|
Abstract:
|
Stable isotope sourcing is used to estimate proportional contributions of sources to a mixture, such as in the analysis of animal diets, plant nutrient use, geochemistry, pollution, and forensics. For modeling, animal ecology is considered because of its complexities. Parameter estimation has been a challenge because there are often many parameters (sources) and few data (isotopes) leading to an underconstrained linear system for the diet probability vector. We explicitly model subject-specific diets because diet differences lead to isotopic differences of individuals within a population. We propose a Bayesian "random effects" model based on documented first principles to estimate the mean population diet, subject-specific diets, and other parameters. We show that our model provides more precise diet estimates than existing models when data informing measurement error is not discarded.
|
