Trophic enrichment of deuterium in aquatic food webs? A cautionary note on the use of δD measurements to infer origins of fish

Stable-hydrogen isotope (δD) measurements of fish tissues have recently been shown to closely correlate with the δD of ambient waters, suggesting that δD measurements can be used to infer origins of fish. However, this approach assumes little trophic isotopic fractionation of hydrogen isotopes in aquatic foodwebs. We investigated whether trophic δD discrimination existed in three fish species: a carnivore (European catfish, Silurus glanis), an omnivore (common carp, Cyprinus carpio) and an herbivore (rudd, Scardinius erythrophthalmus) in the lower Ebro River, Spain. We found δD values increased significantly with fish length for catfish, a species with a known ontogenetic diet shift with increasing size. By contrast, no δD-length relationship was found for carp and rudd. A comparison of species in Flix reservoir provided the following decreasing order for δD values: catfish (-130.86‰) > carp (-140.91‰) > rudd (-158.16‰), that significantly differed by their potential trophic level. However, stable-nitrogen isotope ratios (δ¹⁵N), usually an indicator of trophic level, were not well correlated with δD values in catfish, but other trophic indicators such as tissue Hg and As concentrations were. Our findings suggest that trophic fractionation may occur, and so caution is required when using fish tissue δD to infer origins.