Effects of lipid extraction and the utility of lipid normalization models on δ13C and δ15N values in Arctic marine mammal tissues

Abstract

Animals store lipids, which are 13C-depleted, in their tissues that often must be extracted to correctly interpret δ13C data. However, chemical lipid extraction (CLE) can alter δ15N values and lipid normalization (LN) models are not consistent across fauna. We determined whether lipids should be extracted by assessing effects of CLE and validating LN models for liver and muscle from seven and eight marine mammal species, respectively, and skin from one species. In liver, CLE significantly increased δ13C and δ15N values for all species, whereas only a significant increase in δ13C occurred in skin. For muscle, δ13C and δ15N values were generally greater after CLE, but this was not consistent across species. Extracted lipids were depleted by approximately 7 and 5 ‰ for δ13C and δ15N, respectively, in both muscle and liver compared with protein in all species. The reliability of LN models varied between tissues and species; thus, their use is largely dependent on the precision of stable isotope values needed to address the objectives of a study. A decision framework to decide whether CLE or LN models is required for ecological interpretation of stable isotopes based on species, tissue and study objectives is presented.