Comparing estimators of predator diets

What is it about?

Knowledge of animal diets is critical for understanding their ecology. We developed and compared several estimators of diet based on a method called quantitative fatty acid signature analysis (QFASA), expecting that one estimator would have superior performance and could be recommended for use with confidence. When model assumptions were met (in computer simulations), several estimators had similar performance, though estimators based on the Aitchison distance measure were modestly superior. However, diet estimates based on real data (not simulated) from polar bears were quite different, suggesting that estimators do not perform similarly when model assumptions are violated.

Featured Image

Why is it important?

This research presented new diet estimators and revealed that estimators respond differently to violations of model assumptions. These new findings are expected to stimulate additional research that will hopefully result in a better understanding of these methods and lead to improved estimation methods.