Brian W. Stewart

Strontium isotopes as tracers of ecosystem processes: theory and methods

Abstract The strontium (Sr) isotope method can be a powerful tool in studies of chemical weathering and soil genesis, cation provenance and mobility, and the chronostratigraphic correlation of marine sediments. It is a sensitive geochemical tracer, applicable to large-scale ecosystem studies as well as to centimeter-scaled examination of cation mobility within a soil profile. The 87Sr/86Sr ratios of natural materials reflect the sources of strontium available during their formation. Isotopically distinct inputs from precipitation, dryfall, soil parent material, and surface or groundwater allow determination of the relative proportions of those materials entering or leaving an ecosystem. The isotopic compositions of labile (soil exchange complex and soil solution) strontium and Sr in vegetation reflect the sources of cations available to plants. Strontium isotopes can be used to track the biogeochemical cycling of nutrient cations such as calcium. The extent of cation contributions from in situ weathering and external additions to soil from dust and rain can also be resolved with this method. In this paper, we review the geochemistry and isotopic systematics of strontium, and discuss the use of this method as a tracer of earth surface processes.

Geochemical and Strontium Isotope Characterization of Produced Waters from Marcellus Shale Natural Gas Extraction

Extraction of natural gas by hydraulic fracturing of the Middle Devonian Marcellus Shale, a major gas-bearing unit in the Appalachian Basin, results in significant quantities of produced water containing high total dissolved solids (TDS). We carried out a strontium (Sr) isotope investigation to determine the utility of Sr isotopes in identifying and quantifying the interaction of Marcellus Formation produced waters with other waters in the Appalachian Basin in the event of an accidental release, and to provide information about the source of the dissolved solids. Strontium isotopic ratios of Marcellus produced waters collected over a geographic range of ~375 km from southwestern to northeastern Pennsylvania define a relatively narrow set of values (ε(Sr)(SW) = +13.8 to +41.6, where ε(Sr) (SW) is the deviation of the (87)Sr/(86)Sr ratio from that of seawater in parts per 10(4)); this isotopic range falls above that of Middle Devonian seawater, and is distinct from most western Pennsylvania acid mine drainage and Upper Devonian Venango Group oil and gas brines. The uniformity of the isotope ratios suggests a basin-wide source of dissolved solids with a component that is more radiogenic than seawater. Mixing models indicate that Sr isotope ratios can be used to sensitively differentiate between Marcellus Formation produced water and other potential sources of TDS into ground or surface waters.

Effects of rainfall on weathering rate, base cation provenance, and Sr isotope composition of Hawaiian soils

A climate transect across the Kohala Peninsula, Hawaii provides an ideal opportunity to study soil processes and evolution as a function of rainfall. The parent material is the ∼150 ka Hawi alkali basalt aa flow, and median annual precipitation (MAP) changes from ∼16 cm along the west coast to ∼450 cm in the rain forest near the crest of the peninsula. We measured labile (plant-available) base cation concentrations and 87Sr/86Sr ratios of labile strontium and silicate residue from soil profiles across the transect from 18 to 300 cm MAP. Depletion of labile cations and a shift in labile 87Sr/86Sr ratios toward rainwater values with increasing rainfall clearly show the transition from a mineral-supported to a rainwater-supported cation nutrient budget. In contrast, increases in soil silicate residue 87Sr/86Sr values with increasing MAP result primarily from input of exogenous eolian material (dust derived from Asian loess), with a greater dust fraction at the high MAP sites due to aerosol washout. Most of the soil silicate strontium in high-MAP sites is still derived from the original parent material, but the shallower portions of profiles can be dust-dominated. The variations in labile 87Sr/86Sr with rainfall allow us to calculate weathering rates as a function of MAP. The primary uncertainty is the degree to which Sr in rainwater actually interacts with the labile cation reservoir before being flushed from the system; mass balance calculations for the 150 ka evolution of the profile suggest that only on the order of 5 to 50% of rainwater strontium exchanges with the labile reservoir. Our models suggest that the present-day supply of strontium by weathering increases steadily with rainfall in the low-MAP (<140 cm) sites, then decreases dramatically as the soils become depleted in weatherable parent material. This implies that the initial weathering rate of the high-MAP sites was very high, and that there may be some change in soil weathering behavior in the 100 to 160 cm MAP range. Weathering rates calculated from the labile 87Sr/86Sr are on the same order as other estimates for chemical denudation rates of basaltic terrains.

The importance of sea spray to the cation budget of a coastal Hawaiian soil: a strontium isotope approach

Soil nutrients such as Ca, Mg, and K are traditionally thought to be derived primarily from rock weathering. Here we show that sea spray is a significant source of nutrient elements to modern and buried soils developed on <30,000-year-old Pahala Ash deposits 50 m from the coast at South Point, Hawaii. The soil profiles evolved in a semi-arid climate and have always been above sea level and the water table. Rhizoliths (fossilized root traces) and horizontal laminated carbonate sheets found in buried soils are composed of high-Mg calcite (up to 14 mol% MgCO3). Differences in strontium isotopic composition between marine aerosols (87Sr/86Sr=0.7092) and tephra parent material (∼0.7035) allow quantification of cation sources to the labile soil reservoir and to pedogenic carbonate. Mixing equations indicate that 50–80% of labile soil Sr and approximately half of carbonate Sr was derived from marine sources. Using the Sr isotopic signatures and Sr/Ca ratios of seawater and tephra as end members, we determined that up to 2/3 of the Ca in the labile reservoir and up to 1/3 of Ca in the carbonates has a marine origin. Carbonate 87Sr/86Sr ratios are fairly constant with depth, but labile 87Sr/86Sr ratios indicate decreasing sea spray aerosol influence with depth. This trend could be due either to sequestering of aerosol-derived Sr in the upper part of the profile or to lower aerosol input in the past due to lower sea level. The unusual occurrence of high-Mg pedogenic calcite probably results from high labile Mg/Ca ratios during earlier stages of weathering, coupled with rapid calcite precipitation during soil pore water evaporation.

Quantitative strontium isotope models for weathering, pedogenesis and biogeochemical cycling

Abstract Isotopes of strontium (Sr) are a useful tracer for weathering, atmospheric fluxes, cation biocycling, and pedogenesis. We present basic models for application of strontium isotopes to the soil–vegetation–atmosphere system. The mathematical formulations fall into the general categories of: (1) steady-state models, in which isotopic ratios remain constant over the time scale of interest; and (2) time-dependent models, in which isotope ratios change through time. In the steady-state models, fluxes of Sr and other elements to the system are constant. Steady-state models can be used to infer short-term weathering rates from river and stream isotope compositions, to determine fluxes to a single- or multiple-layer soil exchange/solution system, and to quantify nutrient fluxes to vegetation. Time-dependent models involve a change in isotopic ratio from some initial value to a new value over the time period of interest; in some cases, the change may represent a shift from one steady-state situation to a new one after a shift in one or more of the fluxes feeding the system. Examples of applications of time-dependent models include identifying the dominant cation sources to an evolving soil exchange/solution system, and calculating weathering rates by measuring the isotopic compositions of primary soil minerals. We use time-dependent models to explain differences in the isotopic ratios of labile and carbonate Sr from arid sites in Hawaii (with a basalt parent material isotopic signature) and New Mexico (with an atmospheric isotopic signature). These models suggest that the difference is due to a combination of low atmospheric strontium fluxes and high weathering rates in the Hawaiian profile compared to the New Mexico calcrete profile.

Sm-Nd chronology and petrogenesis of mesosiderites

We have obtained Sm-Nd data from four mesosiderite silicate clasts, including three clasts with a variety of textures from the Vaca Muerta type 1A mesosiderite and one gabbroic clast from the Mt. Padbury mesosiderite. The gabbroic Vaca Muerta Pebble 12 and basaltic Pebble 16 yield identical ^(147)Sm- ^(143)Nd ages of 4.48 ± 0.19 AE and 4.48 ± 0.09 AE, respectively, while the highly recrystallized Pebble 5 gives an age of 4.42 ± 0.02 AE; Mt. Padbury yields an age of 4.52 ± 0.04 AE. All clasts show a correlation of ^(142)Nd /^(144)Nd with ^(147)Sm /^(144)Nd , and provide unequivocal evidence for the presence of live ^(146)Sm at the time of their formation. Calculated initial ^(146)Sm /^(144)Sm values range from 0.004 (Pebble 5) to 0.006 (Pebbles 12, 16, and Mt. Padbury) and are generally consistent with the ^(147)Sm- ^(143)Nd ages. However, discordance of whole-rock leach and residue data and some disagreement between ^(146)Sm- ^(142)Nd relative ages and ^(147)Sm- ^(143)Nd absolute ages indicate small but significant disturbances to the Sm-Nd systematics. The ranges of ages and initial ^(146)Sm /^(144)Sm and ^(143)Nd /^(144)Nd values suggest that each of these silicate clasts underwent a separate, protracted evolution on its parent body prior to mixing with metal. Textural and trace element criteria indicate that individual clasts from the same mesosiderite often had very different igneous sources and thermal histories prior to their incorporation in the meteorite. Pebble 12 is extremely LREE depleted, probably a result of several melt extraction events, whereas Pebbles 5 and 16 and Mt. Padbury have nearly chondritic Sm/Nd with bulk REE concentrations higher than chondrites by factors of 5 to 15. In general, the Sm-Nd systematics of mesosiderite silicates require formation of the silicates on a parent planet which underwent relatively early and extreme differentiation. Preservation of diverse, old ages and the presence of ^(146)Sm imply that metal-silicate mixing did not seriously alter the Sm-Nd isotopic memories of these clasts. We present a model for metal-silicate mixing which combines the cooling history with isotopic reequilibration for the case of thermal blanketing. We show that the total amount of isotopic reequilibration in a sample can be related to the initial temperature, depth of burial, grain size, and diffusion parameters. Application of this model to the silicate clasts measured in this study indicates that if the metal and silicate were thermally equilibrated above the metal solidus temperature during mixing, then the clasts must have been buried no deeper than 1-10 m in regolith during the initial high-temperature cooling phase in order to prevent the Sm-Nd systems from being extensively reset. In order to reconcile these results with the slow cooling rates at lower temperatures determined from studies of exsolution in the metal phase, we infer that heat was transferred rapidly from hot metal to cold silicate material during initial metal-silicate mixing, and that the deeply buried portions of the mixture then cooled slowly after reaching thermal equilibrium at ~600-700°C. The data from this study point to the following history for the mesosiderite parent body: (1) differentiation of a silicate parent body within the first ~50 m.y. of solar system history to form diverse parent magmas; (2) emplacement of primitive and differentiated mafic magmas near the planetary surface, and extensive differentiation in the near surface environment; (3) formation and reworking of regolith breccias through impact gardening at the near surface of the body; (4) mixing of molten Fe-Ni metal with the regolith followed by rapid cooling 100-150 m.y. after the origin of the solar system; (5) slow cooling from temperatures of ~700°C to produce the observed nickel diffusion profiles in the iron phase; (6) mild impact metamorphism and brecciation to obtain the much younger K-Ar ages; and (7) recent collisions or perturbations sending mesosiderite fragments into Earth-crossing trajectories.

Isotopic studies of processes in mafic magma chambers: II. The Skaergaard Intrusion, East Greenland

Isotopic ratios of Nd and Sr have been measured in a suite of samples spanning most of the exposed stratigraphy of the Skaergaard intrusion in order to detect and quantify input (such as assimilated wallrock and fresh magma) into the magma chamber during crystallization. Unlike δ18O and δD, Nd and Sr isotope ratios do not appear to have been significantly affected by circulation of meteoric waters in the upper part of the intrusion. Variations in initial 87Sr/86Sr and εNd suggest that the Skaergaard magma chamber was affected during its crystallization by a small amount (2%–4%) of assimilation of Precambrian gneiss wallrock (high 87Sr/86Sr, low εNd) and possibly recharge of uncontaminated magma. Decreases in εNd and increases in 87Sr/86Sr during the early stages (0%–30%) of crystallization give way to approximately unchanging isotopic ratios through crystallization of the latest-deposited cumulates. Modelling of assimilation-fractional crystallization-recharge processes using these data as constraints shows that the assimilation rate must have been decreasing throughout crystallization. In addition, the isotope data allow replenishment by an amount of uncontaminated magma equal to 20%–30% of the total intrusion mass, occurring either continuously or in pulses over the first 75% of crystallization. Comparison of the recharge models with published Mg/(Mg+Fe2+) data from Skaergaard cumulates shows that the modelled replenishment rates are not inconsistent with available major element data, although significant recharge during the final ∼25% of crystallization can be ruled out. The isotope data show that the Skaergaard magma could have incorporated only a small amount of the gneiss that it displaced from the floor of the chamber; assimilation appears to have taken place primarily across a partially molten zone that formed at the roof from the wallrock that was dislodged during emplacement. In the latest stages of crystallization (>75% crystallized), the Skaergaard magma may have become stratified into two separately-convecting layers, effectively insulating Layered Series cumulates from further contamination.

Geomorphologic evidence for the late Pliocene onset of hyperaridity in the Atacama Desert

The Atacama Desert has experienced a long and protracted period of hyperaridity that has resulted in what may be the most unusual biome on Earth, but the duration of this aridity is poorly constrained. We reconstructed aspects of the fluvial and geochemical history of this region using integrated landscape features (alluvial fans, hillslope soils, soil chemistry, river profiles) in the southern portion of the present desert. Topographic reconstructions of a large watershed (11,000 km 2 ) show deep incision and sediment removal between the late Miocene and the end of the Pliocene, and modest to negligible incision in post-Pliocene times. These changes in incision suggest an ∼50–280× reduction in river discharge, which should reflect corresponding changes in precipitation. Changes in the nature of hillslope soils in the Atacama Desert indicate that in the Pliocene or earlier, hillslopes were mantled with silicate-derived soil. This mantle was stripped off and locally deposited as alluvial fans (late Pliocene to early Pleistocene) that now block or otherwise cause a rearrangement of Pliocene and earlier river channels. Finally, the hillslopes have largely accreted a soil mantle of dust and salt since the apparent late Pliocene stripping, suggesting a decline in annual precipitation of at least 125 mm yr -1 or more (mean annual precipitation [MAP] is now -1 ). Embedded in the long post-Pliocene era of salt accumulation, there are a variety of features suggesting overland flow on hillslopes (rills, striped gravel deposits, piping, and water spouts) and large, infrequent storms that infiltrated gentle alluvial fans (due to the depth of salt-rich horizons). Despite evidence for episodes that punctuate the hyperaridity, the magnitude and duration of these pluvial events have been insufficient to remove the regional accumulations of sulfate, chloride, and nitrate. The late Pliocene cessation of many fluvial features is coincident with recent research on the tropical Pacific, which shows that the Pacific was in a permanent El Nino state until ca. 2.2 Ma, at which time sea-surface temperatures offshore of South America declined greatly relative to those of the western Pacific, in turn setting up the present El Nino–Southern Oscillation (ENSO) climate system. These observations indicate that the latest period of aridity has been prolonged and largely continuous, and it appears to have occurred in step with the onset of the ENSO climate system, beginning ∼2 m.y. ago.

Rare earth element sources and modification in the Lower Kittanning coal bed, Pennsylvania: implications for the origin of coal mineral matter and rare earth element exposure in underground mines

Abstract In this study, we examine the variations in rare earth elements (REE) from the Lower Kittanning coal bed of eastern Ohio and western Pennsylvania, USA, in an attempt to understand the factors that control mineral matter deposition and modification in coal, and to evaluate possible REE mixed exposure hazards facing underground mine workers. The results of this study suggest that the Lower Kittanning coal mineral matter is derived primarily from a clastic source similar to that of the shale overburden. While highly charged cations like silicon, aluminum, and titanium remained relatively immobile within the coal mineral matter, iron (primarily as pyrite) was added from nonclastic sources, either during deposition of the coal mire vegetation or subsequent to burial. Other mobile cations (e.g., alkali and alkaline earth elements) appear to have been added to and/or leached from the originally deposited clastic mineral matter. Most of the sulfur in the Lower Kittanning coal bed is bound as FeS2 in the mineral matter, but a majority of samples contain a small excess of S that is most likely organically bound. In general, the total rare earth element content (TREE) in coal ash is greater than that in the shale overburden. If the primary source of mineral matter is the same as that for the overlying shale, then REE must have been enriched in the coal mineral matter subsequent to deposition. The total rare earth element content of Lower Kittanning coals correlates strongly with Si concentration ([TREE]≈0.0024 [Si]), which provides a threshold for evaluating possible mixed exposure health effects. Chondrite-normalized REE patterns reveal a shale-like light rare earth element (LREE) enrichment for the coal, similar to that of the shale overburden, again suggesting a primarily clastic REE source. However, when normalized to the shale overburden, most of the coal ash samples display a small but distinct heavy rare earth element (HREE) enrichment. We surmise that the HREE were added and/or preferentially retained during epigenesis, possibly associated with groundwater flow through the coal unit, but not necessarily in close association with the addition of iron. At least some of the “excess” HREE could be organically bound within the Lower Kittanning coal.

Diffusive Isotopic Contamination of Mafic Magma by Coexisting Silicic Liquid in the Muskox Intrusion

Shifts in 87Sr/86Sr and 143Nd/144Nd ratios measured in cumulates from the upper levels of the Muskox mafic intrusion indicate that isotopic and bulk chemical exchange were decoupled across a mafic-silicic liquid interface during crystallization of the intrusion. Modeling of diffusive exchange between liquid layers demonstrates that isotopic compositions of silicate liquids in layered magma chambers may be strongly affected by this process on time scales of 103 to 104 years. Diffusive contamination can be used to place constraints on the physical processes and time scales of magmatic systems.