The article presents a Bayesian approach to stable isotope mixing models for source partitioning in ecology. The authors introduce SIAR, an open-source R package that allows for the analysis of stable isotope data with uncertainty and variability. The method uses Bayesian inference to estimate the proportion of dietary sources contributing to a consumer's tissue, incorporating uncertainty in sources, trophic enrichment factors, and external variation. The model accounts for multiple sources and provides probability distributions for dietary proportions, allowing for more accurate and robust estimates. The authors compare SIAR with other models, such as MixSIR, and highlight its advantages in handling complex scenarios with multiple sources and uncertainty. The model is tested with simulated data, showing good performance in estimating true proportions within 95% credibility intervals. The study emphasizes the importance of incorporating variability and uncertainty in isotope analysis, and highlights the potential of SIAR as a tool for investigating complex dietary systems with greater quantitative rigor. The authors also discuss the limitations of the model, including assumptions about normal distribution of variability and the absence of isotopic routing within consumers. Overall, the study demonstrates the effectiveness of Bayesian methods in addressing the challenges of source partitioning in ecological studies.The article presents a Bayesian approach to stable isotope mixing models for source partitioning in ecology. The authors introduce SIAR, an open-source R package that allows for the analysis of stable isotope data with uncertainty and variability. The method uses Bayesian inference to estimate the proportion of dietary sources contributing to a consumer's tissue, incorporating uncertainty in sources, trophic enrichment factors, and external variation. The model accounts for multiple sources and provides probability distributions for dietary proportions, allowing for more accurate and robust estimates. The authors compare SIAR with other models, such as MixSIR, and highlight its advantages in handling complex scenarios with multiple sources and uncertainty. The model is tested with simulated data, showing good performance in estimating true proportions within 95% credibility intervals. The study emphasizes the importance of incorporating variability and uncertainty in isotope analysis, and highlights the potential of SIAR as a tool for investigating complex dietary systems with greater quantitative rigor. The authors also discuss the limitations of the model, including assumptions about normal distribution of variability and the absence of isotopic routing within consumers. Overall, the study demonstrates the effectiveness of Bayesian methods in addressing the challenges of source partitioning in ecological studies.