Content uploaded by Jason R Ali
Author content
All content in this area was uploaded by Jason R Ali
Content may be subject to copyright.
Content uploaded by Jason R Ali
Author content
All content in this area was uploaded by Jason R Ali
Content may be subject to copyright.
Content uploaded by Jason R Ali
Author content
All content in this area was uploaded by Jason R Ali
Content may be subject to copyright.
2. Materials and methods are available as supporting
material on Science Online.
3. A. Grantz, R. A. Scott, S. S. Drachev, T. E. Moore, “Maps
showing the sedimentary successions of the Arctic Region
(58°-64° to 90° N) that may be prospective for
hydrocarbons,”American Association of Petroleum
Geologists GIS-UDRIL Open-File Spatial Library, 2009,
http://gisudril.aapg.org/gisdemo/.
4. U.S. Geological Survey, “Circum-Arctic Resource Appraisal
(north of the Arctic Circle) assessment units GIS data,”
2009, http://energy.usgs.gov/arctic/.
5. USGS World Assessment Team, “U.S. Geological Survey
world petroleum assessment 2000: Description and
results,”USGS Digital Data Series DDS60, 2000,
http://pubs.usgs.gov/dds/dds-060/.
6. R. R. Charpentier, T. R. Klett, E. D. Attanasi, “Database
for assessment of assessment unit-scale analogs
(exclusive of the United States),”U.S. Geological Survey
Open-File Report 2007-1404, 2008, http://pubs.usgs.
gov/of/2007/1404/.
7. R. R. Charpentier, paper presented at the 2008 American
Association of Petroleum Geologists Annual Convention
and Exhibition, San Antonio, TX, 22 April 2008.
8. J. H. Schuenemeyer, “Procedures for aggregation used in
the Circum-Arctic Resource Appraisal,”U.S. Geological
Survey Open File Report, in press.
9. British Petroleum (BP) Public Limited Company, “BP
statistical review of world energy 2008,”London, 2008,
http://www.bp.com/statisticalreview.
10. We thank P.I. McLabe and R.S. Bishop for their comments
on an earlier version of the manuscript.
Supporting Online Material
www.sciencemag.org/cgi/content/full/324/5931/1175/DC1
Materials and Methods
Table S1
8 December 2008; accepted 16 March 2009
10.1126/science.1169467
Volcanism, Mass Extinction, and
Carbon Isotope Fluctuations in the
Middle Permian of China
Paul B. Wignall,
1
*Yadong Sun,
2
David P. G. Bond,
1
Gareth Izon,
3
Robert J. Newton,
1
Stéphanie Védrine,
1
Mike Widdowson,
3
Jason R. Ali,
4
Xulong Lai,
2
Haishui Jiang,
2
Helen Cope,
5
Simon H. Bottrell
1
The 260-million-year-old Emeishan volcanic province of southwest China overlies and is
interbedded with Middle Permian carbonates that contain a record of the Guadalupian mass
extinction. Sections in the region thus provide an opportunity to directly monitor the relative
timing of extinction and volcanism within the same locations. These show that the onset of
volcanism was marked by both large phreatomagmatic eruptions and extinctions amongst
fusulinacean foraminifers and calcareous algae. The temporal coincidence of these two phenomena
supports the idea of a cause-and-effect relationship. The crisis predates the onset of a major
negative carbon isotope excursion that points to subsequent severe disturbance of the
ocean-atmosphere carbon cycle.
The temporal link between mass extinc-
tion events and large igneous province
volcanism is one of the most intriguing
relationships in Earth’s history, with the end-
Permian extinction–Siberian Traps association
being the most celebrated (1,2), but the causal
link is far from resolved. A major problem is
that the site of volcanism can rarely be directly
correlated with the marine extinction record
(3), and so comparison can only be achieved
with the use of geochronological bio- and che-
mostratigraphic correlation techniques, with
their inherent timing inaccuracies. To clarify
some of these relations, we have studied the
Emeishan flood basalt province in southwest
China, where Middle Permian platform lime-
stones pass up into a volcanic pile with inter-
bedded limestones. These record both a marine
extinction record and a major C isotope excur-
sion. Thus, we were able to document multiple
phenomena associated withthe Middle Permian
mass extinction within the same geological
sections.
Middle Permian (Guadalupian) platform
carbonate rocks of the Maokou Formation are
widespread throughout south China. In western
Guizhou, southern Sichuan, and Yunnan Prov-
inces they pass laterally into the flows of the
Emeishan large igneous province (Fig. 1). The
original size of the province is difficult to esti-
mate because much has been eroded (scattered
outcrops of contemporaneous volcanic rocks
are found up to 300 km from the main sections,
Fig. 1), but its main outcrops cover 2.5 × 10
5
km
2
in southwest China; the original volume
was probably substantially less than 1 × 10
6
km
3
(4). Despite their relatively small size, the co-
incidental timing of the Emeishan eruptions
with the Guadalupian mass extinction has led to
suggestions that they may be implicated in this
environmental calamity (2,5).
Sections from both within the volcanic prov-
ince and around its margins record a prolonged
phase of stable carbonate platform deposition
before its termination by abrupt base-level
changes. To the north of the province, in north-
ern Sichuan, the Maokou limestones are capped
by a karstic surface dated byconodont studies to
1
School of Earth and Environment, University of Leeds, Leeds
LS2 9JT, UK.
2
Faculty of Earth Sciences, China University of
Geosciences, Wuhan, Hubei 430074, China.
3
Department of
Earth and Environmental Science, The Open University, Milton
Keynes MK7 6AA, UK.
4
Department of Earth Sciences,
Pokfulam Road, University of Hong Kong.
5
Department of
Bioengineering, University of Strathclyde, Wolfson Building,
106 Rottenrow, Glasgow G4 0NW, UK.
*To whom correspondence should be addressed. E-mail:
p.wignall@see.leeds.ac.uk
Fig. 1. Outcrop map (red)
of the Emeishan large igne-
ous province in southwest
China (4).
www.sciencemag.org SCIENCE VOL 324 29 MAY 2009 1179
REPORTS
on May 28, 2009 www.sciencemag.orgDownloaded from
Fig. 2. Xiong Jia Chang
section (26.5°N, 105.7°E),
40 km west of Zhijin,
western Guizhou Province,
showing the transition
from Maokou Formation
carbonates to volcanics at
the base of the Emeishan
province. Range charts for
foraminifers and calcare-
ous algae show the extinc-
tion to occur immediately
below the first eruptive
unit. These phenomena
also coincide with a sharp
negative carbon isotope
excursion.
29 MAY 2009 VOL 324 SCIENCE www.sciencemag.org1180
REPORTS
on May 28, 2009 www.sciencemag.orgDownloaded from
an interval within the early Capitanian Stage
(6). In contrast, within the outcrop area of the
igneous province, shallow-water carbonate rocks
persisted longer but were abruptly drowned,
whereupon deep-water radiolarian-spiculite chert
became established. Conodont biostratigraphic
control at the Xiong Jia Chang section, in
western Guizhou (Fig. 2) and other sections
(7), shows that the deepening occurred around
the beginning of the Jinogondolella altudaensis
zone in the mid-Capitanian. The brief phase of
deep-water chert deposition was terminated by
Fig. 3. Gouchang sec-
tion, near Ziyun, central
Guizhou, developed 50 km
east of the eastern margin
of the Emeishan province,
showing the coincidence
of the mass extinction
level seen in range trun-
cations of schwagerinids
and calcareous algae.
The post-extinction neo-
schwagerinids are frag-
mented and abraded
specimens interpreted
to be reworked from pre-
extinction strata, but they
may suggest that the
interval of extinction is
longer than depicted. A
major negative shift of
d
13
C isotope values (ana-
lyzed in whole-rock car-
bonate) is seen to occur
after the extinction. Sev-
eral gaps in this otherwise
continuous road section
may reflect the presence
of shale and/or volcanic
ash levels.
www.sciencemag.org SCIENCE VOL 324 29 MAY 2009 1181
REPORTS
on May 28, 2009 www.sciencemag.orgDownloaded from
the onset of Emeishan volcanism. At Xiong Jia
Chang, the basal volcanic rocks consist of a
70-m-thick composite unit comprising aphyric
basalt and interbedded mafic volcaniclastic
rocks. The latter are dominated by vesicular
basalt clasts, many of which display chilled or
glassy margins, which testify to an explosive
style of mafic volcanism. Some clasts show a
fiammé texture, typical of terrestrial emplace-
ment, whereas others are dominated by pala-
gonite-rimmed clasts, indicating subaqueous
emplacement (8). Lithoclasts of platform car-
bonate, spiculite, and isolated marine fossils are
also common (8), indicating that seafloor ero-
sion accompanied emplacement. Such violent
phreatomagmatic-style volcanism is typical of
much of the early to mid-stages of Emeishan
volcanism (9).
The initial phase of volcanism was followed
by the reestablishment of diverse carbonate
facies. Thus, at Xiong Jia Chang, the lowest vol-
canic pile is overlain by deep-water chert, with
euxinic framboid populations (10); this unit is
then abruptly succeeded by carbonate beds be-
longing to the J. prexuanhanensis/J. xuanhanensis
assemblage zone, with a shallow-water micro-
biota (Fig. 2). Intertrappean limestone devel-
opments at other sections include Tubiphytes
sponge reefs up to 30 m thick (11) and steep slope
facies with allodapic limestones (12). The
volcanic interregnum was terminated by the re-
turn of spectacular pyroclastic-phreatomagmatic
volcanism and the deposition of individual mafic
volcaniclastic beds approaching 200 m in thick-
ness (11).
Fossil range data show that the collapse of
carbonate platform deposition, shortly before
the onset of eruptions, coincides with the dis-
appearance of a shallow-water marine biota
dominated by foraminifers and calcareous algae.
Similar platform facies reappear above the basal
volcanic pile, but many of the lost taxa do not
(Fig. 2): The species-level composition of the
calcareous algae shows complete turnover. The
foraminifers also show extinctions, with the loss of
the distinctive keriotheca-walled Schwagerinidae
and Neoschwagerinidae and their replacement
with assemblages consisting of small, long-
ranging lagenides, the miliolinid Hemigordius,
and small fusulinaceans (Codonofusiella and
Reichelina).
This mid-Capitanian age for the mass
extinction is substantially earlier than previous
estimates (13,14). C isotope data (15) from
Xiong Jia Chang reveal a 5 to 6 per mil (‰)
negative shift above the lowest volcanic bed
(Fig. 2). This big excursion postdates the
extinction interval, which is within a phase of
stable values around the J. shannoni/J. altu-
daensis zonal boundary. Similar faunal turnover
is seen in the carbonate platform sections at
Gouchang (Fig. 1), approximately 50 km east of
the Emeishan volcanic margin, and here too it
coincides with stable d
13
C values below a major
negative excursion (Fig. 3). A similar large-
amplitude negative excursion was reported from
the mid-Capitanian in northern Sichuan Prov-
ince, to the north of the Emeishan volcanic
province (6). This excursion is now seen to be
widespread (sites are up to 1000 km apart) and
from several types of depositional environment
(16), which suggests that it is a global signal. A
further large negative excursion has been re-
ported at the end of the Capitanian Stage (17),
indicating that the post-extinction interval was
marked by several large fluctuations.
The close temporal link between the onset of
eruptions and extinction suggests a cause-and-
effect scenario. Cooling and acid rain (caused
by SO
2
effusion and sulfate aerosol formation)
and consequent environmental deterioration are
candidates for this link (18,19). The dominance
of pyroclastic volcanism (rather than more
quiescent-style flood basalt eruptions) in the
initial eruptive stages of the Emeishan province
and the large scale of the flows (30 to 200 m
thick) suggest that such effects are likely to have
been severe. The subsequent negative shift of C
isotope values is too large to be attributed to
relatively heavy volcanic CO
2
(d
13
C=–5‰),
but it may record the release of much lighter
thermogenic C from the site of volcanism (20).
This was in the aftermath of the biotic crisis,
but the high diversity of the post-extinction
biota suggests that the light C flux is not linked
to any prolongation of the environmental stress
that caused the extinction.
Our study of the volcano-sedimentary record
of southwest China reveals that the Middle
Permian marine crisis precisely coincided with
the onset of Emeishan volcanism. This provides
evidence for a potential link between mass ex-
tinction and the eruption of this igneous prov-
ince, although the absolute time scale for the
event is not yet known. The subsequent nega-
tive d
13
C excursion implies that the C cycle
was destabilized for some time after the extinc-
tions, perhaps by C release from thermogenic
sources.
References and Notes
1. V. Courtillot, Evolutionary Catastrophes: The Science
of Mass Extinctions (Cambridge Univ. Press,
Cambridge, 1999).
2. P. B. Wignall, Earth Sci. Rev. 53, 1 (2001).
3. G. Keller, T. Adatte, S. Gardin, A. Bartolini, S. Bajpai,
Earth Planet. Sci. Lett. 268, 293 (2008).
4. J. R. Ali, G. M. Thompson, X.-Y. Song, Y.-L. Wang,
Lithos 79, 475 (2005).
5. M.-F. Zhou et al., Earth Planet. Sci. Lett. 196, 113
(2002).
6. X.-L. Lai et al., Palaeogeogr. Palaeoclimatol. Palaeoecol.
269, 78 (2008).
7. Composite sections around Pingdi, on the
Guizhou-Yunnan border, show the development of
Maokou Formation platform carbonates overlain by
deeper-water radiolarian cherts and slumped
carbonates, with a conodont fauna spanning the
J. shannoni/ J. altudaensis zones. These are overlain
by the lowest flood basalts of the Emeishan
volcanic pile.
8. See supporting material on Science Online.
9. I. Ukstins Peate, S. E. Bryan, Nat. Geosci. 1, 625
(2008).
10. Pyrite framboid populations from deep-water cherts
in the basal J. prexuanhanensis/J. xuanhanensis
assemblage zone at Xiong Jia Chang are dominated
by framboids with a mean diameter of less than
7mm, a characteristic of populations formed in
anoxic bottom water (21).
11. A section on the eastern access road to Lugu Lake
(27°41.711′N, 100°58.788′E) in northern Yunnan
shows the lower Emeishan volcanic pile to be dominated
by thick, mafic volcaniclastic flows. These are overlain
by a 30-m-thick Tubiphytes sponge-cement reef, a further
mafic volcaniclastic unit approaching 200 m thick,
another limestone approximately 20 m thick, and finally
a continuous section of basalt lavas, in excess of 1 km
thick. These contain flows showing well-developed
pillows.
12. The Pingchuan section (27°40.779′N, 101°53.104′E),
a roadside section developed immediately to the east of
the small town of Pingchuan (southern Sichuan),
shows the main succession of Emeishan flood basalts to
be underlain by alternations of clast-supported breccia
beds (containing clasts of Maokou limestone, basalt,
and dacite) and laminated cherty micrites yielding
goniatites.
13. Y.-G. Jin et al., Episodes 29, 253 (2006).
14. K. Kaiho et al., Palaeogeogr. Palaeoclimatol. Palaeoecol.
223, 172 (2005).
15. Carbonate-C isotopes were measured at the stable
isotope laboratory at the University of Leeds on CO
2
generated by the addition of anhydrous phosphoric acid
to around 20 mg of powdered whole rock in a vacuum
(22). Values are corrected with standard methods
(23) and reported relative to the Vienna Pee Dee
belemnite (VPDB) standard. The analytical precision
for this analysis, based on replicate analyses of an
in-house strontium carbonate standard, is estimated
at 0.1‰.
16. The negative d
13
C excursion at Xiong Jia Chang
(analyzed in whole-rock carbonate) occurs within a
section showing abrupt deepening from platform
carbonate to deep-water, calcareous chert deposition
associated with the development of the Zhijin Basin, a
north-south rift structure developed immediately
before volcanism (24). The excursion at Gouchang
occurs within a shallow-water platform carbon ate
succession. A potential third example of the ex cursion is
seen in the upper part of unit 2 of the Maokou
Formation at the Chaotian secti on in northern
Sichuan, where it occurs within lagoonal carbonates
on the southern margin of the North Yangtze
Basin (6).
17. W. Wang, C.-Q. Cao, Y. Wang, Earth Planet. Sci. Lett.
220, 57 (2004).
18. S. Self, M. Widdowson, T. Thordarson, A. E. Jay, Earth
Planet. Sci. Lett. 248, 517 (2006).
19. S. Self, S. Blake, K. Sharma, M. Widdowson, S. Sephton,
Science 319, 1654 (2008).
20. G. J. Retallack, A. H. Jahren, J. Geol. 116, 1 (2008).
21. P. B. Wignall, R. Newton, Am. J. Sci. 298, 537
(1998).
22. J. M. McCrea, J. Chem. Phys. 18, 849 (1950).
23. H. Craig, Geochim. Cosmochim. Acta 12, 133
(1957).
24. B. He, Y.-G. Xu, Y.-M. Wang, Z.-Y. Luo, J. Geol. 114, 117
(2006).
25. We thank for funding the Natural Environment
Research Council (grant no. NE/ D011558/1), Natural
Science Foundation of China (grants nos. 40872002,
406210022, and 40232025), Chinese State
Administration of Foreign Experts Affairs
(grant B08030), and Hong Kong Research Grant
Council (grant no. HKU700204).
Supporting Online Material
www.sciencemag.org/cgi/content/full/324/5931/1179/DC1
Fig. S1
6 February 2009; accepted 26 March 2009
10.1126/science.1171956
29 MAY 2009 VOL 324 SCIENCE www.sciencemag.org1182
REPORTS
on May 28, 2009 www.sciencemag.orgDownloaded from