Abstract
Measures of trophic position (TP) are critical for understanding food web interactions and human-mediated ecosystem disturbance. Nitrogen stable isotopes (δ(15) N) provide a powerful tool to estimate TP but are limited by a pragmatic assumption that isotope discrimination is constant (change in δ(15) N between predator and prey, Δ(15) N = 3.4‰), resulting in an additive framework that omits known Δ(15) N variation. Through meta-analysis, we determine narrowing discrimination from an empirical linear relationship between experimental Δ(15) N and δ(15) N values of prey consumed. The resulting scaled Δ(15) N framework estimated reliable TPs of zooplanktivores to tertiary piscivores congruent with known feeding relationships that radically alters the conventional structure of marine food webs. Apex predator TP estimates were markedly higher than currently assumed by whole-ecosystem models, indicating perceived food webs have been truncated and species-interactions over simplified. The scaled Δ(15) N framework will greatly improve the accuracy of trophic estimates widely used in ecosystem-based management.