You are here:
Fish tissue lipid-C:N relationships for correcting ä13C values and estimating lipid content in aquatic food web studies
Citation:
Hoffman, J., M. Sierszen, AND A. Cotter. Fish tissue lipid-C:N relationships for correcting ä13C values and estimating lipid content in aquatic food web studies. Rapid Communications in Mass Spectrometry. Wiley InterScience, Silver Spring, MD, 29(21):2069–2077, (2015).
Impact/Purpose:
To undertake large-scale ecosystem assessments incorporating stable isotope analysis of carbon, it is necessary to standardize treatment of tissues for comparability. The difference in lipid content among tissue samples is an important confounding source of variability in carbon stable isotope ratios. Various mathematical approaches have been proposed to normalize carbon stable isotope ratios for lipid content. Here, we evaluate those approaches and propose a generalized normalization procedure that accurately estimates lipid content in tissue samples and provides a suitable carbon stable isotope ratio correction across a wide range of lipid content.
Description:
Normalizing 13C values of animal tissue for lipid content is necessary to accurately interpret food web relationships from stable isotope analysis. This is because lipids are 13C-depleted relative to proteins and carbohydrates, and because lipid content varies among species, tissue types, time and space. Various studies have explored using arithmetic mass balance rather than chemical extraction of lipids to normalize 13C values. Our goal was to evaluate two different, commonly-used models for estimating tissue lipid content (the mass-balance model based on C:N ratio, and that proposed by McConnaughey and McRoy (1979)) and to determine the effect of model choice on 13C values normalized using arithmetic mass balance. To do so, we used a collection of fish from Lake Superior spanning a wide range in lipid content (5% to 73% lipid). Lipid content was positively related to bulk muscle tissue C:N ratio. The models produced similar estimates of lipid content based on tissue C:N, within 6% for tissue C:N values <7. Normalizing 13C values using an arithmetic mass balance approach based on either model yielded similar results, with a small bias (<1‰) compared to chemical extraction. More broadly, the study allowed us to evaluate some basic elements influencing variability in lipid normalization models (extraction quality, choice of model parameters) and thereby provides some perspective on the achievable precision of normalized 13C values.
