Bayesian estimation of predator diet composition from fatty acids and stable isotopes

Area of Science:

• Ecology
• Quantitative Ecology
• Bioinformatics

Background:

• Stable isotope (SI) and fatty acid profile (FAP) analyses are key tools for diet assessment in ecology.
• Integrating SI and FAP data for quantitative diet estimation has been historically limited and qualitative.

Purpose of the Study:

• To develop a novel Bayesian mixing model for quantitative analysis of fatty acid profiles (FAP) in predator diets.
• To integrate FAP models with existing stable isotope (SI) models for enhanced ecological resolution.
• To provide an open-source R package for applying these advanced diet estimation methods.

Main Methods:

• Developed a Bayesian mixing model for FAP analysis, mirroring established SI mixing model frameworks.
• Incorporated fat content and conversion coefficients to refine diet proportion estimates.
• Applied the model to simulated and experimental datasets to demonstrate performance and utility.

Main Results:

• The proposed Bayesian FAP model provides robust point estimates and probability distributions for diet proportions.
• Model integration with SI analysis significantly increases the resolution of predator-prey relationship studies.
• Demonstrated successful application and performance through simulations and experimental data analysis.

Conclusions:

• The novel Bayesian FAP mixing model offers a powerful, quantitative approach to diet reconstruction.
• This framework enhances the integration of multiple dietary analysis techniques for ecological research.
• The open-source R package facilitates broader adoption and advancement in quantitative diet studies.