Bringing the analysis of animal orientation data full circle: model-based approaches with maximum likelihood

Area of Science:

• Ecology
• Zoology
• Statistics

Background:

• Animal orientation studies frequently use directional data analyzed with circular statistics.
• Existing circular statistical tests have limitations, especially for multimodal data.
• Likelihood-based inference is underutilized in animal orientation research.

Purpose of the Study:

• To address the rarity of likelihood-based inference in animal orientation.
• To introduce the R package CircMLE for maximum likelihood analysis of circular data.
• To provide a convenient tool for applying model-based approaches in orientation studies.

Main Methods:

• Discussion of assumptions and limitations of common circular statistical tests.
• Development and implementation of the CircMLE R package.
• Application of CircMLE to simulated and empirical datasets, including Chinook salmon.

Main Results:

• CircMLE facilitates maximum likelihood analysis of circular data.
• The package demonstrates utility on both simulated and real-world orientation data.
• Provides a practical interface for advanced statistical modeling in this field.

Conclusions:

• CircMLE enhances the application of likelihood-based methods in animal orientation.
• The R package simplifies complex statistical analyses for researchers.
• Promotes more robust hypothesis testing, especially for complex data patterns.