RETRIEVING RELATIVE PALEOTEMPERATURES FROM OXYGEN ISOTOPE RATIOS OF EARLY PALEOZOIC CARBONATES

Attempts to establish Paleozoic ocean temperatures using oxygen isotope ratios measured within least-altered marine carbonates have been confounded by three effects. First, the oxygen isotopic composition of ancient oceans is unknown, and may have varied significantly over Earth’s history (Jaffrés et al., 2007). Second, oxygen isotope ratios may be original but reflect changes in paleosalinity in addition to paleotemperature (Samtleben, 1996; Bickert, 1997). Third, original ratios may be altered to lower values by post-depositional interaction with isotopically light fluids derived from meteoric waters at burial metamorphic temperatures (Knauth and Kennedy, 2009).

Previous authors have incorporated one or more of these effects in their models in order to explain covariation of δ¹⁸O and δ¹³C observed in the early Paleozoic, even where strata have been exposed to low-grade regional metamorphism (Jeppsson et al., 2007). We are experimenting with the use of clumped heavy isotopes of carbon and oxygen to assess the temperature differences between δ¹³C maxima and minima in well-studied Silurian chemostratigraphic profiles. Linear correlations between oxygen and carbon isotope ratios in Silurian and Devonian carbonates have similar slopes, suggesting that a relative temperature signal may survive post-depositional alteration. In each period, the isotopically heaviest oxygen values are associated with the peaks of large positive carbon isotope excursions, representing times of increased carbon sequestration and CO₂ drawdown (Cramer and Saltzman, 2007). Applying this approach to diagenetically-shifted δ¹⁸O values from the late Cambrian SPICE positive carbon isotope excursion in Siberia (Kouchinsky et al., 2008) implies global cooling during this demonstrably worldwide event.