Diagenesis of lower Cretaceous pelagic carbonates, North Atlantic: Paleoceanographic signals obscured

Abstract

The stable isotope and minor element geochemistry of Neocomian (Lower Cretaceous) pelagic carbonates of the North Atlantic Basin (Deep Sea Drilling Project Sites 105, 367, 387, 391, and 603) was examined to develop a diagenetic model for pelagic limestones. In particular, we hoped to test the fidelity of whole-rock geochemical records as paleoceanographic indicators for pelagic deposits of pre-Aptian age, in which individual microfossils are not available for analysis. Data indicate that in addition to depth of burial, rhythmic variations in primary carbonate content have strongly controlled diagenetic patterns and associated geochemical signatures in these Neocomian sequences. Samples become increasingly depleted in Sr and 18O with increasing CaCO3 content. Within individual sedimentary sections, substantial decreases in Sr/Ca ratios and ??18O values are evident over a range of 4 to 98% CaCO3. However, even over a relatively narrow range of 50 to 98% CaCO3 a 2.5%c variation in ??18O values and a change of a factor of 1.7 in Sr/Ca ratios are observed. Carbon isotope compositions do not vary as extensively with CaCO3 content, but carbonate-rich intervals tend to be relatively depleted in 13C. Petrographic analysis reveals that these geochemical patterns are related to the transfer of CaCO3 from carbonate-poor intervals (calcareous shales and marlstones) to adjacent carbonate-rich intervals (limestones) during burial compaction and pressure solution. This process results in the addition of diagenetic cement to carbonate-rich intervals to produce a bulk composition that is relatively depleted in Sr and 18O and, at the same time, enables the retention of more-or-less primary carbonate that is relatively enriched in Sr and 18O in adjacent carbonate-poor intervals. Thus, although cyclic variations in CaCO3 content are primary in the Neocomian sequences examined, measured variations in Sr/Ca ratios and ??18O values are not and, as such, do not provide reliable proxies for past variations in climate, oceanographic conditions, or global ice volume.