Abstract
Rationale Carbon (δ 13 C) and nitrogen (δ 15 N) stable isotope analysis is a powerful tool to investigate diverse questions in fish ecology, such as their trophic position or migration strategies. These questions appear particularly important to protect endangered European eel. However, elevated lipid content in eel muscle can bias δ 13 C values, as lipids are 13 C‐depleted compared to proteins and carbohydrates. Methods We measured δ 13 C and δ 15 N values of bulk and lipid‐free samples of eel muscle. Lipid‐free samples were obtained after the extraction of lipids with cyclohexane. Lipid‐corrected δ 13 C values, using five different mathematical equations based on bulk δ 13 C values, were compared to lipid‐free δ 13 C values. We also evaluated the effect of lipid extraction on δ 15 N values. The analyses were based on linear regression performed on 333 individuals captured in nine lagoons and four rivers. Results Independently to the capture site or habitat (river or lagoon), the predicted lipid‐corrected δ 13 C values were highly consistent with the measured lipid‐free δ 13 C values ( R 2 > 0.90). The application of specific equations for each habitat or capture site only slightly increases these R 2 (1.5% or less). The lipid extraction treatment significantly decreased by 0.2‰ the δ 15 N values compared to bulk samples. Conclusions Given the excellent prediction of mathematical equations and the small decrease of δ 15 N values after lipids extraction, we propose to use mathematical correction to estimate δ 13 C values of eel muscle. As the habitats or sites did not strongly influence the results, the coefficients from our study can be applied to other studies on European eel.