Trophic experiments to estimate isotope discrimination factors

摘要

In Caut, Angulo & Courchamp (2008a) rats were fed with experimental diets of distinct isotopic values (δ~(13)C and δ~(15)N) in order to infer the discrimination factors. We showed negative relationships between discrimination factors and diet isotopic values. In Caut, Angulo & Courchamp (2009), our aim was to generalize these relationships to other taxonomic groups with a view to providing ecologists with a general and flexible method to obtain discrimination factors for diet reconstruction studies when difficult to obtain otherwise.2.Perga & Grey (2010) claims that there is an artefact of experimental design (due to protein contents) and partly reanalyzes our data, taking into account the type of diet (mixed or not; carnivorous, omnivorous or herbivorous). However, (i) the role of protein content through isotopic routing is not clear; (ii) a reanalysis of our data shows that the relationships between discrimination factors and diet isotopic values remain significant even when following Perga's suggestions.3.Auerswald et al. (2010) criticized the imprecise definitions, mathematical artefacts and invalid statistical analysis of Caut, Angulo & Courchamp (2009). However, (i) their arguments are not new; they challenge the use of stable isotope in ecological trophic studies in general, and lack both alternative suggestions and ecological perspective; (ii) none of their statistical and mathematical objections has a significant effect on the results.4.Synthesis and applications. Neither reply reflects the current context of debate. The complementarity between experimental studies and their application in field studies is indispensable for the use of stable isotopes in ecology. Protocols should be very well assessed in order to represent field conditions with accuracy. Further analyses are necessary to remove some of the flaws in the use of stable isotopes (carrying out an in-depth analysis of isotopic routing, turnover, and dynamic during trophic experiments to ensure accurate attainment of equilibrium), but none of the issues raised by Perga & Grey (2010) and Auerswald et al. (2010) alter our main conclusions.