AbstractNitrogen isotope analyses of amino acids (δ15N‐AA) are being increasingly used to decipher trophic dynamics. Interpretation of δ15N‐AA in consumers relies on the assumption that consumer physiological status and nutritional status of prey have negligible influences on the trophic discrimination factor (TDF), hence a constant TDF value is used in trophic position (TP) equations. Recent experiments have shown that this is not always the case and there is also a need to validate derived TP estimates in the field. We take advantage of the uniquely long time series of environmental monitoring data and archived (frozen) samples from the species‐poor Baltic Sea. We analyzed δ15N‐AA in similar sized individuals of cod and in its prey herring from four decades, 1980–2018; including time periods where dramatic reduction in condition status of cod has occurred. We expected that TDF in trophic AAs would increase during periods of poor cod condition, resulting in inflated TP estimates. We found that calculated TP and empirical estimates of TDF (difference in δ15N in trophic AAs between cod and herring) for cod increased in recent decades and that this was linked to condition status, herring (prey) lipid content and the hypoxic state of the ecosystem. Statistically adjusting TP for condition and prey lipid content as well as environmental stress (hypoxia) resulted in lower cod TP which better resembled the observed decrease in herring TP in recent decades. TP calculated from stomach analysis data in cod individuals over the same period showed no trend over time and confirmed that adjusted TP estimates mirror the real dietary TP better than unadjusted. By simultaneously measuring condition/nutritional status in both predator and prey it is possible to adjust for them as confounding variables and decipher actual consumer TP, partly overcoming the issues of unknown and variable TDF‐values. Our study also highlights the importance of including environmental stressors (here hypoxia) when interpreting TP and reconstructing food webs.
Read full abstract