Collective Computation in Animal Fission-Fusion Dynamics.

Area of Science:

• Behavioral Ecology
• Social Systems
• Collective Computation

Background:

• Collective computation of social structure can reduce environmental uncertainty and aid adaptation.
• Fission-fusion social structures are common in primates, influencing foraging and social dynamics.

Purpose of the Study:

• To investigate how spider monkeys (Ateles geoffroyi) use social knowledge for collective computation of subgroup size distributions.
• To determine if these distributions are adaptive for foraging in variable environments.
• To assess individual strategies for staying or leaving subgroups based on social cues.

Main Methods:

• Analysis of time-series subgroup membership data from spider monkey groups.
• Development of a stochastic model of collective computation using individual stay-leave strategies.
• Simulation of various collective decision-making circuits to generate subgroup size distributions.
• Use of transfer entropies to measure the match between computed distributions and environmental food abundance.

Main Results:

• Spider monkeys employ multiple stay-leave strategies, indicating the importance of individual social knowledge.
• Individuals use information from multiple conspecifics to make decisions.
• The monkeys collectively compute subgroup size distributions that efficiently exploit ephemeral food sources.
• Simulations showed that artificially optimized distributions outperformed observed ones, suggesting adaptive lags or constraints.

Conclusions:

• Spider monkeys exhibit collective computation for adaptive foraging in dynamic environments.
• Individual social knowledge and collective decision-making are key to optimizing subgroup size.
• Measurement error, constraints, and adaptive lags may limit the efficiency of collective computation in this system.