A novel method for identifying behavioural changes in animal movement data.

Area of Science:

• Ecology
• Animal Behavior
• Quantitative Biology

Background:

• Analyzing animal movement is crucial for understanding how organisms navigate complex environments.
• Statistical challenges in movement analysis include data dimensionality, autocorrelation, measurement error, and irregular sampling intervals.
• Animal movement data often reflect heterogeneous behaviors that are difficult to model using traditional methods.

Purpose of the Study:

• To introduce a novel statistical method, behavioral change point analysis (BCPA), for analyzing animal movement data.
• To develop a likelihood-based approach for identifying significant structural changes in movement patterns.
• To provide a robust and computationally efficient tool for uncovering hidden behavioral structures.

Main Methods:

• Modeling animal movement as a subsampling of continuous stochastic processes.
• Implementing a likelihood-based behavioral change point analysis (BCPA).
• Assessing the robustness of BCPA to data gappiness and measurement error.

Main Results:

• BCPA successfully identified significant structural changes in movement data.
• Application to a northern fur seal (Callorhinus ursinus) GPS track revealed a complex diurnal behavioral profile.
• The method demonstrated robustness to greater errors typical of ARGOS tracking systems.

Conclusions:

• Behavioral change point analysis (BCPA) is an effective and robust method for analyzing animal movement data.
• BCPA enhances the empirical interpretation of movement tracks, revealing complex behavioral patterns.
• This approach can facilitate the development of mechanistic behavioral models for animal movement.