SIAR Mixing Model
Stable Isotope Analysis in R (SIAR) Mixing Model · Also known as: isotope mixing model, Bayesian mixing model, source apportionment, diet analysis
The Stable Isotope Analysis in R (SIAR) mixing model is a Bayesian framework for estimating the proportional contributions of dietary sources to a consumer, using stable isotope ratios. Developed by Parnell and colleagues (2010) and implemented in the R package siar (and its successor MixSIAR), this method integrates isotopic data from potential food sources and consumers to infer diets. It accounts for uncertainty in isotope fractionation (the shift in isotope ratios between diet and tissue) and natural variation among source populations, producing probability distributions rather than point estimates of diet composition.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use SIAR mixing models to estimate the diet of wild animals or humans when direct dietary observation is impractical or impossible. Requires stable isotope data from consumers and potential sources, and known (or estimated) fractionation factors. Works well when sources are isotopically distinct and the number of sources is not too large (typically 3-5 sources for robust inference with 2 isotopes). Not suitable when sources overlap extensively in isotope space or when fractionation is highly variable.
Strengths & limitations
- Provides probabilistic estimates (credible intervals) rather than point estimates, quantifying uncertainty in diet reconstruction
- Bayesian framework naturally incorporates prior information and allows formal hypothesis testing via model comparison
- Accounts for variation in source isotope values and fractionation factors through hierarchical modeling
- Non-invasive (requires only tissue samples, not stomach contents or observation) and applicable to extinct species via fossils
- Can be extended to multiple isotopes and to incorporate elemental concentration data
- Accuracy depends critically on accurate fractionation factors; errors in fractionation propagate to diet estimates
- Sources must be sufficiently separated in isotope space; overlapping sources yield uninformative posterior distributions
- Assumes that fractionation is constant across consumer individuals and populations; violation leads to biased estimates
- Cannot distinguish between sources if they have identical or highly similar isotopic signatures (e.g., two plants with the same photosynthetic pathway)
- Requires large sample sizes for stable posterior distributions; small sample sizes yield wide credible intervals
Frequently asked
How do I choose the right fractionation factors for my study?
Use published values if available for your consumer taxa and tissue type. If not, published meta-analyses (e.g., Post 2002) provide ranges. Conduct sensitivity analysis: rerun models with different plausible fractionation values to assess how results change. If fractionation is highly uncertain, incorporate that uncertainty into the model as a random effect rather than treating it as fixed.
What should I do if my dietary sources overlap in isotope space?
Add more isotopes (δ34S, δ2H, or δ18O) if possible; more isotopes increase the dimensionality of isotope space and may separate overlapping sources. If sources remain inseparable, acknowledge that the model cannot distinguish them; consider combining them or using additional data (DNA, fatty acids) to supplement isotopes. Overlapping sources lead to wide posterior distributions reflecting genuine uncertainty.
How many source samples do I need to estimate their isotope signatures?
At least 5 to 10 samples per source is a rough guideline to estimate mean and variance. For rare sources or those with high spatial variation, more samples are preferable. Report standard deviations or confidence intervals for each source's mean isotope values and incorporate this uncertainty into the mixing model.
Can I use SIAR to estimate the diet of humans?
Yes. SIAR has been applied to dietary reconstruction in archaeological populations and modern nutritional studies. Use appropriate fractionation factors for humans and include all plausible food sources. Account for cooking effects on isotope values if relevant. SIAR provides probabilistic diet estimates but cannot pinpoint individual-level diets—only population-level averages.
How do I interpret posterior distributions for diet proportions?
Report the median (or mean) diet proportion and credible interval (e.g., 95% interval). Wide intervals reflect high uncertainty and often indicate that sources are not well-separated or sample sizes are small. Posterior distributions may be multimodal, indicating that multiple diet combinations are consistent with the data. Compare distributions across groups using overlapping credible intervals or formal Bayesian hypothesis tests.
Sources
- Parnell, A. C., Inger, R., Bearhop, S., & Jackson, A. L. (2010). Source partitioning using stable isotopes: coping with too much variation. PLoS ONE, 5(3), e9672. DOI: 10.1371/journal.pone.0009672 ↗
- Jackson, A. L., Inger, R., Parnell, A. C., & Bearhop, S. (2011). Comparing isotopic niche widths among sympatric species: the role of phylogenetic relatedness. Ecology Letters, 14(8), 841-851. link ↗
- Phillips, D. L., Inger, R., Bearhop, S., Jackson, A. L., Moore, J. W., Parnell, A. C., Semmens, B. X., & Ward, E. J. (2014). Best practices for use of stable isotope mixing models in food-web studies. Canadian Journal of Zoology, 92(10), 823-835. DOI: 10.1139/cjz-2014-0127 ↗
How to cite this page
ScholarGate. (2026, June 3). Stable Isotope Analysis in R (SIAR) Mixing Model. ScholarGate. https://scholargate.app/en/ecology/siar-mixing-model
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Bioaccumulation ModelEcology↔ compare
- Food Web TopologyEcology↔ compare
- Leslie MatrixEcology↔ compare
- Population Viability AnalysisEcology↔ compare