Toward the quantification of a conceptual framework for movement ecology using circular statistical modeling.

Area of Science:

• Movement ecology
• Animal behavior
• Biophysical modeling

Background:

• Existing random walk models for animal movement trajectories lack ecological interpretability and struggle with tortuous or oriented paths.
• Previous models often fail to account for the complex interplay between an animal's internal state and external environmental factors.

Purpose of the Study:

• To propose a novel, ecologically grounded movement model with interpretable parameters and mathematical tractability.
• To enhance the modeling of turning angles to better capture real-world animal movement patterns.
• To develop a framework for disentangling the effects of external factors (e.g., wind) from an animal's internal state on its movement.

Main Methods:

• Developed a new movement model based on a circular auto-regressive model, extending a generalized linear model to circular variables.
• Incorporated improved modeling of turning angles to capture more complex movement behaviors.
• Applied the model to GPS tracking data of seabirds (Calonectris leucomelas) to assess its applicability and parameter interpretability.

Main Results:

• The proposed model successfully captured diverse movement patterns in seabird GPS trajectories.
• Model parameters demonstrated clear ecological interpretations within a movement ecology framework.
• The model effectively distinguished the influence of external factors, such as wind, from the animal's directed movement decisions.
• Analysis revealed a seabird's navigation adjustments in response to wind during a homing flight.

Conclusions:

• The new movement model provides a more robust and ecologically relevant tool for analyzing animal trajectories.
• The ability to separate external influences from internal states offers significant advancements in understanding animal movement decisions.
• This model enhances the study of movement ecology by offering interpretable parameters and improved mathematical tractability.