Evaluating Bayesian stable isotope mixing models of wild animal diet and the effects of trophic discrimination factors and informative priors

Abstract

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

Short‐ and long‐term diets of the threatened longhorned pygmy devil ray, Mobula eregoodoo determined using stable isotopes

“…In the analyses, tissue type (muscle and liver) was a fixed factor to compare long‐ and short‐term resource use, as muscle tissue turnover rates are approximately twice that of liver (Madigan et al ., 2012). To account for the uncertainties of DTDFs, a precautionary approach was taken by including a large uncertainty around this value (Swan et al ., 2020). To this end, the standard deviations of both Δ 13 C and Δ 15 N were included from the sources mentioned earlier, especially because DTDFs of zooplanktivores can be large relative to other trophic levels ( e.g., Barton et al, 2019; Wyatt et al, 2019).…”

“…The trophic position of mature adult M. eregoodoo was calculated using In the analyses, tissue type (muscle and liver) was a fixed factor to compare long-and short-term resource use, as muscle tissue turnover rates are approximately twice that of liver (Madigan et al, 2012). To account for the uncertainties of DTDFs, a precautionary approach was taken by including a large uncertainty around this value (Swan et al, 2020). To this end, the standard deviations of both Δ 13 C and Posterior distributions were taken from models run with 10 5 iterations, with the first 10 4 iterations discarded as burn-in.…”

Combining stable isotopes, morphological, and molecular analyses to reconstruct the diet of free‐ranging consumers

“…We evaluated if the use of priors in SIA improved the precision of diet estimates based on the range of credible intervals around the mean for each food source between models with and without priors (O'Donovan et al, 2018). We retained in the results' section only SIA whose priors were included given that they reflect the optimal use of SIA (Phillips et al, 2014;Swan et al, 2020). Diet estimates for SIA without prior are available in Appendix 2.…”

“…Still, our findings and those of others suggest that SIA are a powerful tool for dietary studies, especially those addressing the contribution of distinct food sources or food categories over longtime scales and for longitudinal monitoring. Nevertheless, their use should be cautious depending on the studied system and research objectives because reliable estimates with SIA require a minimum of a priori information about the consumer's foraging patterns (Moore & Semmens, 2008;Swan et al, 2020).…”

“…The optimal use of SIA also requires a minimum of a priori information on the consumer diet and/or on potential food sources available, questioning its use as a single approach to describe diet (Caut et al, 2008;Phillips et al, 2014). Such a priori information, used as priors in Bayesian-based SIA (Phillips et al, 2014), can be derived from the literature or preferably from results obtained with other methods of diet description in the same consumer populations, thereby allowing considering regional specificities in food habits (Chiaradia, Forero, McInnes, & Ramirez, 2014;Franco-Trecu et al, 2013;Swan et al, 2020). Because many studies are relying solely on SIA to determine the dietary choice of targeted species (but see, e.g., Horswill et al, 2018;O'Donovan et al, 2018), it is highly relevant to investigate how diet estimates based on SIA match those derived from morphological and molecular approaches for a shared period of time.…”