The oxygen isotopic compositions (δ18O) of minimally altered phosphate minerals and fossils, such as conodont elements, are used as a proxy for past ocean temperature. Phosphate is thermally stable under low to moderate burial conditions and is ideal for reconstructing seawater temperatures because the P-O bonds are highly resistant to isotopic exchange during diagenesis. Traditional bulk methods used to measure conodont δ18O include multiple conodont elements, which can reflect different environments and potentially yield an aggregate δ18O value derived from a mixture of different water masses. In situ spot analyses of individual elements using micro-analytical techniques, such as secondary ion mass spectrometry (SIMS), can address these issues.
Here we present 108 new δ18O values using SIMS from conodont apatite collected from four Lower to Upper Ordovician stratigraphic successions from North America (Nevada, Oklahoma, and the Cincinnati Arch region of Kentucky and Indiana, USA). The available elements measured had a range of thermal alteration regimes that are categorized based on their conodont alteration index (CAI) as either low (CAI = 1−2) or high (CAI = 3−4). Though individual spot analyses of the same element yield δ18O values that vary by several per mil (‰), most form a normal distribution around a mean value. Isotopic variability of individual spots can be minimized by avoiding surficial heterogeneities like cracks, pits, or near the edge of the element and the precision can be improved with multiple (≥4) spot analyses of the same element. Mean δ18O values from multiple conodonts from the same bed range between 0.0 and 4.3‰ (median 1.0‰), regardless of low or high CAI values. Oxygen isotopic values measured using SIMS in this study reproduce values similar to published trends, namely, δ18O values increase during the Early−Middle Ordovician and plateau by the mid Darriwilian (late Middle Ordovician). Twenty-two of the measured conodonts were from ten sampled beds that had been previously measured using bulk analysis. SIMS-based δ18O values from these samples are more positive by an average of 1.7‰ compared to bulk values, consistent with observations by others who attribute the shift to carbonate- and hydroxyl-related SIMS matrix effects. This offset has implications for paleo-temperature model estimates, which indicate that a 4 °C temperature change corresponds to a 1‰ shift in δ18O (‰). Although this uncertainty precludes precise paleo-temperature reconstructions by SIMS, it is valuable for identifying spatial and stratigraphic trends in temperature that might not have been previously possible with bulk approaches.