Report

A Stratified Redox Model for the Ediacaran Ocean

See allHide authors and affiliations

Science  02 Apr 2010:
Vol. 328, Issue 5974, pp. 80-83
DOI: 10.1126/science.1182369

Abstract

The Ediacaran Period (635 to 542 million years ago) was a time of fundamental environmental and evolutionary change, culminating in the first appearance of macroscopic animals. Here, we present a detailed spatial and temporal record of Ediacaran ocean chemistry for the Doushantuo Formation in the Nanhua Basin, South China. We find evidence for a metastable zone of euxinic (anoxic and sulfidic) waters impinging on the continental shelf and sandwiched within ferruginous [Fe(II)-enriched] deep waters. A stratified ocean with coeval oxic, sulfidic, and ferruginous zones, favored by overall low oceanic sulfate concentrations, was maintained dynamically throughout the Ediacaran Period. Our model reconciles seemingly conflicting geochemical redox conditions proposed previously for Ediacaran deep oceans and helps to explain the patchy temporal record of early metazoan fossils.

Numerous lines of geochemical and stable isotopic evidence have indicated that the Ediacaran (635 to 542 million years ago) ocean underwent a stepwise and protracted oxidation [e.g., (14)]. Some geochemical studies suggested that ocean basins were fully oxygenated by the late Ediacaran (1, 2, 4), yet others provided seemingly conflicting evidence for anoxic deep waters (5, 6), with ferruginous conditions [Fe(II)-enriched] persisting into the Cambrian (5). Although a stratified ocean maintained through the Ediacaran Period (7) may help reconcile these seemingly conflicting views, the details remain unclear.

The Doushantuo Formation in the Nanhua Basin, South China, presents a unique opportunity to study Ediacaran ocean chemistry across spatial and temporal scales [e.g., (8)]. It is composed of a succession of both shallow- and deep-water siliciclastic, carbonate, and phosphatic sedimentary rocks deposited immediately after the last globally extensive Neoproterozoic glacial episode (9), widely known as the Marinoan glaciation. Zircon U-Pb ages indicate that the deposition of Doushantuo Formation lasted from ~635 to ~551 million years ago (10), spanning most of the Ediacaran Period.

To investigate the marine redox structure, we characterized the composition of sedimentary Fe mineral species and measured S isotope signatures for sulfides and sulfates (11) at four sections of Doushantuo Formation, which encompass the full range of sedimentary facies from continental shelf to slope to deep basin (fig. S1). We focused on quantifying the fractional abundance of Fe in several highly reactive mineral species (FeHR): pyrite (FePy), Fe(III) oxides, magnetite, and carbonate minerals relative to total Fe (FeT) contents. High FeHR/FeT ratios indicate anoxic conditions (12). If anoxic, low associated FePy/FeHR ratios indicate ferruginous bottom waters, whereas high FePy/FeHR points to euxinic conditions, defined as having an anoxic and H2S-containing water column (5, 12). In most modern and ancient sediments deposited beneath anoxic bottom waters, FeHR/FeT exceeds 0.38, but this threshold value can be reduced to 0.15 [±0.10 (SD)] for thermally altered ancient sedimentary rocks (13) such as Doushantuo Formation because of conversion of FeHR to nonreactive iron during burial. For a euxinic water column, FePy/FeHR in the underlying sediments usually exceeds 0.8 (12). Previous Fe speciation data obtained from Paleo- and Mesoproterozoic sedimentary rocks (14, 15) revealed two distinct redox end members in marine basins characterized by either euxinic or ferruginous deep waters (fig. S2). In contrast, the iron speciation data from Doushantuo Formation are not confined to a single end member (Fig. 1A), suggesting nonuniform redox conditions for deep waters of Nanhua Basin.

Fig. 1

(A) A crossplot of FePy/FeHR versus FeHR/FeT shows a co-occurrence of euxinic with ferruginous conditions for deep waters of Nanhua Basin. Horizontal and vertical (solid for thermally immature and dashed for thermally mature rocks) lines indicate the boundaries for distinguishing euxinic from ferruginous and anoxic from oxic water columns. Dashed lines indicate the most appropriate boundary values for Doushantuo Formation. (B) FePy/FeHR versus Al content shows two distinct redox profiles (path A and path B) that constrain the spatial location of the metastable sulfidic water mass to a region between the inner shelf and shelf margin. Samples in the shaded area were deposited in a stratigraphic interval during an early Ediacaran transgressive period and are outliers for path B.

The inner shelf Jiulongwan section records sedimentary deposition in the shallowest water facies but well below wave base (3) and far from shore, along a broad continental shelf. The associated Fe data from this section plot in both the euxinic and ferruginous fields (Fig. 1A); late Ediacaran black shales from this section all plot in the euxinic zone (table S1). In contrast, samples from the Zhongling section, a deeper shelf margin setting, plot mainly in the ferruginous region, with only three samples having FePy/FeHR ratios > 0.8 (Fig. 1A). The deepest water samples from the slope Minle and basinal Longe sections, which are all early Ediacaran black shales, contain very low levels of pyrite (thus low FePy/FeHR) and yield Fe mineral parameters suggesting ferruginous bottom waters. Considering all the data, the paleoenvironmental trends suggest a co-occurrence of euxinic waters on the shelf with ferruginous deep water toward the center of the basin.

Carbonate lithologies dominate the shallow platform of Nanhua Basin, grading into shales in basinal settings (16, 17), reflecting preferential precipitation of carbonates in shallower waters and enhanced hydrodynamic sorting of fine aluminosilicates into deeper waters. Accordingly, a stratigraphic decrease in carbonate content or, inversely, increasing Al content in samples (fig. S3) from the same site broadly reflects increasing water depth with rising sea level. Doushantuo samples exhibit two dominant trends when FePy/FeHR ratios are viewed in light of Al concentration (Fig. 1B). Almost all inner shelf samples show an increase in FePy/FeHR with increasing Al (path A in Fig. 1B), suggesting generally more sulfidic conditions with higher sea levels (greater distance from shore) but ferruginous conditions when sea level was lower. In contrast, most samples from distal sections show a reverse pattern (path B in Fig. 1B), with low FePy/FeHR values found for samples with moderate-to-high Al content [>1 weight % (wt %)] regardless of organic content, suggesting a dominantly ferruginous and thus sulfate-limited deep-water setting. Taken together, these opposing trends can only be explained by a metastable mid–water-column sulfidic zone located between the inner shelf and the shelf margin nested within a ferruginous water mass (Fig. 2). The location and dimensions of this sulfidic zone would have fluctuated temporally. In most cases, euxinia is suggested independently by diagnostic Mo enrichments (fig. S4) above typical crustal values (4).

Fig. 2

Schematic representation of a stratified redox model for the Ediacaran Nanhua Basin. Most prevalent is a sulfidic water wedge, located at intermediate water depths within a ferruginous deep water mass and maintained by low riverine sulfate input and consumption of sulfate by bacterial sulfate reduction on the continental shelf. A lateral shore-to-basin sulfate concentration gradient (bottom inset) is assumed to be metastable.

We did not observe direct evidence for oxic shallow waters in the sections that we sampled, although such conditions are expected in shallow settings given appreciable atmospheric O2 in the Ediacaran atmosphere (18) and early benthic animal fossils at shallow sites (1922). Therefore, it is likely that the sulfidic and ferruginous zones in the deep anoxic waters persisted beneath shallower oxic and ferruginous layers (Fig. 2).

It is difficult to envisage that a metastable sulfidic water mass could have coexisted with ferruginous deep water for 84 million years without mixing if seawater SO42– concentration in Nanhua Basin was high. Net reduction in the surface S inventory may have limited the resupply of sulfate during the Neoproterozoic (23), particularly if reduced weathering during the Neoproterozoic glaciations suppressed the riverine flux (24). In combination with efficient sulfate removal from seawater through bacterial sulfate reduction and pyrite formation (2426), enhanced by hydrothermal release of dissolved iron (27) during the glaciations and thereafter, this reduced delivery of sulfate reset Ediacaran ocean chemistry back to the ferruginous conditions (5) with extremely low SO42– concentrations that have been prevalent in the Archean (28). Under such conditions, the continental sulfate supply was exhausted before reaching the deep basinal areas, resulting in a dynamically maintained lateral sulfate concentration gradient from shore to basin (Fig. 2, bottom inset) and, in turn, a metastable euxinic zone. Lack of sufficient sulfate to support extensive organic matter remineralization in distal marine settings is consistent with the large and long-stable pool of organic matter suggested for the deep Ediacaran ocean by the unusual carbon isotope systematics expressed in carbonate rocks and sedimentary organic matter of this age (1, 3, 29).

Several lines of evidence support the existence of a persistent sulfate concentration gradient in Nanhua Basin. First, larger ratios of organic C to pyrite S (Corg/Spy) in the distal sections (mean = 34, n = 49) compared with those of the inner shelf site (mean = 4, n = 42), similar to those found under modern low sulfate conditions (30), suggest lower sulfate concentrations in the distal regions (fig. S5). Second, an average offset of 30 per mil (‰) in the S isotope composition of pyrite (δ34Spy) was observed between most of the inner shelf and shelf margin sections (Fig. 3A), with 34S-enriched pyrites formed in the deeper water settings. Furthermore, differences in S isotope ratios (Δ34S) between coeval carbonate-associated sulfate (δ34SCAS) and δ34Spy were on average ~10‰ lower for the shelf margin (~20‰) compared with the inner shelf section (~30‰), suggesting lower SO42– concentration availability along the shelf margin (Fig. 3B). Lastly, concentrations of carbonate-associated sulfate were consistently much lower for the shelf margin rocks compared with those from the inner shelf, with concentrations in the deeper sections frequently too low to permit isotopic analysis (table S1).

Fig. 3

Chemostratigraphic comparisons of (A) δ34Spy and (B) Δδ34S for inner shelf (Jiulongwan) and shelf margin (Zhongling) sections. The sections are correlated on the basis of alignment with published sequence stratigraphic data, and three similar transgressive-regressive sedimentary cycles can be identified (3, 16). The lateral S-isotope gradient is also apparent when the sections are aligned by using carbonate C isotope stratigraphy (fig. S7). Values are reported relative to VCDT (Vienna Cañon Diablo Troilite) standard.

Modern seawater SO42– concentration is ~28 mM, and the isotopic fractionation associated with bacterial sulfate reduction under such high sulfate availability is often large (up to 46‰). In contrast, limited fractionation occurs when sulfate concentration is low, particularly when it falls below a biological threshold of ~200 μM (28). These observations provide upper and lower limits for our estimates of SO42– concentration in the Ediacaran Nanhua Basin, albeit within a broad range. The Δ34S in the inner shelf Jiulongwan section increases abruptly from 1.5‰ in the basal cap carbonate to ~30‰ in the overlying ~40 m and thereafter, with a few samples having Δ34S > 30‰ in the upper section (Fig. 3B). These trends are consistent with an increase of SO42– concentration from <200 μM during post-Marinoan deglaciation to >200 μM thereafter and extending into the late Ediacaran. Although Δ34S values from the shelf margin section are no more than 24‰, such fractionations are large enough to point to local sulfate levels > 200 μM during the late Ediacaran. The Δ34S data do not provide a clear upper limit for late Ediacaran sulfate concentrations, but the isotopic and concentration gradients inferred from our study demand a sulfate level that was only a small fraction of the modern 28 mM.

Because the sulfidic zone could expand into previously oxygenated areas of the shelf during transgression and during pulses of high productivity, the generally patchy record of metazoans observed through the Ediacaran (3, 1922) might be explained by fluctuating oceanic redox conditions in and around the continental shelf. The finding of Ediacaran metazoan resting cysts, in the form of large ornamented acritarch fossils most prevalent in lower to middle Ediacaran strata from Doushantuo Formation (19) and from Australia and Russia (22), has previously been interpreted as an evolutionary response of early benthic metazoans to prolonged episodes of anoxia in shelf and platform bottom waters (22). Such a control is broadly consistent with our ocean redox model. For the inner shelf facies of Doushantuo Formation, available data highlight a broad correlation between sedimentary horizons containing the most diverse assemblage of acritarch fossils or animal embryos and intervals with the lowest pyrite contents (fig. S6), suggesting that the presence of hydrogen sulfide, rather than merely the absence of oxygen, hindered colonization of the shelf sea floor by early animals.

Supporting Online Material

www.sciencemag.org/cgi/content/full/science.1182369/DC1

Materials and Methods

Figs. S1 to S7

Table S1

References

References and Notes

  1. See section of Materials and Methods in supporting material available on Science Online.
  2. We are extremely grateful to R. Raiswell, S. Bates, W. Gilhooly, B. Gill, J. Owens, A. Khong, P. Marenco, N. Planavsky, C. Reinhard, M. Rohrssen, C. Scott, S. Severmann, J. Huang, L. Feng, H. Chang, and Q. Zhang for laboratory and field assistance and helpful discussions. The NSF Earth Sciences program (grant EAR-0720362 to G.D.L. and T.W.L. and grant EAR-0719493 to A.L.S.), National Science Foundation of China Fund (grant 40532012 to X.C.), the Chinese Academy of Sciences Fund (grant KZCX3-SW-141 to X.C.), the NASA Astrobiology Institute and the Agouron Institute provided funding.
View Abstract

Navigate This Article