LAUSR

Abstract Accurate trophic position (TP) estimates are important for the development of ecosystem-based management plans. TPs can be quantified by carbon (δ13C) and nitrogen (δ15N) stable isotopes in tissues, but these can disagree with observed and perceived feeding ecology. A recent method that has used a scaled diet-tissue discrimination factor (DTDF), reflecting the inverse relationship between DTDF and δ15N, was found to better describe TPs of predatory fish species in marine ecosystems, but this has not been tested in freshwater ecosystems. Here, we compare methods of TP estimations in the Lake Huron-Erie corridor (HEC), a system where high diversity of prey items has contributed to the concern that foraging ecology of piscivorous fish species is poorly understood. Using δ15N and δ13C, we quantified TP of longnose gar (Lepisosteus osseus), largemouth bass (Micropterus salmoides), and northern pike (Esox lucius) to assess the efficacy of a scaled DTDF compared to traditional DTDF isotope methods and stomach content analysis (SCA). The scaled DTDF method produced TP estimates that were at times consistent with SCA and were generally higher and with a greater range among individuals than non-scaled DTDFs. The scaled method was not sensitive to baseline choice nor influenced by incorporating carbon source in the model. Greater variability of TP estimates using a scaled DTDF suggests more complex trophic structuring in the upper trophic level guild of the HEC. These results, particularly the lack of baseline sensitivity, provide support for using the scaled DTDF in freshwater food web characterization.