Ca and Mg cycles in deep time
As part of our effort to understand better Earth's deepest δ¹³C excursion, also known as the 'Wonoka-Shuram' excursion, we have expanded the approaches used to study it with measurements of calcium and magnesium isotopes on the Wonoka Formation of South Australia. Calcium in the modern ocean is isotopically homogenous, owing to its long residence time of ~1 Myr. Ca inputs to the ocean system are mainly from riverine and hydrothermal sources, and the only quantitatively significant output is CaCO3 precipitiation and burial (figure right). CaCO3 minerals are fractionated from seawater - 40Ca is preferentially taken up into the solid, leaving seawater enriched in 44Ca - although the degree of this fractionation varies with mineralogy (calcite vs. aragonite), temperature, and precipitation rate. Modern magnesium isotope systematics share many similarities with calcium. Its long residence time (~10 Myr) leads to isotopic homogeneity in the modern ocean. Carbonate minerals similarly sequester the light isotope of 24Mg over 26Mg, with limestone generally being more fractionated than dolomite, resulting in a seawater δ26Mg heavier than riverine inputs (figure above). One important difference between the Mg and Ca cycles is that Mg has two important sinks - dolomite formation and the formation of Mg-rich clays from the alteration of seafloor basalts. Unlike other isotopic systems which have been used to study the 'Wonoka-Shuram' excursion (87Sr86Sr, δ34S of carbonate-associated sulfate, δ13Corg), calcium and magnesium species are major constituents of carbonate rock (Ca for both limestone and dolomite, Mg for dolomite) and thus are not subject to the same concerns regarding diagenesis and isotopic resetting as are trace systems. Additionally, as the residence times for carbon, calcium and magnesium are orders-of-magnitude different from each other (~105, 106 and 107 years, respectively), synchronous changes in each of the isotopic systems are not expected if each of the records represents secular, long-term evolution of seawater composition.
Ca and Mg isotopes in the Wonoka Formation
Data from four measured sections spanning the basin (figure left) reveal stratigraphically coherent trends, with variability of ~1.5‰ in δ26Mg and ~1.2‰ in δ44Ca. This Ca isotope variability dwarfs the 0.2-0.3‰ change seen coeval with the Permian-Triassic mass extinction, the largest recorded in the rock record, and is on par with putative changes in the δ44Ca value of seawater seen over the Phanerozoic Eon. Changes in both isotopic systems are too large to explain with changes in the isotopic composition of Ca and Mg in global seawater given modern budgets and residence times, and thus must be products of alternative processes. Relationships between δ44Ca and [Sr] and δ26Mg and [Mg] are consistent with mineralogical control (e.g., aragonite vs. calcite, limestone vs. dolostone) of isotopic variability. The most pristine samples in the Wonoka dataset, preserving Sr concentrations (in the 1000's of ppm range) and δ44Ca values inherited from an originally aragonitic polymorph, have δ13Ccarb of -8 to -7‰, thereby providing strong geochemical evidence that extremely negative δ13Ccarb values are primary products of the Ediacaran surface environment.
Isotopic measurements by multi-collector ICP-MS
Measurements of Ca and Mg isotopes were made in collaboration with John Higgins in his stable isotope geochemistry lab at Princeton University. Sample powders of 5-10 mg were developed by micro-drilling individual laminae of polished slabs. Powders were dissolved in a buffered solution of anhydrous acetic acid and ammonium hydroxide (pH of ~5). To develop single-element analytes for Ca and Mg isotopic analysis, samples were processed using a Thermo Dionex 5000+ ion chromatography (IC) system (figure above, left panel). The automated IC system runs samples through an in-line CS16 cation exchange column which separates various cations and measures peak intensities using changes in conductivity (middle panel). Collection windows are specified to collect pure Ca or Mg cuts. Isotopic analyses were performed on a Thermo Neptune Plus inductively-coupled plasma mass spectrometer (ICP-MS; right panel). For Ca, accuracy of our measurements is assessed by analysis of the common SRM 915b Ca standard and modern seawater. The external reproducibility of our measurements is defined by full sample replicates (column chemistry followed by mass spectrometry), and is ~0.2 (2σ) for δ44/40Ca, based on measurements of an in-house synthetic limestone standard. For Mg, to assess accuracy, we analyze both CAMBRIDGE-1 and seawater. Precision of Mg isotope measurements is assessed similarly to Ca by analysis of an IC-separated, in-house synthetic dolomite standard (Mg/Ca = 1 + minor/trace elements), and long-term external reproducibility on δ26Mg is ~0.1 (2σ).