Sulfate Burial Constraints on the Phanerozoic Sulfur Cycle

See allHide authors and affiliations

Science  20 Jul 2012:
Vol. 337, Issue 6092, pp. 331-334
DOI: 10.1126/science.1220224


The sulfur cycle influences the respiration of sedimentary organic matter, the oxidation state of the atmosphere and oceans, and the composition of seawater. However, the factors governing the major sulfur fluxes between seawater and sedimentary reservoirs remain incompletely understood. Using macrostratigraphic data, we quantified sulfate evaporite burial fluxes through Phanerozoic time. Approximately half of the modern riverine sulfate flux comes from weathering of recently deposited evaporites. Rates of sulfate burial are unsteady and linked to changes in the area of marine environments suitable for evaporite formation and preservation. By contrast, rates of pyrite burial and weathering are higher, less variable, and largely balanced, highlighting a greater role of the sulfur cycle in regulating atmospheric oxygen.

Sulfate (SO42–) is the fourth most abundant ion in modern seawater and a major component of the alkalinity budget, which governs the pH of seawater (1). Bacterial sulfate reduction accounts for ~50% of sedimentary organic matter respiration (2), and precipitation of pyrite (FeS2) is one of the major exit channels of sulfur from the ocean (3). Because reduction of riverine sulfate and burial of the sulfide leave oxidized products in the ocean-atmosphere system, pyrite burial is considered a major indirect source of oxygen to the atmosphere (4, 5).

Several time series data sets constrain aspects of the Phanerozoic sulfur cycle (Fig. 1A). The sulfur isotope composition, δ34S, of carbonate-associated sulfate, sulfate evaporites, and barite (BaSO4) records the δ34S of seawater sulfate, whereas the δ34S of sedimentary pyrite captures the products of microbial sulfate reduction (68). The chemical composition of fluid inclusions in halite constrains the concentration of major ions in seawater, including sulfate (9, 10).

Fig. 1

Observational constraints on the Phanerozoic sulfur cycle. (A) Sulfur isotope composition of seawater sulfate and sedimentary pyrite (8), and seawater sulfate concentration from fluid inclusions in halite (9, 10). Isotope compositions are reported relative to the standard VCDT (Vienna Canyon Diablo Troilite). (B) North America and the Caribbean (NAC) submerged continental area estimates and the rate of change of global mean sea level (23). (C) Estimates of NAC total and per-area sulfate evaporite burial rates. (D) NAC sulfate evaporite burial rates scaled to global fluxes and corrected for decay of the surviving record (20), including uncertainty (shaded envelope).

Variability in the δ34S records of seawater sulfate and sedimentary pyrite is typically interpreted to reflect changes in the fraction of sulfur removed from the oceans as pyrite, fpyr. Because pyrite is depleted in 34S by several percent relative to the sulfate reservoir from which it formed, times of high seawater sulfate δ34S are interpreted as times of high rates of pyrite burial. By assuming a steady state and constant input magnitude and δ34S, or by scaling inputs and outputs to modern values, models of the Phanerozoic sulfur cycle explain long-term trends in δ34S values by changes in fpyr between ~0.2 and ~0.6 (4, 1113). Recognizing that the magnitude and δ34S of the influxes to the ocean have likely varied in time, thereby influencing the isotopic record, some models included parameterized influxes and solved mass balance equations for the outfluxes and the value of fpyr (4, 13). The parameterizations are uncertain, however, because they are largely based on a scaling of modern influxes by debated factors, such as the relative rates of seafloor spreading and continental runoff (14, 15).

It is possible to measure the sink of sulfate evaporites from seawater and obtain estimates of the influx magnitude and δ34S by mass balance, though previous volume estimates of Phanerozoic evaporites (mostly halite, but some sulfate) have been considered too coarse or uncertain to accurately constrain past rates of sulfate burial (1618). We quantified sulfate burial over Phanerozoic time, using a comprehensive macrostratigraphic database (19, 20), which includes 23,843 lithostratigraphic rock units in 949 geographic locations across North America and the Caribbean (NAC). Data were binned by age, and sulfate burial rates were obtained by dividing evaporite volume by bin duration. Macrostratigraphy-based estimates of sulfate burial rates are higher than those derived from other compilations. This is due to the improved spatial and lithological resolution of this data set, which includes sedimentary rocks in the surface and subsurface, and many comparatively thin but widespread deposits not included in previous compilations. Notably, the NAC burial rates are highly variable, with values 2 to 14 times the average occurring mainly in Paleozoic intervals (Fig. 1C).

The macrostratigraphic database currently provides comprehensive coverage only in NAC, but can be scaled globally (Fig. 1D) by using mechanistic relationships between the observations and environmental controls on sulfate evaporite deposition (20). The volume-weighted average ratio of global to North American sulfate deposit volumes is ~8 (16, 17). In comparison, the area-weighted ratio of global to NAC submerged continental area in latitudes of net evaporation, estimated from paleogeographic reconstructions (20, 21), is ~7. This close agreement reflects a primary requirement for massive sulfate evaporite deposition—hydrographic isolation of large, marine-fed basins at latitudes of net evaporation (22). Such basins are created by rifting, small changes in sea level or the development of a barrier to circulation (22), often at the shoreward edge of submerged continental shelves. Indeed, the long time-scale variability in the burial rate data is well explained by the estimated NAC submerged continental area at latitudes ±10° to 50° (linear product moment correlation coefficient of 0.47 at the temporal resolution of the paleogeographic reconstructions), and we use this relationship to derive average global burial rates. We add shorter time scale variability to this average using correlations between NAC fluxes and the rate of change in eustatic sea level (20, 23). Additional factors that influence evaporite deposition, such as geodynamic controls on basin subsidence and climate patterns at the regional-to-local scale (22), are not explicitly represented by this scaling methodology. They are, however, implicitly included in the scaling because they have contributed to the observed NAC sulfate burial rates (20).

Intervals of rapid sulfate evaporite burial (NAC and global) occur with equal frequency during times of high and low marine sulfate concentrations (Fig. 1). This reveals that sulfate burial rates were disconnected from changes in the activity product of calcium and sulfate [aCa2+aSO42–; e.g., (13)], which have varied by up to ~15% (9, 10). We suggest, instead, that the first-order control on sulfate burial was the episodic availability of suitable environments in which evaporation of seawater could lead to saturation, precipitation, accumulation, and long-term preservation of sulfate evaporites. Unsteady sulfate burial rates, governed by the interactions between sea level, tectonics, and paleogeography, suggest that the critical statistic derived from isotope mass balance studies, fpyr, convolves information regarding the relative activity of sulfate-reducing microbiota with the availability of environments suitable for sulfate burial. For example, a decrease in the area of shallow seas due to a drop in sea level or continental migrations could decrease sulfate burial rates and increase fpyr. The rate of pyrite burial itself need not change. Scaled globally and corrected for the decay of surviving rock with time (20), the macrostratigraphic data reveal an average Phanerozoic sulfate evaporite burial rate of ~3.3 × 1011 to 4.5 × 1011 mol year–1, depending on bin duration. This is only ~10 to 30% of the estimated riverine influx of sulfate to the oceans [~1.5 × 1012 to 3.5 × 1012 mol year–1; (12, 24)], implying that the value of fpyr has been ~0.7 to 0.9 if the sulfur cycle operated close to steady state.

The macrostratigraphic data motivated us to reevaluate models of the sulfur cycle by integrating constraints from sulfur isotope measurements and sulfate concentration data from fluid inclusions. In any isotope mass balance model, the time-dependent concentration and δ34S of seawater sulfate can be expressed by two equations

dMdt=JinJeJp (1)dMδdt=JinδinJeδJp(δΔ)(2)

Here, M is the concentration of seawater sulfate and δ is its δ34S value. Jin is the total influx of sulfur to the oceans, with contributions from evaporite weathering, oxidative weathering of sedimentary and igneous sulfide minerals, and volcanic outgassing of sulfur volatiles. The δ34S of this influx, δin, depends on the relative contributions of these three components. Je and Jp are the burial fluxes of sulfate evaporites and pyrite, respectively, and Δ is the average difference (in permil) between the δ34S of contemporaneous sulfate evaporite and pyrite sedimentary sinks. The values of M, δ, and Δ can be constrained by δ34S and fluid inclusion data, and Je, by macrostratigraphic data.

It is possible to solve Eqs. 1 and 2 for the outputs using the inputs (“parameterized input”) or for the inputs using the outputs (“parameterized output”), and the solutions can be evaluated with independent physical understanding of the sulfur cycle. We examined three variants of the parameterized input model (20). The first assumed constant values of Jin, δin, and Δ (11, 12, 24) and either a steady state or a dynamic mass balance. In the second and third models, we used influx parameterizations similar to those of (18) and (13). We examined three variants of the parameterized output model (20). The first is a model of constant fpyr. The second, motivated by the observation that pyrite burial is common in shallow, organic-rich sediments (25), is a model in which Jp scales with normalized global submerged continental area. This scaling may be weak because although high submerged areas may lead to high bacterial sulfate reduction rates, inefficient delivery of reactive iron across the shelf could limit pyrite burial to near-shore environments (26). We therefore considered a third model of constant Jp. Using the macrostratigraphy-based sulfate burial rates as additional constraints, we solved Eqs. 1 and 2 for the time-dependent values of Jin and δin.

Of the parameterized output models (Fig. 2, A to C), the models of constant fpyr yield unrealistically low values of Jin (often lower than estimates of the volcanic influx and occasionally negative) and wildly varying values of δin. The latter is related to the former; if Jin is small, then the change in δin required to drive an observed change in seawater sulfate δ34S must be large. The models of submerged continental area-dependent or constant Jp yield reasonable values of Jin and low values of δin (near the average sedimentary pyrite δ34S) that become higher during times of increased Je. This is consistent with the idea of rapid sediment recycling (4), where low values of δin closely follow high relative rates of pyrite burial due to oxidative weathering of recently deposited pyrite; high δin values result from weathering of newly deposited sulfates. Notably, dynamic and steady-state solutions diverge only during short intervals of rapid change in seawater sulfate concentrations, implying that the system operates much of the time close to steady state. All three parameterized output models, when constrained by the macrostratigraphic data, indicate high average values of fpyr (~0.7 to 0.9).

Fig. 2

Model results. (A) fpyr in the parameterized output models. (B and C) Jin and δin required for mass balance constrained by time series in Fig. 1. (D) fpyr in the constant-input models. (E) Je calculated with the dynamic model (mass balance) compared to the macrostratigraphic data, and Jin (=Jout) required to reproduce seawater sulfate δ34S in the steady-state model. (F) fpyr in the parameterized input models. (G) Je calculated by mass balance compared to the macrostratigraphic data. Shaded envelopes reflect uncertainty in Je estimates (see Fig. 1).

The parameterized input models yield values of fpyr similar to those in previous modeling studies of the Phanerozoic sulfur cycle (Fig. 2, D and F). However, the model of constant δin at steady state occasionally requires unrealistically low values of Jin to reproduce the δ34S records [Fig. 2E; e.g., 200 Ma (millions of years ago)], supporting the notion that the δ34S value of inputs has varied with time. The values of Je required for mass balance under the other parameterizations are ~three times as large as our burial rate estimates (Fig. 2, E and G, and fig. S1). We hypothesize that this is because parameterizations for the sulfate influx into the ocean are calibrated to modern riverine fluxes, which are higher than the expected long-term average due to the exposure and weathering of recently deposited evaporites (supplementary text). This notion is supported by anomalously high per-area sulfate deposition rates obtained from the NAC macrostratigraphic compilation during the last 10 million years (Fig. 1C); by widespread massive Neogene-age (~23 to 2.6 Ma) evaporites in Europe, Asia, and Africa (22, 2731); and by the recognition that deposition rates decrease with increasing observation time scale to constant, long-term values (32). This implies that the high abundance of Neogene-age sulfate evaporites represents gross deposition. Ultimately much of this material will not be preserved, and net deposition rates will converge to long-term Phanerozoic rates (33).

The results presented here describe a Phanerozoic sulfur cycle in which the majority of net inputs and outputs are oxidative weathering and burial of sedimentary pyrite, respectively. Over the time intervals resolved by our macrostratigraphic data, the long-term value we estimate for fpyr is generally high, and large downward excursions in its value are associated with high rates of sulfate evaporite burial, rather than times of less pyrite burial. Although spatial heterogeneity and short–time scale variability in the value of Jp have likely occurred (34), global models of constant or slowly varying Jp (in response to changes in submerged continental area) yield results that are consistent both internally and with existing observations of seawater sulfate concentration and δ34S. Large and stable pyrite weathering and burial fluxes highlight the importance of oxidation-reduction feedbacks between carbon, iron, and sulfur (24) and imply a greater role for the sulfur cycle in regulating Phanerozoic atmospheric oxygen.

Supplementary Materials

Materials and Methods

Supplementary Text

Figs. S1 to S5

References (3546)

Database S1

References and Notes

  1. Methods are available on Science Online.
  2. Acknowledgments: We thank D. Canfield and J. Adkins for helpful discussion, and C. Scotese for help with the paleogeographic reconstructions. I.H. acknowledges support from a Texaco Postdoctoral Fellowship in Geological and Planetary Sciences at the California Institute of Technology and a Sir Charles Clore Prize for Outstanding Appointment in the Experimental Sciences at the Weizmann Institute of Science. S.E.P. was funded by NSF grant EAR-0819931. W.W.F. acknowledges support from the Agouron Institute and a David and Lucile Packard Foundation Fellowship for Science and Engineering. The binned macrostratigraphic data are available as a supplementary table on Science Online.
View Abstract

Navigate This Article