Effects of extraction protocols on the oxygen isotope composition of conodont elements

• Oxygen isotopic results from conodonts record no statistically significant differences related to processing technique.
• Standard conodont extraction techniques using formic and acetic acid are safe for oxygen isotopic studies.
• Partially dissolved conodont elements record offset δ18O values and should be avoided in geochemical studies.
• Conodont element yields do not systematically vary between splits processed using unbuffered and buffered acid solutions.

Oxygen isotopic measurements of conodont elements can provide important paleoclimatic, paleoecological, and paleoenvironmental information. However, preparation techniques are often cited as a potential, albeit poorly constrained, source of error. Prior to isotopic measurements, conodont elements are typically liberated from a calcium carbonate matrix using dilute solutions of acetic or formic acid. These extraction techniques are well established for paleontological studies, but the protocols were established before conodonts became a common target for isotopic analyses. Whereas concerns are often raised about the possibility of bias introduced through phosphate leaching, especially with stronger acid, empirical observations regarding error in isotopic measurements introduced due to the preparation techniques used are limited and generally anecdotal.This study was designed to constrain the effects different types of acids used for carbonate digestion have on the measured oxygen isotopic compositions of bioapatite (specifically, conodont elements and brachiopod shells). We performed parallel analyses on splits from seven Late Ordovician samples that were processed with both buffered and unbuffered solutions of acetic and formic acid. In addition, to further assess the potential for δ18O alteration related to partial apatite dissolution, we forced partial dissolution of isolated conodont elements by placing them in an unbuffered formic solution for varying intervals of time prior to analysis. Measured δ18O values from conodont elements and brachiopod shell fragments show no statistically significant differences related to processing technique, but isolated conodont elements with visible evidence of partial dissolution due to prolonged exposure to unbuffered formic acid record consistently offset δ18O values. Such partially dissolved elements should be avoided in geochemical studies. If specimens exhibiting apparent etching are avoided, though, we conclude that standard digestion protocols for conodont element extraction from carbonates do not introduce a bias (either a systematic offset or an increase in variance) on measured δ18O values from phosphate-bound oxygen in bioapatite above the limits of the resolution of this study.