Incorporating concentration dependence in stable isotope mixing models

Area of Science:

• Stable isotope analysis
• Ecological modeling
• Biogeochemistry

Background:

• Stable isotopes (e.g., C, N) are natural tracers for quantifying source contributions in mixtures.
• Standard linear mixing models assume equal source contributions for all elements, which fails when elemental concentrations differ significantly.

Purpose of the Study:

• To develop and validate a concentration-weighted linear mixing model for stable isotope analysis.
• To improve the accuracy of source contribution estimations when elemental concentrations vary among sources.

Main Methods:

• Developed a concentration-weighted linear mixing model, generalizing to n elements and n+1 sources.
• Conducted sensitivity analyses for C and N in three-source scenarios.
• Applied the model to a case study of bears and a captive feeding study of mink.

Main Results:

• Varying N concentration significantly impacted estimated source contributions in sensitivity analyses.
• The concentration-weighted model provided more accurate dietary proportions for mink compared to the standard model.
• The standard model failed to yield valid proportions in the mink feeding study.

Conclusions:

• The concentration-weighted model is recommended when elemental concentrations differ substantially among sources.
• This model enhances applications in dietary studies, food webs, water source tracing, and soil dynamics.
• The model offers improved accuracy for ecological and geochemical stable isotope analyses.