Calcium isotopic evidence for rapid recrystallization of bulk marine carbonates and implications for geochemical proxies

Strontium and calcium isotopic data for bulk carbonate solids and pore fluids from ODP Sites 1170 and 1171 are presented. The data suggest that bulk carbonate sediments actively exchange with coexisting pore fluids over tens of millions of year time scales. Recrystallization rates constrained by Sr isotopes and Sr elemental data are ∼3% per Ma at 1170A and ∼7% per Ma at 1171A. The pore fluid chemistries at both sites are affected by advection, which occurs in the downwards direction at 1170 (∼−25 m/Ma) and upwards at 1171A (∼250 m/Ma). Both the direction and the rate of advection are reflected by the width of the diffusive boundary layer for Sr at both 1170A (∼300 m) and 1171A (∼50 m), compared to ODP Site 807A (∼150 m) where no chemically-detectable advection is occurring. Recrystallization is supported not only by interpretations of pore fluid data, but also by the alteration of the bulk solid. This is especially true at 1171A, where advection drives significant alteration of Sr/Ca, Mg/Ca, and 87Sr/86Sr. Numerical simulations of pore fluid geochemical and isotopic evolution over tens of millions of years, conducted with a depositional, time-dependent reactive transport model, suggest that recrystallization rates in the upper tens of meters of the sedimentary section at both sites are more rapid than suggested by the Sr geochemical data. When the Sr-constrained rates are applied to the pore fluid Ca isotope data, the model does not predict pore fluid δ44Ca within analytical uncertainty. The simulations indicate rates that are initially ∼20% to 40% per Ma in young, <1 Ma sediments. The Ca isotope data cannot be explained by either inaccurate diffusion coefficients, inaccurate temporal evolution of pore fluid Ca concentrations, or upwards advection. Ultimately, such high rates in young sediments can impact paleoclimate and paleoenvironmental proxies used by geoscientists to study the past. Diagenetic effects due to rapid recrystallization, demonstrated for the oxygen isotope and Mg/Ca paleotemperature proxies, can alter paleotemperature reconstructions by as much as 4 °C. This suggests a means for affecting not only absolute temperature estimates but also systematic differences between the two paleotemperature tools.