Caution on isotopic model use for analyses of consumer diet

Isotopic models are increasingly used to determine the relative contribution of different food sources to an animal's diet. However, these models are based on restrictive assumptions and provide estimates rather than exact values of contributions to consumer diet. The sources of inaccuracies in isotopic models are not well understood and laboratory experiments may be useful to evaluate model performance. In this paper we assess the accuracy of the three main isotopic models in controlled laboratory experiments, involving a breed of Norway rats (Rattus norvegicus (Berkenhout, 1769)), in which the isotopic values of resources are known. At the same time, we measure errors resulting from the use of fixed or specific discrimination factor values for each resource and tissue. We show that the results of the three main isotopic models deviate considerably from the correct values in some cases. Estimations obtained using specific discrimination factors (corresponding to each experimental diet and tissue) were more accurate than those obtained using fixed discrimination factors (obtained from the literature). In addition, estimations varied depending on the tissue used, with the liver giving more accurate results than muscle or hair. We discuss the assumptions and limitations of isotopic models and highlight the importance of taking these assumptions into account when the most accurate results are sought. Finally, we propose some recommendations for the correct use of isotopic models, emphasizing the need to use specific discrimination factors for each species, tissue, and diet isotopic value.