Lipid correction for carbon stable isotope analysis of deep-sea fishes

Stable isotope analysis of fish tissue can aid studies of deep-sea food webs because sampling difficulties severely limit sample sizes of fish for traditional diet studies. The carbon stable isotope ratio (δ13C) is widely used in food web studies, but it must be corrected to remove variability associated with varying lipid content in the tissue. A lipid correction has not been determined for any deep-sea fish. These fishes are ideal for studying lipid correction because lipid content varies widely among species. Our objective was to evaluate an application of a mass balance δ13C correction to a taxonomically diverse group of deep-sea fishes by determining the effect of lipid extraction on the stable isotope ratios, examining the quality of the model parameters derived for the mass balance correction, and comparing the correction to published results. We measured the lipid extraction effect on the nitrogen stable isotope ratio (δ15N) and δ13C of muscle tissue from 30 North Atlantic species. Lipid extraction significantly increased tissue δ15N (+0.66‰) and δ13C values, but the treatment effect on δ13C was dependent on C:N, a proxy for lipid content. We compared the lipid-extracted δ13C to the δ13C predicted by the mass balance correction using model variables estimated from either all individuals (pooled) or species-by-species or using published values from other species. The correction using the species-by-species approach performed best; however, all three approaches produced corrected values that were generally within 0.5‰ of the measured lipid-free δ13C and that had a small over-all bias (<0.5‰). We conclude that a generalized mass balance correction works well for correcting δ13C in deep-sea fishes, is similar to that developed for other fishes, and recommend caution when applying a generalized correction to fish with high lipid content (C:N >8).