Diagenetic effects on conodont oxygen-isotopic compositions: implications for in-situ analysis and paleothermometry

Conodont oxygen-isotope (δ18O) paleothermometry is widely used in paleoclimate studies, yet significant δ18O variability among conodont tissue types (i.e., albid crown vs. hyaline crown vs. basal body) implies a substantial diagenetic and/or ontogenetic influence, but the specifics of this relationship remain poorly constrained. Here, we conducted high-resolution secondary ion mass spectrometry analysis of various tissues and layers (i.e., outer surface vs. interior axial) on six coniform elements (108 total data points) of Early Ordovician paraconodonts. In addition, we undertook micro-laser Raman spectral analysis on interior layers of the Ordovician paraconodonts, and laser ablation multi-collector inductively coupled plasma mass spectrometric analysis of both Ordovician coniform elements, and pectiniform and ramiform elements of Early Triassic euconodonts as a supplement. Our results reveal variable degrees of post-depositional alteration for stratigraphically adjacent specimens as well as among the various tissues and layers of individual specimens. Raman spectral and petrographic studies reveal greater crystallinity of albid crown and hyaline crown relative to darker-colored basal body, which is characterized by more abundant skeletal microporosity and defects related to the presence of amorphous carbon, resulting in greater radiogenic Sr uptake during diagenesis. The more strongly recrystallized albid and hyaline crown tissues show lower and more scattered δ18O values (+13.5 ‰ to + 17.8 ‰, avg. + 15.9 ± 0.8 ‰) as well as lower G-band FWHM (i.e., full width at half maximum, mostly < 65 cm−1) and D1/G ratio (i.e., ratio of D1-band to G-band peak intensities; mostly < 0.7), exhibiting strong positive correlations of δ18O to D1/G as well as v1-PO3- 4FWHM. In contrast, the more coarsely crystalline basal body shows systematically higher and narrower δ18O ranges (∼+15.1 ‰ to ∼+17.5 ‰, avg. + 16.6 ± 0.5 ‰), and larger G-band FWHM (mostly > 65 cm−1) and D1/G ratios (mostly > 0.7) that are not correlated to δ18O. The axial versus outer layers of conodonts exhibit similar patterns to albid/hyaline crown versus basal body in weakly altered specimens. Our findings demonstrate that δ18O variation in conodonts is predominantly controlled by diagenesis, with ontogenetic effects playing a minor role. Consequently, conodont δ18O values, even from low thermal maturity specimens, should be interpreted with caution as a primary paleoenvironmental proxy. Due to widespread diagenetic alteration, published paleotemperature estimates derived from light-colored albid/hyaline crowns require a systematic correction on the order of ∼+0.8‰ to + 1.3‰, although relative δ18O trends over time remain robust. The results of this study of diagenetic effects on conodont elements validate the general reliability of the conodont oxygen-isotope paleothermometer while establishing optimized analytical and calibration protocols. These advances provide a framework for more precise deep-time paleotemperature reconstructions, thereby enabling integrated studies on the co-evolution of paleoclimate, paleoenvironment, and paleobiology.