Reports: DNI253802-DNI2: Magnesium Isotopes and the Origin of Marine Dolomite - New Insights into an Old Problem

John Andrew Higgins, PhD, Princeton University

The chemistry of shallow water marine carbonate sediments has been used to reconstruct the temperature and isotopic composition of seawater as well as the global carbon and oxygen cycles over >3 billion years of Earth history. The underlying assumption in these studies is that the chemical composition of the sediment accurately preserves a record of ancient open-ocean seawater. One of the principal ways in which this assumption is violated is through diagenesis - the chemical changes that occur during the transformation of sediment into rock. Diagenetic alteration of carbonate sediments can occur in association with either meteoric or marine fluids, is observed at length scales from microns to stratigraphic units, and may take the form of recrystallization (no change in mineralogy) or neomorphism (change in mineralogy). In some cases diagenetic alteration is regarded as the dominant source of variability in the geologic record (i.e. d18O) whereas in other cases (i.e. d13C) the effects of diagenesis are generally thought to be small though not insignificant. The effects of diagenesis, in particular early marine diagenesis, on many shallow water carbonate-bound geochemical proxies (e.g. d34S values of carbonate-associated sulfate (CAS), d44Ca, d26Mg, d7Li, d11B, d98/95Mo, d238/235U values, and I/Ca ratios) are either unknown or limited to a handful of studies. One of the more widely used metrics for characterizing diagenetic alteration in ancient carbonate rocks – the Mn/Sr ratio – can be shown to depend on variables other than early marine diagenesis (mineralogy, bottom water oxygenation, sedimentation rate, etc.) and at best is only a qualitative indicator.

Recent studies by Blättler et al., (2014) and Fantle & Higgins (2014) have shown both large variability and co-variation between the stable isotope ratios of magnesium (26Mg/24Mg) and calcium (44Ca/40Ca) in authigenic and diagenetic carbonates from the Monterey Formation (Miocene) in offshore southern California and the Marion Plateau off of northeast Australia (Ocean Drilling Program (ODP) Site 1196. In both cases the variability and co-variation are driven primarily by the extent to which diagenesis/authigenesis occurred in fluid-buffered ('open') or sediment-buffered ('closed') conditions. Additional studies on Ca and Mg isotope fractionation in carbonates have shown that it depends on mineralogy – aragonite tends to be depleted in 44Ca and enriched in 26Mg compared to calcite - as well as the rate of mineral precipitation and solution chemistry. Importantly, the observed variability in the Ca and Mg isotopic composition of bulk carbonate sediments associated with changes in mineralogy and early diagenesis is large compared to other potential sources (i.e. changes in seawater d26Mg and d44Ca values), suggesting that these measurements can provide novel insights into both the primary mineralogy and early diagenetic history of carbonate-bound geochemical proxies.

Using the Petroleum Research Fund from the American Chemical Society and additional funding from NSF I have assembed a large data set (N = 676) of d44Ca and d26Mg values in Neogene carbonate sediments and associated pore fluids from the Bahamas and the Eucla Shelf. The nine studied sites encompass a wide range of depositional, mineralogical, and diagenetic environments and were selected to explore whether these differences are associated with variations in sedimentary carbonate d44Ca and d26Mg values. Four of the sites are from cores drilled along a 25-km transect across the western flank of the Great Bahama Bank (GBB) from the platform top and margin to the toe of the slope. Sites along the transect document the effects of Neogene sea-level changes and platform progradation on the supply and diagenetic alteration of aragonitic platform-derived sediments. Previous studies have shown that these effects lead to a systematic and globally observed change in the d13C values of shallow water carbonate sediments that is decoupled from changes in the global carbon cycle as recorded by deep-sea carbonate sediments. Additional studied sites in the Bahamas include four shallow (<100 meter) cores taken from a north-south transect through the interior of the Little Bahama Bank. Miocene to Pleistocene-age carbonate sediments from these sites have experienced multiple episodes of diagenetic alteration including pervasive dolomitzation by seawater (35). Finally, a ninth site, International Ocean Drilling Program (IODP) Site 1131, a Quaternary carbonate succession on the uppermost slope adjacent to the Eucla Shelf (Great Australian Bight), was selected for comparison to the Bahamas. All studied sites are geologically young in that they are composed of carbonate sediments deposited over the last 10 million years. Reconstructions of seawater d44Ca and d26Mg values over this time period indicate that changes have been small (<0.2‰). As a result, the observed variability in sediment d44Ca and d26Mg values can be interpreted as largely due to mineralogy, diagenesis, and differences in isotopic fractionation due to precipitation rate.

At all studied sites we observe coherent stratigraphic variability in carbonate Ca and Mg isotopes that can be attributed largely to the effects of mineralogy and diagenesis under both fluid-dominated and rock-dominated conditions. The magnitude of the stratigraphic variability in Ca isotopes is large - >1‰ - and in many cases co-varies with bulk sediment geochemistry (Sr/Ca), d13C and d18O values. We interpret this variability as a consequence of changes in the extent to which the carbonate sediment has been neomorphosed and/or recrystallized under fluid-buffered or sediment-buffered conditions. Measured pore-fluid profiles of d44Ca and d26Mg values indicate that both fluid-dominated and rock-dominated carbonate diagenesis is ongoing on the western flank of the GBB and provides independent evidence that the rates of early marine diagenesis are sufficient to reset the chemistry of shallow water carbonate sediments on million-year timescales. This work demonstrates the importance of mineralogy and early diagenesis in shaping the geochemistry (d13C, d18O, major, minor, and trace elements) of shallow water carbonate sediments and highlights the utility of using paired measurements of Ca and Mg isotopes as a powerful tool to understand the paleo-environmental information stored in the carbonate rock record.