An information-theory approach to geometry for animal groups

Area of Science:

• Animal behavior
• Ethology
• Information theory

Background:

• Quantifying group-level social interactions in animals is challenging.
• Advances in GPS tracking and computational power offer new analytical possibilities.
• Understanding group geometry and interaction impacts on individuals remains an open problem.

Purpose of the Study:

• To present an information theory-based tool for analyzing animal group behavior.
• To demonstrate the tool's utility using GPS data from a Siberian husky pack.

Main Methods:

• Utilized an information theory framework to analyze group movement patterns.
• Applied GPS tracking data from 15 freely interacting Siberian huskies.
• Developed predictive models based on individual and group locations.

Main Results:

• Individual movement freedom was limited to 4%; locations were 96% predictable by group members.
• Dominant individuals showed less influence from others; same-sex individuals had stronger mutual influence.
• Kinship increased inter-individual dependency, and pack stability decreased before feeding.

Conclusions:

• Information theory coupled with bio-logging technology offers a powerful approach for studying complex animal social interactions.
• This method moves beyond dyadic analysis to understand group dynamics.
• Findings highlight significant constraints on individual movement within social groups.