Varying organic content in fish otoliths: Effects on SIMS-based δ18O measurements and possible corrections

Varying organic contents in otoliths have complex and sometimes counterintuitive effects on intra-otolith δ18O measurements. This is often addressed by roasting otoliths or applying fixed corrections, however, the underlying chemical and physiological mechanisms involved are poorly understood and have not been tested in a quantitative manner, potentially rendering such corrections unreliable. Using high-resolution secondary ion mass spectrometry (SIMS) measurements of δ18O values paired with OH/16O ion ratios as organic proxy, we derived quantitative relationships of these measurements over the entire life of a large sample of northern pike (Esox lucius) otoliths from freshwater and brackish habitats. We assessed OH/16O ion count ratio as an organic tracer, and estimated its relationship with δ18O determinations. We developed a pointwise correction approach that accounted for variations in otolith organic contents. OH/16O ion ratio profiles agreed with other organic proxies, confirming them as reliable tracer of organic content. We detected an inverse relationship between δ18O values and OH/16O ion ratio, with elevated OH/16O ion ratios near otolith cores. OH/16O ratios decreased with distance to the core. Pairwise corrections for the effect of OH/16O ratios on δ18O values resulted in a mean offset between uncorrected and corrected values of 0.52 ‰, suggesting an approximately 2°C bias towards warmer temperature if uncorrected data were to be used for δ18O thermometry. Simultaneous determination on organic- and inorganic-bound oxygen resulted in a negative offset of δ18O, which varies with the life history of individual fish. Varying offsets in δ18O values within individual life histories could be accounted for using our correction. We recommend future SIMS-based δ18O thermometry studies to estimate the local organic content to assess whether correction is warranted. We further offer more general recommendations on how future studies may assess whether corrections for organics are necessary. [Display omitted] •SIMS-measured OH/16O correlates closely with organic-associated elements.•Intraotolith organic content biases SIMS δ18O measurements towards lower values.•Organics in otoliths may bias temperature reconstruction in northern pike by 2 °C.•Pairwise correction with OH/16O offers a biologically realistic correction of otolith δ18O in SIMS approaches.