Isotopic Evidence for Massive Oxidation of Organic Matter Following the Great Oxidation Event

See allHide authors and affiliations

Science  23 Dec 2011:
Vol. 334, Issue 6063, pp. 1694-1696
DOI: 10.1126/science.1213999


The stable isotope record of marine carbon indicates that the Proterozoic Eon began and ended with extreme fluctuations in the carbon cycle. In both the Paleoproterozoic [2500 to 1600 million years ago (Ma)] and Neoproterozoic (1000 to 542 Ma), extended intervals of anomalously high carbon isotope ratios (δ13C) indicate high rates of organic matter burial and release of oxygen to the atmosphere; in the Neoproterozoic, the high δ13C interval was punctuated by abrupt swings to low δ13C, indicating massive oxidation of organic matter. We report a Paleoproterozoic negative δ13C excursion that is similar in magnitude and apparent duration to the Neoproterozoic anomaly. This Shunga-Francevillian anomaly may reflect intense oxidative weathering of rocks as the result of the initial establishment of an oxygen-rich atmosphere.

The generally high values of carbon isotope ratio (δ13C) [~+10 per mil (‰)] during the Paleoproterozoic Lomagundi-Jatuli (LJ) event [2200 to 2060 million years ago (Ma) (13)] indicate that rates of organic matter deposition exceeded that of oxidative weathering, reflecting high rates of net oxygen production during what has come to be known as the Great Oxidation Event [GOE (4, 5)]. The swings to markedly low δ13C values in the Neoproterozoic indicate excess organic matter oxidation over burial in sediments and atmospheric O2 depletion. The most notable of these, the globally distributed Shuram-Wonoka anomaly (6), is characterized by a negative shift in δ13C of sedimentary carbonates of between –8 to –18‰. However, covariation between oxygen isotope ratio (δ18O) and δ13C in many Shuram-Wonoka carbonate sections, together with the absence of a parallel δ13C excursion in contemporaneous organic matter, has led some to conclude that the Shuram-Wonoka anomaly is an artifact of interaction with fluids in the subsurface, either in a subaerial weathering environment where groundwaters have low δ18O values typical of rainfall and are enriched in 12C through the decomposition of organic matter in soils or in a deep burial diagenetic environment at elevated temperatures (7, 8). There has been no evidence of a large negative δ13C anomaly in the Paleoproterozoic (9), but sampling of this earlier interval has been relatively limited.

We sampled carbonate rocks and organic-carbon–bearing shales from the Zaonega Formation (ZF) in FAR-DEEP (Fennoscandia Arctic Russia–Drilling Early Earth Project) drill cores 12A and 12B from the Paleoproterozoic Onega Basin on the southeastern margin of the Fennoscandian Shield (10). FAR-DEEP recovered more than 3500 m of volcanic and sedimentary rocks from the Paleoproterozoic Era of the Kola Peninsula and Karelian regions of northwestern Russia. The Onega Basin sedimentary and volcanic rocks accumulated in a marine basin within a rifted active continental margin. This succession includes shungite deposits, a noncrystalline and nongraphitized form of carbon representing petrified oil (11), as well as enigmatic organosiliceous rocks and petrified oil fields with up to 99 weight % (wt. %) C. Overlying and underlying formations constrain the age the ZF between 1980 ± 27 Ma (mean ± 1 SD) and 2090 ± 70 Ma (1214).

In the lower part of the succession (below 250 m), the δ13C of organic matter has values typical of Proterozoic shales (~–25‰), whereas the concretionary and vein calcites in the rhythmically bedded, turbiditic graywacke-siltstone-mudstone that dominates the sedimentary rocks of this part of the ZF have very low δ13C and δ18O values, likely reflecting diagenetic and metamorphic rather than seawater conditions (Fig. 1). This interpretation is supported by the relatively small isotopic difference between carbonate and organic matter (Δ13Ccarb-org), averaging 13.8 ± 2.1‰ (mean ± 1 SD) for carbonates below 250 m compared with 25.3 ± 2.7‰ for carbonates above 250 m in the section [supporting online material (SOM) text].

Fig. 1

(Left to right) Lithostratigraphy; carbon and oxygen isotope compositions of carbonates (including samples inferred to have been affected by diagenesis, designated as altered); carbon isotope composition of organic matter, including bulk, acid-insoluble (carbonaceous), and migrated fractions; and nitrogen isotopic composition of both combined mineral and carbonaceous (bulk) and acid-insoluble carbonaceous materials from FAR-DEEP cores 12A and 12B. Samples that are clearly migrated pyrobitumens are indicated with black symbols in the third and fourth panels. VPDB, Vienna Pee Dee belemnite.

A prominent, two-step negative δ13C excursion of about –14 ‰ occurs from 220 to 100 m core depth in carbonates, bulk organic matter, and HF/HCl insoluble organic matter (carbonaceous material including kerogen and pyrobitumen). The first step is smaller, about –4‰, and occurs over 20 m of core (from 220 to 200 m). The second is larger, ~–10‰, again over about 20 to 30 m of section (from 130 to 100 m). Near the top of the core interval, there is a return to somewhat less negative δ13C in both carbonates and organic matter. δ18O varies over a relatively small range (±1‰) within the interval of the negative anomaly in δ13C and with no trend up section. δ18O values are at the lower envelope of the Proterozoic range of values (15), indicating substantial diagenetic or metamorphic alteration of the oxygen isotope composition of the carbonates. This alternation apparently did not significantly affect the carbon isotope composition of the carbonates.

Nitrogen isotope values (δ15N) of bulk sediment are low (2 to 6‰) below 300 m core depth and even lower in the acid insoluble (carbonaceous) fraction, ranging between 0 and 2‰. From 300 to 215 m, the δ15N of the bulk sediment and carbonaceous material δ15N increases markedly to ~10 and 5‰, respectively. Bulk sediment δ15N remains generally elevated, but, at ~130 m, carbonaceous material returns to near pre-excursion values (1 to 3‰). The isotopic difference between bulk rock and carbonaceous δ15N is likely the result of a combination of primary environmental signals recorded in the carbonaceous material and the incorporation of isotopically heavy diagenetic and metamorphic ammonium into clays (16) (SOM text).

In contrast to the Shuram-Wonoka anomaly, the ZF reveals parallel δ13C excursions in carbonates and organic matter but no change in carbonate δ18O, suggesting it is not primarily a result of diagenetic alteration. We interpret the carbon isotope anomaly to reflect the geologically rapid oxidation of a massive amount of ancient organic matter. The source of organic matter could have been sedimentary material deposited during the preceding LJ interval of elevated δ13C (13). Subsequent exposure and oxidation may have occurred during uplift and rifting associated with the breakup of a Paleoproterozoic supercontinent (17). Oxygen-rich groundwaters would have interacted with kerogen in shales, driving massive oxidation and release of CO2 and generating negative δ13C excursions, just as they apparently led to supergene iron-ore enrichment worldwide (17). Alternatively, oxidation of global anoxic or euxinic basins (18) enriched in dissolved organic matter (19) could explain the excursions in both the N and C isotopes, especially if the event proves to be considerably shorter than the 110-million-year maximum duration for deposition of the ZF (1214) and thus required a proportionately smaller transfer of 13C-depleted organic carbon to the inorganic pool.

The dynamic response of the nitrogen cycle in the ZF is consistent with increased availability of O2 and oxidation of reduced N-species (NH4+ and organic N) in the Onega Basin water column and is similar to the response documented in Archean sequences (16, 20) during apparent transient oxygenation events. Carbonaceous-matter δ15N values are initially low, likely reflecting a biological N2-fixation source for nutrient nitrogen. Transition from a largely anoxic Onega Basin water column to one that is oxidized in surface waters would have allowed for enhanced microbial conversion of ammonium to nitrite and nitrate (nitrification), and nitrate reduction and anaerobic ammonium oxidation at depth. The coupled processes of nitrification and nitrate reduction fractionate N isotopes, with 14N preferentially incorporated into N2 or N2O, leading to 15N-enrichment of the residual dissolved inorganic nitrogen pool (21). Thus, the observed δ15N increase of 5‰ from 300 to 215 m core depth in carbonaceous material reflects an expansion of the redox cycling of nitrogen. The ~3‰ decrease in δ15N values above 130 m may have been a consequence of O2 consumption during organic matter oxidation resulting in the crossing of basinal redox thresholds and limitation of nitrogen redox cycling. Additionally, the drop in δ15N could also, in part, reflect an intensification of nitrogen fixation in response to fixed-nitrogen deficits.

The apparently correlative, organic-C-rich shales of the 2083 ± 6 to 2050 ± 30 Ma Francevillian Series of Gabon (22, 23) display a negative δ13C excursion that, after accounting for basinal differences in productivity or thermal maturity leading to a several permil offset, is strikingly similar in overall magnitude and detailed stratigraphic trend to that of the ZF rocks analyzed here (Fig. 2). Moreover, a recent analysis of the carbonate rocks from the Francevillian documents an 8 to 10‰ negative excursion in δ13C (24) that corresponds to the second phase of decline in δ13C of the organic matter. We therefore suggest that this “Shunga-Francevillian” excursion was global in extent.

Fig. 2

Comparison of bulk organic matter δ13C values from the Onega Basin (Shunga) of Fennoscandia (FAR-DEEP cores 12A and 12B) and the Francevillian Basin of Gabon (22). The depth axis for the Shunga samples is as in Fig. 1. The Francevillian depth axis is in relative stratigraphic position, as reported by Gauthier-Lafaye and Weber (22), and aligned with the Shunga profile by matching the two prominent steps in δ13C. FB, FC, and FD are recognized stratigraphic units in the Franceville Formation.

The Shunga-Francevillian anomaly may represent the oxidative recycling of a fraction of the organic matter sequestered during the preceding LJ high-δ13C interval, triggered by the attainment of atmospheric and groundwater oxygen levels sufficient [~1% of modern (25)] to mobilize uranium (leading to the Oklo natural fission reactors in the Francevillian), generate supergene iron ores (17), and oxidize fossil organic matter during weathering. The well-established cessation of mass-independent fractionation of sulfur isotopes (S-MIF) ~2400 Ma has led to the notion that the GOE was an abrupt event occurring in the early Paleoproterozoic (26). A recent quantitative reanalysis of oxygen indicators preserved in Paleoproterozoic paleosols (27) indicates instead that oxygen levels rose gradually in the early Paleoproterozoic but more rapidly between 2100 and 2000 Ma, achieving a value that was ~1% of modern. Evidence presented here supports the conclusion that the GOE played out over hundreds of millions of years, gradually crossing the low atmospheric O2 threshold for pyrite oxidation {10−8 to 10−5 of the present atmospheric level [PAL (28, 29)} by ~2500 Ma and the loss of S-MIF [~10−5 PAL (30)] by ~2400 Ma and then increasing at an ever-increasing rate through the Paleoproterozoic, achieving levels ~1% PAL by 2000 Ma (27).

Supporting Online Material

Materials and Methods

SOM Text

Figs. S1 to S4

Tables S1 and S2

References (3151)

References and Notes

  1. Materials and methods are available on Science Online.
  2. Acknowledgments: This work was supported by the International Continental Drilling Program and by NSF, NASA Astrobiology Institute, Natural Environment Research Council grant NE/G00398X/1, Norwegian Research Council grant 191530/V30, and Geological Survey of Norway. Thanks to the FAR-DEEP drilling team, especially D. V. Rychanchik and A. E. Romashkin, Karelian Science Centre, for assistance during the drilling and archiving operations. The data reported in this paper are tabulated in the SOM.
View Abstract

Navigate This Article