Concentration-dependent mixing models predict values of diet-derived stable isotope ratios in fish otoliths

The use of artificially enriched stable isotopes to mark the calcified structures of fish is becoming a popular marking tool and is a valuable method for determining relative contributions of water and diet sources to otolith elements. Until now, source contribution studies on otoliths have relied on a linear mixing model. The incorporation of concentration dependence in mixing models, however, could allow more accurate estimates of the contributions of sources to a mixture. Here we experimentally determine the rate of incorporation and mark success in the otoliths of larval marine fish fed a spiked stable isotope diet using Ba-138 or Sr-86 over a four week period in a marine system. We then compare the experimental results with the estimates from a traditional linear mixing model as well as a concentration-dependent mixing model parameterized with marine values. The experiment demonstrated that dietary spiking of Ba-138 was successful and values stabilized after two weeks of exposure. Conversely, diet spiking of Sr-86 did not result in a detectable shift in otolith Sr isotope ratios, even after four weeks of exposure. Both results agreed with predictions from the concentration-dependent mixing model using measured concentrations of elements in sources for this marine system, whereas the linear mixing model failed to predict the observed shift in otolith Ba isotope ratios. These results support the use of dietary stable isotopes of Ba for efficient marking of fish calcified structures in marine systems as well as highlighting the importance of concentration dependence in source contribution mixing models for otoliths. (C) 2014 Elsevier B.V. All rights reserved.