Atmospheric Carbon Injection Linked to End-Triassic Mass Extinction

See allHide authors and affiliations

Science  22 Jul 2011:
Vol. 333, Issue 6041, pp. 430-434
DOI: 10.1126/science.1204255


The end-Triassic mass extinction (~201.4 million years ago), marked by terrestrial ecosystem turnover and up to ~50% loss in marine biodiversity, has been attributed to intensified volcanic activity during the break-up of Pangaea. Here, we present compound-specific carbon-isotope data of long-chain n-alkanes derived from waxes of land plants, showing a ~8.5 per mil negative excursion, coincident with the extinction interval. These data indicate strong carbon-13 depletion of the end-Triassic atmosphere, within only 10,000 to 20,000 years. The magnitude and rate of this carbon-cycle disruption can be explained by the injection of at least ~12 × 103 gigatons of isotopically depleted carbon as methane into the atmosphere. Concurrent vegetation changes reflect strong warming and an enhanced hydrological cycle. Hence, end-Triassic events are robustly linked to methane-derived massive carbon release and associated climate change.

The end-Triassic mass extinction (ETME) [~201.4 million years ago (1)], one of the five major extinction events of the Phanerozoic (2), is marked by up to 50% marine biodiversity loss and major terrestrial ecosystem changes (25). This event closely matches a distinct and globally observed negative carbon-isotope excursion (CIE) in δ13CTOC records (TOC: total organic carbon) (6, 7) and a potential fourfold increase in atmospheric CO2 concentrations (8). Previously, the end-Triassic C-cycle perturbation has been attributed to large-scale carbon release caused by a major volcanic episode, with emplacement of the Central Atlantic Magmatic Province (CAMP) during the break-up of Pangaea. However, deposition of this large igneous province continued for at least ~600 ky (ky, thousand years) (9, 10), much longer than the ~20- to 40-ky duration of the end-Triassic extinction event (11, 12). The magnitude of the observed negative CIE also varies largely between different geological basins, possibly due to changes in source and preservation of the sedimentary organic matter. These observations question the reality of an end-Triassic global carbon cycle turnover and its potential causal relationship to the extinction event. We determined compound-specific C-isotope records from the western Tethys Ocean [including the global stratotype section and point (GSSP) for the base of the Jurassic], which span the ETME. Changes in the stable carbon isotopic composition of long-chain n-alkanes, derived from epicuticular plant waxes, directly reflect changes in the carbon isotopic composition of CO2 in the atmosphere. Furthermore, the δ13C of these organic molecules is, depending on burial history, unaffected by diagenetic alteration of organic matter (13). Hence, this allows accurate reconstruction of the end-Triassic C-cycle perturbation in a biostratigraphically well-constrained framework.

Sediments for this study are from an upper Rhaetian (latest Triassic) interval of 54 cm in the Kuhjoch and 52 cm in the Hochalplgraben outcrops (7). These sediments were deposited in the intraplatform Eiberg Basin at the continental margin of the western Tethys Ocean (14). The studied interval in both sections directly succeeds the transition from limestones of the Kössen Formation (Fm) to marls of the Kendlbach Fm and coincides with marine and terrestrial extinctions and assemblage changes (Fig. 1). Long-chain n-alkanes (C23 to C35) are preserved in these >200-million-year-old, but thermally immature (15), sediments and originate from higher-plant leaf waxes, with moderate odd-over-even carbon number predominance (16). The weighted-average of the stable carbon isotopic composition of the odd-numbered C23 to C35 n-alkanes shows a ~8.5 per mil (‰) negative excursion from Rhaetian base values of ~−29‰ (Fig. 2). The observed negative excursion in compound-specific C-isotope data strongly suggests 13C depletion of the terrestrial higher-plant carbon reservoir, and hence is sound evidence for end-Triassic atmospheric-13C depletion coinciding with the ETME. This negative CIE in our molecular records is >2.5‰ stronger than in δ13CTOC records from Austria and the UK (6, 7) and 5.5‰ stronger than previously assumed for the modeled end-Triassic negative CIE (17). It is also two times as large as observed in the Newark-Hartford basin compound-specific δ13C record (10).

Fig. 1

End-Triassic mass extinction event coincides with δ13CTOC negative excursion. The δ13CTOC record from Kuhjoch (7), the GSSP for the base of the Jurassic (47°41'15"N, 13°21'30"E). The end-Triassic negative CIE closely matches marine extinctions and continental and marine assemblage changes (3, 6, 14, 34) (LO, last occurrence; FO, first occurrence). The negative CIE stratigraphically coincides with the onset of CAMP emplacement (11).

Fig. 2

N-alkane biomarker C-isotope and climate proxy records. The end-Triassic mass-extinction interval is marked by a distinct negative CIE in δ13Cn-alkane records from (A) Kuhjoch and (B) Hochalplgraben (47°28'20"N, 11°24'42"E). The ~8.5‰ magnitude of the negative CIE in the weighted-average δ13C(C23-C25-C27-C29-C31-C33-C35) record of Hochalplgraben is ~3.5‰ larger than observed in the δ13CTOC record of the same section. The black curves show individual n-alkane C-isotope records for both sections. The shaded areas show the bandwidth of δ13CC23 to δ13CC35 values per sample, in both sections. The extracted amount of n-alkanes from the lower three δ13CTOC samples from Kuhjoch was too low for reliable δ13Cn-alkane measurements. The (onset of the) observed negative CIE coincides in both sections with increased continental temperatures and an enhanced hydrological cycle, based on statistical analyses of terrestrial palynomorph distributions (16).

Major vegetation changes could have modified the magnitude of the end-Triassic negative CIE in compound-specific C-isotope records because of differential carbon isotopic fractionation among plant groups. Palynological data show a strong increase of Cheirolepidiaceaen (Classopollis meyeriana) conifer pollen concurrent with the onset of the observed negative CIE (18). A rapid transition from a mixed angiosperm-conifer flora to a purely angiosperm flora at the Paleocene-Eocene Thermal Maximum (PETM), however, amplified the observed negative CIE by 1 to 2‰ (19). Also, modern conifer-derived n-alkanes are relatively enriched in 13C because of a lower stomatal conductance relative to other plant groups (20). Assuming that physiological mechanisms in Mesozoic conifers were similar to those in conifers of the Cenozoic and today, the observed 8.5‰ end-Triassic higher-plant n-alkane δ13C excursion may thus even be dampened relative to atmospheric values. The smaller magnitude of this excursion in the Newark-Hartford δ13Cn-alkane record may be due to a lower sampling resolution during the ETME (10). δ13Cn-alkane values in this record may also be affected by thermal degradation of n-alkanes or mixing of organic sources, as suggested by n-alkane chain-length distribution and large variations between δ13Cn-alkane, δ13CTOC, and δ13Cwood (10). Nevertheless, the similar trend in compound-specific C-isotope records during the ETME in North America (10) and the Western Tethys Ocean (this study) strongly suggests major end-Triassic atmospheric 13C depletion.

The end-Triassic negative CIE in δ13CTOC records was, based on model calculations, previously ascribed to the release of ~8000 to 9000 gigatons (Gt) of carbon as volcanogenic gaseous CO2 from the CAMP with subsequent destabilization of ~5000 Gt of carbon from the methane-hydrate reservoir (17). Also, recent Pco2 (partial pressure of CO2) estimates from pedogenic carbonates, succeeding subsequent CAMP basalt units in eastern North American rift basins, suggest major basaltic CO2 outgassing (21). However, this modeled release of carbon from two reservoirs results in a ~3‰ depletion of the exogenic carbon pool only. The onset of the observed negative CIE probably occurred within ~10 to 20 ky, based on the astronomically constrained ~20- to 40-ky duration of the complete event (1012). This implies a rapid release of large amounts of isotopically depleted carbon to the end-Triassic atmosphere, coincident with the ETME. The ~8.5‰ magnitude and the short (~20 to 40 ky) duration of the observed negative CIE do not, therefore, match with CAMP-related CO2 release as the main source for isotopically depleted carbon. A simple mass balance calculation using end-Triassic boundary conditions (16, 17) shows that ~8.5‰ atmospheric-13C depletion can also be explained by the release of ~12,000 Gt of carbon as methane from clathrates (with δ13C values of −60‰). Alternatively, thermal metamorphism of subsurface organic-rich strata, associated with sill intrusions and flood basalt emplacement, was proposed as a major source of 13C-depleted thermogenic methane to the end-Permian, Rhaetian, Toarcian, and Eocene atmosphere (2225). It possibly also contributed to the magnitude of the end-Triassic C-cycle perturbation at the ETME (23, 26). Carbon release purely from this source (with δ13C values of −35 to −50‰) would involve an input of possibly as much as 38,000 Gt. None of these mechanisms is mutually exclusive, and all three may have contributed to the release of 13C-depleted carbon at the ETME, with thermogenic methane and gaseous CO2 release from CAMP initiating a positive feedback in the global exogenic carbon cycle, causing the release of methane from clathrates. The relative contribution of these end members for 13C-depleted carbon release is yet unknown. Continued carbon release from primary outgassing or contact metamorphism during CAMP emplacement (21) likely caused prolonged early Jurassic Pco2 increase and decreased δ13CTOC values (12). However, given the duration and magnitude of the observed end-Triassic negative CIE, a strong contribution from the methane-clathrate reservoir may be likely.

The injection of ~12,000 to 38,000 Gt of carbon during the ETME probably had a profound impact on global climate. Statistical analyses of palynological data spanning this time interval (16) show a strong warming event and an enhanced hydrological cycle directly coinciding with the onset of the negative CIE (Fig. 2). Marine and terrestrial assemblage changes and extinctions directly coincide with the onset of this warming event (Fig. 2). This suggests a strong causal relationship between massive carbon release, associated climate change, and terrestrial and marine ecosystem turnover. Terrestrial ecosystem changes have likely been further enhanced by the release of toxic gases (e.g., sulfur dioxide) (26). Massive carbon release to the atmosphere and subsequently the ocean also has strong effects on ocean acidification (27). At the ETME, this resulted in decreased carbonate precipitation and possibly dissolution, leading to reduced marine ecosystem stability and extinctions (28).

Similar events of rapid carbon release to the atmosphere [e.g., at the PETM (29) and in the early Toarcian (30)] suggest a ~100-ky recovery period for the δ13C composition of exchangeable carbon reservoirs, in line with the residence time of carbon in the exogenic carbon pool (31). A relatively short duration of the observed negative CIE at the ETME may be attributed to enhanced surface-ocean productivity and increased burial of carbon at the continental margins (18). Diminished background release of isotopically light carbon, due to changes in the gas-hydrate capacitor (31), may have further enhanced rapid recovery of the δ13C signal of the exogenic carbon pool. Furthermore, a sea-level lowstand at the ETME possibly enhanced erosion of the extensive Triassic carbonate ramps and dilution of the exchangeable carbon reservoirs with relatively enriched carbon.

The ETME interval, with rapid and large-scale carbon release, may be regarded as a natural deep-time analog to today’s anthropogenic carbon emissions. Cumulative anthropogenic carbon release of >5000 Gt (27) likely will enhance greenhouse warming by several degrees (32) and substantially lower oceanic pH values (27). Earth’s biosphere also is projected to experience major disruption of ecosystems, with associated loss of biodiversity (33). A direct link between massive carbon release and the ETME suggests that modern-day ecosystems could experience a further loss in biodiversity, not only by habitat reduction but also by carbon release–driven rapid climate changes.

Supporting Online Material

Materials and Methods

Figs. S1 and S2

Tables S1 and S2


References and Notes

  1. Supporting Online Material is available at Science Online.
  2. Acknowledgments We thank G. Nobbe for lab assistance with stable-isotope measurements, and H. Visscher and A. Sluijs for suggestions and comments during the initial drafting of the manuscript. We also thank H. Svensen and three anonymous reviewers for suggestions and comments during the review of this paper. W.M.K., M.R., and N.R.B. acknowledge funding from the Utrecht University High Potential program. This contribution is publication 20110601 of the Netherlands Research School of Sedimentary Geology.

Stay Connected to Science

Navigate This Article