 202.33 193.55
4 0.0014 179.3 6.69 0.65 0.98 826 0.05014 0.43 0.2196 0.52 0.03176 0.24 0.57 201.57 201.56 201.45
5 0.0027 41.1 1.68 0.55 1.32 439 0.05020 0.63 0.2226 0.71 0.03216 0.25 0.48 204.06 204.10 204.25
6 0.0028 167.4 6.12 0.83 0.83 1208 0.05022 0.32 0.2202 0.41 0.03180 0.24 0.63 201.83 202.10 205.23
7 0.0010 a) 38.9 1.46 0.76 0.90 1069 0.05084 0.35 0.2197 0.41 0.03134 0.20 0.52 198.95 201.68 233.61
8 0.0010 a) 5.5 0.22 0.69 1.22 185 0.05018 2.45 0.2198 2.54 0.03176 0.24 0.42 201.56 201.70 203.34
9 0.0010 a) 13.0 0.53 0.37 1.45 746 0.05007 0.54 0.2191 0.61 0.03174 0.23 0.48 201.43 201.19 198.24
10 0.0010 a) 19.2 0.73 0.40 1.04 1013 0.05016 0.36 0.2196 0.43 0.03175 0.21 0.55 201.51 201.60 202.36

Sample L-19
11 0.0063 295.5 11.14 0.54 0.13 7392 0.08106 0.07 0.3653 0.24 0.03269 0.23 0.96 207.34 316.16 1222.89
12 0.0050 227.1 7.42 1.80 0.50 1287 0.05013 0.16 0.2171 0.30 0.03140 0.25 0.85 199.33 199.47 200.99
13 0.0118 72.8 2.42 6.04 0.58 301 0.05030 0.29 0.2170 0.39 0.03129 0.25 0.67 198.64 199.43 208.89
14 0.0034 144.7 4.92 0.70 0.67 1426 0.05006 0.17 0.2162 0.32 0.03132 0.26 0.85 198.83 198.70 197.79
15 0.0080 93.4 3.10 1.47 0.54 1029 0.05018 0.21 0.2174 0.34 0.03142 0.25 0.79 199.46 199.77 206.60
16 0.0093 61.8 2.09 2.18 0.63 550 0.05021 0.38 0.2175 0.48 0.03141 0.26 0.62 199.39 199.81 204.71
17 0.0101 169.2 5.74 0.58 0.64 6105 0.05011 0.11 0.2173 0.26 0.03144 0.24 0.91 199.58 199.62 200.08
18 0.0143 139.8 4.78 0.58 0.68 7155 0.05010 0.08 0.2172 0.25 0.03144 0.23 0.95 199.58 199.58 199.62
19 0.0083 166.7 5.86 0.89 0.75 3227 0.050212 0.08 0.2179 0.24 0.03147 0.23 0.94 199.77 200.13 204.71
a) Estimated weight.
b) Calculated on the basis of radiogenic Pb208/Pb206 ratios, assuming concordancy.
c) Corrected for fractionation and spike.
d) Corrected for fractionation, spike, blank and common lead (Stacey and Kramers, 1975).
e) Corrected for initial Th Disequilibrium, using an estimated Th/U ratio of 4 for the melt.
 U. Schaltegger et al. / Earth and Planetary Science Letters 267 (2008) 266–275 271

0.20 Ma (N = 30; MSWD = 1.17) for both ETH and Univ. of Sinemurian boundary Fig. 2, ash bed B). A total of nine zircons
Geneva measurements, and at 419.08 ± 0.19 Ma (N = 27; have been analyzed (Table 1); one grain is clearly biased by
MSWD = 0.70) for measurements at Univ. of Geneva alone. old inherited lead (analysis 11), pointing to the presence of
Proterozoic xenocrystic zircon in the melt. The rest of the
3.2. Results zircons plot close to or on concordia (Fig. 5b). Analyses 12 and
15–19 yield a mean 206Pb/238 U age of 199.53 ± 0.19/0.29 Ma
The ash bed of sample 86 is located at the very top of the (MSWD = 0.42; Table 1, Fig. 5b), which is considered to be
Rhaetian (Fig. 2, Ash bed A), just above the last local occurrence representative of zircon crystallization and ash bed deposition.
of Choristoceras crickmayi and 5 m below the first occurrence Analyses 13 and 14 have been excluded from age calculations
of Psiloceras spelae (Fig. 3). The sample yielded abundant because their 206 Pb/238 U ages of 198.64 and 198.83 Ma, res-
prismatic to long-prismatic zircons. Ten grains were selected for pectively, are slightly younger and probably have suffered a
analysis, seven of which are concordant with a mean 206 Pb/238U small amount of lead loss.
age of 201.58 ± 0.17/0.28 Ma (uncertainty without/with decay
constant uncertainties; MSWD = 0.72; Table 1, Fig. 5a). This age 4. Discussion
is considered to be zircon crystallization and deposition of
this ash tuff. Analyses 3 and 5 yielded 206Pb/238U ages of 203.07 4.1. The age of the TJB and its synchrony with CAMP
and 204.06 Ma, which are older and are outside of analytical magmatism
uncertainty of the other points. Analysis 7 is discordant and
indicating lead loss. The currently accepted age estimate of the TJB (199.6 ±
The ash bed of sample L-19 is located between the last local 0.3 Ma) is based on a weighted mean 206Pb/238U zircon age
occurrence of Badouxia canadensis and the first occurrence of from a marine section in the Queen Charlotte Island of Canada
Coroniceras sp. and lies therefore very close to the Hettangian– Pálfy et al. (2000). This age was calculated from three

Fig. 5. U–Pb concordia diagrams of sample 86 (a) and sample L19 (b).
272 U. Schaltegger et al. / Earth and Planetary Science Letters 267 (2008) 266–275

concordant multigrain analyses; the entire dataset, however Although our age for the TJB is statistically older than the
contains analyses that show excess scatter in their 206 Pb/238 U age of the NMB, several sources of uncertainty prevent a direct
values, probably reflecting combined effects of lead loss and comparison of the initiation of flood basalt magmatism with the
inheritance. More recently, Mundil et al. (2005) and Mundil and ammonite extinction event recorded in the Peruvian section.
Pálfy (2005) commented that those data were possibly biased by First, the ash bed of sample 86 (Fig. 2) is about 1 m above the
unresolved Pb loss, resulting that the former TJB estimate is last Triassic genus Choristoceras and 5 m below the first oldest
slightly too young — a conclusion that is consistent with the Jurassic Psiloceras (P. spelae), which bracket the location of the
age of ca. 201.58 Ma reported in this study. real TJB. From our age determinations we can estimate the
Pálfy et al. (2000) compared their age to the timing of CAMP rate of deposition to be c. 15 kyr/m, which means that the “true”
magmatism, which occurred within 1 m.y. at an age of ca. TJB may be by c. 75 kyr or more younger. Second, the two age
200 Ma (Marzoli et al., 1999; Hames et al., 2000; Marzoli et al., determinations were carried out using two different isotope
2004). Courtillot and Renne (2003) supported an estimated tracer solutions that were not intercalibrated. Despite our efforts
duration of some 500 ± 100 kyr for the entire CAMP magmatism to quantify and propagate these uncertainties into the age,
based of magneto- and cyclostratigraphic arguments, which is possible unknown systematic uncertainties in tracer calibration
in agreement with the data and estimations of Marzoli et al. may be apparent. Third, the sample of the NMB dated by
(2004), Knight et al. (2004), Hames et al. (2000) and others. A Hodych and Dunning (1992) does not necessarily represent the
major problem of correlating age determinations from CAMP earliest basalts present in the area nor does the 190 m thick flow
basalts wish ash beds in sedimentary successions concerns at the base of NMB necessarily represent one single event. The
the systematic offset of the K–Ar and U–Pb decay schemes: authors assume that their dated NMB sample is probably up to
Numerous previous studies have noticed that 40Ar-39Ar dates 200 kyr younger than the TJB. Consequently, though the ages
are systematically 0.3 to 1.0% younger than U–Pb ages of the for the NMB reported by Schoene et al. (2006) and Hodych and
same rocks (Renne et al., 1998; Min et al., 2000; Renne, 2000; Dunning (1992) likely represent the age of early NMB eruption,
Villeneuve et al., 2000; Min et al., 2001; Schmitz and Bowring, its correlation with the TJB and the oldest CAMP basalts from
2001), which cannot entirely be related to sequential closure other areas remains questionable. For example, Whiteside et al.
of the two isotopic systems during cooling. There is a broad (2007) conclude that the earliest CAMP basalts are younger
agreement that much of the bias can be accounted for by an than the TJ crisis based on palynological data from the Bay of
inaccurate 40K decay constant or physical constants (e.g., Renne Fundy section of the Newark basin, where the NMB is located,
et al., 1998; Min et al., 2003). This fact led to the suggestion implying that there is no causal relationships between the TJ
that the Ar–Ar system should be calibrated against the U– extinction and the onset of that large igneous province. Their
Pb system, resulting in a 0.8% higher age of approximately conclusion is mainly based on palynological data established in
28.24 Ma for the Fish Canyon sanidine (instead of 28.02 Ma). an exceedingly short stratigraphic section, and reexamination of
Such an age has also been suggested from cross-calibration with their data leads to the conclusion that all the pollens recorded in
astronomically tuned volcanic ashes (Kuiper et al., 2004, 2005). that section are of Triassic age: the relatively abundant Coral-
Villeneuve et al. (2000) arrived, however, at a much lower lina meyeriana first occurs in the upper Triassic, as demon-
207
Pb/235U zircon-calibrated age of 27.98 ± 0.10 Ma for Fish strated by Kuerschner et al. (2007). This draws into question the
Canyon Tuff sanidine, identical to the one of Renne et al. (1994) placement of the TJB relative to the NMB made by Whiteside
based on astronomical tuning, and close to the generally accepted et al. (2007). A more robust argument is made by the very
value of 28.02± 0.28 Ma achieved by Renne et al. (1998) through precise and convincing conchostracan data established by
intercalibration with other Ar–Ar standards. Schoene et al. (2006) Kozur and Weems (2005) further south in the Newark Basin.
showed that the amount of correction will be different for ages Here, the Orange Mountains Basalt is located below the
calculated using the 238U or 235U decay schemes, due to potential lowermost Jurassic Bulbilimnadia sheni conchostracan zone.
inaccuracies in the two uranium decay constants. The conclusion of Kozur and Weems (2005) that the Rhaetian is
If there is a direct correlation between the end-Triassic ex- partly missing in the Newark basin is corroborated by the
tinction and CAMP volcanism, a U–Pb age for the TJB would magnetostratigraphic work of Gallet et al. (2007). Kozur and
therefore be significantly older than the 40Ar–39Ar age of the Weems' (2005) correlation implies that the first CAMP lava
coeval CAMP rocks. A more accurate correlation is only pos- flows in the Newark Basin already started below the TJB. We
sible using U–Pb age determinations of CAMP volcanics. Such also note that the CAMP basalts in Morocco have been
data are available from volcanic units within the continental considered to be stratigraphically below the NMB and the OMB
Newark Supergroup, such as the Palisades and Gettysburg sills (Knight et al., 2004; Deenen et al., 2007). Therefore, there are
(200.9 ± 1 and 201.3 ± 1 Ma) as well as the basal flow of the not enough radiochronological data available at the moment to
North Mountain Basalt (NMB, 201.7 + 1.4/− 1.1 Ma; Dunning know the precise age of the truely oldest CAMP basalts. This
and Hodych, 1990; Hodych and Dunning, 1992; 201.27 ± needs to be further tested, most importantly by U–Pb dates of
0.06 Ma; Schoene et al., 2006). The NMB is often regarded as similarly high-precision to those reported in Schoene et al.
being the oldest CAMP basalt flow in North America Kozur and (2006) and this study, and measured with the same U–Pb tracer
Weems, 2007; Whiteside et al., 2007), and therefore the precise solutions.
age of 201.27 ± 0.06 Ma Schoene et al. (2006) and our date for Despite difficulties in correlating disparate stratigraphic sec-
the TJB (201.58 ± 0.17 Ma) warrants discussion. tions and the present sparcity of high-precision geochronologic
 U. Schaltegger et al. / Earth and Planetary Science Letters 267 (2008) 266–275 273

dates, several other lines of evidence support a CAMP origin for genera have a broader scope than the ones used in the Lower
the mass extinction event. The synchrony of magmatism and Triassic, mainly because the phylogenetic relationships between
extinction has been corroborated by the palynological investi- the different groups is less well known within that earlier time,
gations of Kuerschner et al. (2007), also showing that the major resulting in a much more splitted taxonomy. However, despite
botanical change occurs at the same time as the marine major all these different aspects we have addressed above, we think
crisis and implying that the continental ecosystem in general that the duration of 1 to 2 m.y. seems to be a reasonable estimate
must have been affected simultaneously. Vertebrate extinction, for Lower Jurassic biotic recovery.
as deduced from tetrapod remains (Olsen et al., 1987) and their
trace fossil record (Silvestri and Szajina, 1993), is coincident 4.3. A revised Lower Jurassic timescale
with the peak in floral turnover (Olsen et al., 2002). It is thus
highly probable that such floral changes were concomitant with Our new results, of course, ask for a modification of the
major climatic perturbations (see Guex et al., 2004). We may Mesozoic timescale published by Ogg (2004) where the TJB is
thus suggest that the mass extinction at the TJB is compati- located at 199.6 Ma, i.e 2 million years younger than the one
ble with a volcanic triggered global biotic crisis. Further high- established in the present paper. The same applies for the
resolution biostratigraphic work such as our ammonoid cor- Hettangian–Sinemurian boundary which was located at 196.5
relations combined with high-precision U–Pb geochronology instead of 199.5 Ma. This is a major new point affecting an im-
volcanic units from well-calibrated stratigraphic sections and portant part of the current age assignments of the early Jurassic
the lowest CAMP basalts will help to prove contemporaneity of stages.
large igneous provinces and major extinction events in Earth's
history. 5. Conclusions

4.2. Duration of the Hettangian stage and tempo of lower We are presenting new biostratigraphic and radio-isotopic
Jurassic post-extinction biotic recovery data from a complete mid-Rhaetian to Sinemurian marine sec-
tion that was recently discovered in the Utcubamba Valley
An important question raised by the short Hettangian du- (northern Peru). Biostratigraphic correlation was carried out
ration is whether 2 m.y. is a typical recovery period? There are by means of ammonites and age information was determined
very few precise numerical dates allowing a comparison with from zircon bearing volcanic tuffs, which yielded precise and
other periods. The best ones are those established recently by accurate 206Pb/238U ages of 201.58 ± 0.17/0.28 Ma (without/
Ovtcharova et al. (2006) for the post-Permian extinction with decay constant uncertainty) for the Triassic–Jurassic boun-
recovery of ammonoids in the Lower Triassic. They found that dary, and 199.53 ± 0.19/0.29 Ma for the Hettangian–Sinemurian
the duration of the recovery interval (Griesbachian, Dienerian boundary. Our new, both precise and accurate U–Pb ages are c.
and Smithian substages), was close to 2 m.y., very similar to the 0.8 to 1.0% older than precise 40Ar/39Ar ages reported for basalts
duration of the ammonoid recovery in the Hettangian found in of the Central Atlantic Magmatic Province (CAMP) in northern
this study. During that first part of the Lower Triassic, if we and southern Americas and northwestern Africa. The fact that
40
follow the compilations published by Tozer (1981) and the work Ar/39Ar ages are often approximately 1% too young when
of Brayard et al. (2006), there are about 60 ammonoid genera, 45 compared to both U–Pb radioisotopic as well as astronomically
being restricted to the Smithian stage, all of them deriving from tuned stratigraphic ages has been repeatedly recorded and is
the genus Ophiceras which contains mainly evolute and smooth interpreted in terms of using an inaccurate 40K decay constant in
forms. During the Hettangian recovery period, about 30 genera Ar–Ar dating.
were developed during a comparable duration, deriving from the Taking the systematic offset between U–Pb and Ar–Ar ages
genus Psiloceras which is homeomorph of the early Triassic into account, our new data allow for the first time a firm con-
Ophiceras (see Guex, 2006). firmation for synchrony between the volcanism of the Central
In other words, if we estimate grossly the rate of recovery as Atlantic Magmatic Province (CAMP) and an important marine
the number of genera per million years, the Hettangian recov- faunal extinction at the Triassic–Jurassic boundary. They also
ery appears to be slightly “slower” than the post-Permian. We provide an excellent basis to estimate the tempo of the biotic
suppose that the difference between the two is due to a greater recovery after end-Triassic extinction. Given the possibility that
environmental instability during the Lower Triassic times (Pruss the precisely dated North Mountain basalt in the Newark basin
et al., 2005; Payne and Kump, 2007), and we believe that such does not reflect the oldest CAMP basalt, we need to continue
instabilities increase the morphological variability of the marine looking for older basalts in other areas of the CAMP, notably in
invertebrates Guex (2006). This, in turn, increases the number Morocco, in order to prove contemporaneity of volcanism and
of morphological taxa without real biological significance. mass extinction at the T/J boundary at the 100 to 500 kyr level.
Furthermore, it should also be noticed that the taxonomic phi-
losophies applied within the two periods are different and Acknowledgements
are not strictly comparable. This could have an influence on
the apparent diversity of the two periods. Our estimate of the We thank M. Senn, M. Chiaradia, for technical help in mass
Hettangian ammonite biodiversity is mainly based on the tax- spectrometry and chemistry. The study was supported by the
onomic concepts of Guex (1995), Guex et al. (2004), where the Swiss National Foundation (JG project 200020-111559, US
274 U. Schaltegger et al. / Earth and Planetary Science Letters 267 (2008) 266–275

project nr. 200020-113387), and by the CNRS project Eclipse. Kuerschner, W.M., Bonis, N., Krystyn, L., 2007. Carbon-isotope stratigraphy and
JG thanks Oscar von Bischhofshausen (Cloudforest Expedi- palynostratigraphy of the Triassic–Jurasic transition in the Tiefengraben
section — Northern Calcareous Alps (Austria). Palaeoecology 244, 257–280.
tions) and Arcenio Balcazar Saldana for their efficient logistic Kuiper, K.F., Hilgen, F.J., Steenbrink, J., Wijbrans, J.R., 2004. 40Ar/39Ar ages of
help in the field, Jose Machare, Victor Benavides and Silvia tephras intercalated in astronomically tuned Neogene sedimentary sequences
Rosas for their scientific support in Peru, Andrea Marzoli and in the eastern Mediterranean. Earth Planet. Sci. Lett. 222, 583–597.
Sandra Kamo for their useful review of the original manuscript. Kuiper, K.F., Wijbrans, J.R., Hilgen, F.J., 2005. Radioisotopic dating of the
Tortonian Global Stratotype Section and Point: implications for intercalibra-
tion of 40Ar/39Ar and astronomical dating methods. Terra Nova 17, 385–397.
References Ludwig, K.R., Isoplot/Ex. V.3, USGS Open-File Repository(2005).
Mattinson, J., 2003. CA (chemical abrasion)-TIMS: high-resolution U–Pb
Bachmann, O., Charlier, B.L.A., Lowenstern, J.B., 2007. Zircon crystallization zircon geochronology combining high-temperature annealing of radiation
and recycling in the magma chamber of the rhyolitic Kos Plateau Tuff damage and multi-step partial dissolution analysis. Eos Transactions AGU,
(Aegean arc). Geology 35, 73–76. Fall Meeting Supplement Abstract V22E-06.
Black, L.P., Kamo, S.L., Allen, C.M., Davis, D.W., Aleinikoff, J.N., Valley, J. Mattinson, J.M., 2005. Zircon U–Pb chemical abrasion (“CA-TIMS”) method:
W., Mundil, R., Campbell, I.H., Korsch, R.J., Williams, I.S., Foudoulis, C., combined annealing and multi-step partial dissolution analysis for improved
2004. Improved 206Pb/238U microprobe geochronology by the monitoring of precision and accuracy of zircon ages. Chem. Geol. 220, 47–66.
a trace-element-related matrix effect; SHRIMP, ID-TIMS, ELA-ICP-MS and Marzoli, A., Renne, P.R., Piccirillo, E.M., Ernesto, M., Bellieni, G., De Min, A.,
oxygen isotope documentation for a series of zircon standards. Chem. Geol. 1999. Extensive 200-million-year-old continental flood basalts of the
205, 115–140. Central Atlantic Magmatic Province. Science 284, 616–618.
Brayard, A., Bucher, H., Escarguel, G., Fluteau, F., Bourquin, S., Galfetti, T., Marzoli, A., Bertrand, H., Knight, K.B., Cirilli, S., Buratti, N., Verati, C.,
2006. The Early Triassic ammonoid recovery: paleoclimatic significance of Nomade, S., Renne, P.R., Youbi, N., Martini, R., Allenbach, K., Neuwerth,
diversity gradients. Palaeogeogr. Palaeoclimatol. Palaeoecol. 239, 374–395. R., Rapaille, C., Zaninetti, L., Bellieni, G., 2004. Synchrony of the Central
Cherniak, D.J., Watson, E.B., 2001. Pb diffusion in zircon. Chem. Geol. 172, 5–24. Atlantic magmatic province and the Triassic–Jurassic boundary climatic and
Courtillot, V.E., Renne, P.R., 2003. On the ages of flood basalt events. C. R. biotic crisis. Geology 32, 973–976.
Geosci. 335, 113–140. Min, K., Mundil, R., Renne12, P.R., Ludwig, K.R., 2000. A test for systematic
Deenen, M.H.L., Reitsma, M.J., Krijgsman, W., Langereis, C.G., van Bergen, errors in 40Ar/39Ar geochronology through comparison with U/Pb analysis
M.J., 2007. The CAMP controversy, new data from the Argana Basin, of a 1.1-Ga rhyolite. Geochim. Cosmochim. Acta 64, 73–98.
Morocco. Geophys. Res. Abstr. 9, 06839. Min, K.W., Renne, P.R., Huff, W.D., 2001. 40Ar/39Ar dating of Ordovician K-
Dunning, G.R., Hodych, J.P., 1990. U/Pb Zircon and Baddeleyite Ages for the bentonites in Laurentia and Baltoscandia. Earth Planet Sci. Lett. 185, 121–134.
Palisades and Gettysburg Sills of the Northeastern United States — implications Min, K.W., Farley, K.A., Renne, P.R., Marti, K., 2003. Single grain (U–Th)/He
for the age of the Triassic–Jurassic Boundary. Geology 18, 795–798. ages from phosphates in Acapulco meteorite and implications for thermal
Gallet, Y., Krystyn, L., Marcoux, J., Besse, J., 2007. New constraints on the history. Earth Planet. Sci. Lett. 209, 323–336.
End-Triassic (Upper Norian -Rhaetian) magnetostratigraphy. Earth Planet. Mundil, R., Pálfy, J., 2005. Triassic–Jurassic time scale and mass extinction: current
Sci. Lett. 255, 458–470. status and new constraints. Geochim. Cosmochim. Acta 69, A320–A321.
Gerstenberger, H., Haase, G., 1997. A highly effective emitter substance for mass Mundil, R., Ludwig, K.R., Metcalfe, I., Renne, P.R., 2004. Age and timing of
spectrometric Pb isotope ratio determinations. Chem. Geol. 136, 309–312. the Permian mass extinctions: U/Pb dating of closed-system zircons. Science
Guex, J., 1995. Ammonites hettangiennes de la Gabbs Valley Range (Nevada). 305, 1760–1763.
Mém. Géol. Lausanne 27, 132. Mundil, R., Pálfy, J., Matzel, J., 2005. New contraints on the timing of the Triassic–
Guex, J., 2006. Reinitialization of evolutionary clocks during sublethal environmental Jurassic extinction and recovery. GSA Annual Meeting, Salt Lake City.
stress in some invertebrates. Earth Planet. Sci. Lett. 242, 240–253. Ogg, J.G., 2004. The Triassic period. In: Gradstein, F.M., Ogg, J.G., Smith, A.G.
Guex, J., Bartolini, A., Atudorei, V., Taylor, D., 2004. High-resolution ammonite (Eds.), A geological time scale. Cambridge University Press, pp. 271–306.
and carbon isotope stratigraphy across the Triassic–Jurassic boundary at New Olsen, P.E., Shubin, N.H., Anders, M.H., 1987. New Early Jurassic Tetrapod
York Canyon (Nevada). Earth Planet. Sci. Lett. 225, 29–41. assemblages constrain Triassic–Jurassic Tetrapod extinction event. Science
Hallam, A., Wignall, P.B., 1999. Mass extinctions and sea-level changes. Earth- 237, 1025–1029.
Sci. Rev. 48, 217–250. Olsen, P.E., Kent, D.V., Sues, H.D., Koeberl, C., Huber, H., Montanari, A.,
Hames, W.E., Renne, P.R., Ruppel, C., 2000. New evidence for geologically Rainforth, E.C., Fowell, S.J., Szajna, M.J., Hartline, B.W., 2002. Ascent of
instantaneous emplacement of earliest Jurassic Central Atlantic magmatic dinosaurs linked to an iridium anomaly at the Triassic–Jurassic boundary.
province basalts on the North American margin. Geology 28, 859–862. Science 296, 1305–1307.
Hillebrandt, A.v., 2000. Die Ammoniten-Fauna des südamerikanischen Hettan- Ovtcharova, M., Bucher, H., Schaltegger, U., Galfetti, T., Brayard, A., Guex, J.,
gium (basaler Jura); Teil 3. Paleontograpica, Abteilung A: Palaeozoologie- 2006. New Early to Middle Triassic U–Pb ages from South China:
Stratigraphie 258, 65–116. calibration with ammonoid biochronozones and implications for the timing
Hodych, J.P., Dunning, G.R., 1992. Did the Manicouagan impact trigger End-of- of the Triassic biotic recovery. Earth Planet. Sci. Lett. 243, 463–475.
Triassic mass extinction. Geology 20, 51–54. Pálfy, J., Mortensen, J.K., Carter, E.S., Smith, P.L., Friedman, R.M., Tipper, H.W.,
Knight, K.B., Nomade, S., Renne, P.R., Marzoli, A., Bertrand, H., Youbi, N., 2000. Timing the end-Triassic mass extinction: first on land, then in the sea?
2004. The central Atlantic magmatic province at the Triassic–Jurassic Geology 28, 39–42.
boundary: paleomagnetic and 40Ar/39Ar evidence from Morocco for brief, Payne, J.L., Kump, L.R., 2007. Evidence from recurrent Early Triassic massive
episodic volcanism. Earth Planet. Sci. Lett. 228, 143–160. volcanism from quantitative interpretation of carbon isotope fluctuations.
Kozur, H.W., Weems, R., 2005. Conchostracan evidence for a late Rhaetian to Earth Planet. Sci. Lett. 256, 264–277.
early Hettangian age for the CAMP volcanic event in the Newark Super- Pruss, S.B., Corsetti, F.A., Bottjer, D.J., 2005. The unusual sedimentary rock
group and a Sevatian (late Norian) age for the immediately underlying beds. record of the Early Triassic; a case study from the southwestern United
Hallesches Jahrb. Geowiss. B27, 21–51. States. Palaeogeogr. Palaeoclimatol. Palaeoecol. 222, 33–52.
Kozur, H.W., Weems, R., 2007. Upper Triassic conchostracan biostratigraphy of Renne, P.R., 2000. 40Ar/39Ar age of plagioclase from Acapulco meteorite and
the continental rift basins of Eastern North America: it's implication for the problem of systematic errors in cosmochronology. Earth Planet. Sci. Lett.
correlating Newark supergroup events with the Germanic basin and the 175, 13–26.
international geologic time scale. In: Lucas, S.G., Spielmann, J.A. (Eds.), The Renne, P.R., Deino, A.L., Walter, R.C., Turrin, B.D., Swisher, C.C., Becker, T.A.,
Global Triassic Bulletin, New Mexico Museum of Natural History and Science, Curtis, G.H., Sharp, W.D., Jaouni, A.R., 1994. Intercalibration of astro-
pp. 137–188. nomical and radioisotopic time. Geology 22, 783–786.
 U. Schaltegger et al. / Earth and Planetary Science Letters 267 (2008) 266–275 275

Renne, P.R., Swisher, C.C., Deino, A.L., Karner, D.B., Owens, T.L., DePaolo, Stacey, J.S., Kramers, J.D., 1975. Approximation of terrestrial lead isotope evo-
D.J., 1998. Intercalibration of standards, absolute ages and uncertainties in lution by a two-stage model. Earth Planet. Sci. Lett. 26, 207–221.
40
Ar/39Ar dating. Chem. Geol. 145, 117–152. Taylor, D., Guex, J., 2002. The Triassic Jurassic System boundary in the John
Rosas, S., 1994. Facies, diagenetic evolution and sequence analysis along a SW- Day Inlier, East-Central Oregon. Or. Geol. 64, 3–28.
NE profile in southern Pucara Basin (Upper Triassic-Lower Jurassic), Cen- Tozer, E.T., 1981. Triassic Ammonoidea: geographic and stratigraphic distribu-
tral Peru. Heidelb. Geowiss. Abh. 80, 1–330. tion. In: House, M.R., Senior, J.R. (Eds.), The Ammonoidea: the evolution,
Schmitz, M.D., Bowring, S.A., 2001. U–Pb zircon and titanite systematics of classification, mode of life and geological usefulness of a major fossil group.
the Fish Canyon Tuff: an assessment of high-precision U–Pb geochronology Systamtic Association Special, vol. 18. Academic Press, London - New York,
and its application to young volcanic rocks. Geochim. Cosmochim. Acta 65, pp. 397–431.
2571–2587. Villeneuve, M., Sandeman, H.A., Davis, W.J., 2000. A method for intercalibration
Schoene, B., Crowley, J.L., Condon, D.J., Schmitz, M.D., Bowring, S.A., 2006. of U–Th–Pb and 40Ar/39Ar ages in the Phanerozoic. Geochim. Cosmochim.
Reassessing the uranium decay constants for geochronology using ID-TIMS Acta 64, 4017–4030.
U–Pb data. Geochim. Cosmochim. Acta 70, 426–445. Whiteside, J.H., Olsen, P.E., Kent, D.V., Fowell, S.J., Et-Touhami, M., 2007.
Sepkoski, J.J., 1994. Extinction and the fossil record. Geotimes 39, 15–17. Synchrony between the Central Atlantic magmatic province and the Triassic–
Silvestri, S.M., Szajina, M.J., 1993. Biostratigraphy of vertebrate footprints in Jurassic mass-extinction event? Palaeogeogr. Palaeoclimatol. Palaeoecol. 244,
the Late Triassic section of the Newark basin, Pennsylvania: reassessment of 345–367.
stratigraphic ranges. In: Lucas, S.G., Morales, M. (Eds.), The nonmarine
Triassic 3, New Mexico Museum of Natural History and Science Bulletin.