The ability of the phosphoric acid digestion technique to extract carbon dioxide from biogenic carbonates and reliably reproduce delta(18)O and delta(13)C signatures from standard reference materials (NBS-18, NBS-19) was tested and shown to produce accurate, unbiased measurements of non-biologic materials. The effects of roasting preparation methods commonly reported when analyzing biogenic carbonates were also tested in a series of experiments using reference standards and otoliths obtained from aquacultured Arctic charr and rainbow trout. Roasting had no effect on the isotope measurement of reference standards. No significant differences between mean oxygen isotope signatures from paired experiments with roasted and non-roasted fish otoliths were found. However, otolith oxygen isotope measurements were significantly enriched in comparison to rearing water-based measurements for both species. Agreement between expected isotopic equilibrium and measured otolith delta(18)O values varied as a function of roasting temperature and between species. Criteria for the selection of appropriate roasting temperatures are suggested and favour 350 degrees C in freshwater fish where unbiased estimates of average rearing water temperatures and known differences in rearing temperatures were obtained. Carbon isotopic disequilibria were observed for both species. A mixing model analysis established differences in the percentage of metabolically derived carbon in studied otoliths, with Arctic charr deriving a greater proportion of otolith delta(13)C from metabolism as a result of higher metabolic rates.
Copyright 2003 John Wiley & Sons, Ltd.