Document Type


Date of Original Version





The stable isotope geochemistry of gorgonian octocoral skeletons facilitates detailed time series reconstructions of nutrient biogeochemistry. However, comparisons among reconstructions from different locations require realistic estimates of the uncertainty surrounding each measured geochemical value. Here, we determine quantitative uncertainties related to 1) standard skeletal pretreatment in preparation for stable isotopic analysis and 2) biological variability associated with a heterogeneous isotopic composition of the gorgonin skeleton. We found that the 5% HCl pretreatment required for the δ13C measurements does not significantly impact the δ15N values of the skeleton nor the reproducibility of the δ15N measurements. In contrast, while 5% HCl pretreatment significantly altered bulk δ13C values via removal of CaCO3, it did not change amino acid δ13C values in the organic skeleton. We found that the variance of repeat measurements of skeleton samples formed contemporaneously and homogenized skeleton for both δ13C and δ15N exceeded that of instrumental uncertainty of an acetanilide standard. This indicates that instrumental uncertainty underestimates the true precision of an isotopic measurement of the organic skeleton. Furthermore, measurements of contemporaneous skeleton around the circumference of an octocoral colony yielded variability exceeding that of homogenized skeleton. Based on these results, we find that 1) both δ13C and δ15N values can be measured simultaneously in pretreated skeleton, 2) growth bands should be homogenized prior to analysis, and 3) reported error should include uncertainty due to biological effects determined from repeat analysis of homogenized skeleton and not just instrument error to reduce false significant differences. Our results present an important protocol for processing proteinaceous octocoral skeletons and propagating uncertainty to more accurately reconstruct nutrient dynamics from proteinaceous deep-sea octocoral skeletons.