Evolution of learned strategy choice in a frequency-dependent game.

Area of Science:

• Evolutionary Game Theory
• Behavioral Ecology
• Agent-Based Modeling

Background:

• Frequency-dependent games involve strategy choices that can be innate or learned.
• Previous research on producer-scrounger games suggested learned strategies are common, but theoretical models indicated learning might not spread through a population.
• A gap exists in understanding the evolutionary dynamics of learned strategy adoption in populations.

Purpose of the Study:

• To investigate the conditions under which learned strategy choice can evolve and spread throughout an entire population in the producer-scrounger game.
• To test the efficacy of two distinct learning rules for strategy adoption using computational simulation.
• To determine the influence of environmental fluctuation and phenotypic asymmetry on the evolution of learned behaviors.

Main Methods:

• Utilized an agent-based evolutionary simulation to model the producer-scrounger game.
• Implemented and tested two general learning rules governing strategy choice.
• Analyzed the impact of varying degrees of environmental fluctuation and player phenotypic asymmetry on evolutionary outcomes.

Main Results:

• Learned strategy choice was acquired by all individuals in the simulation when environmental fluctuation was sufficiently high or when players exhibited phenotypic asymmetry.
• In stable environments or without significant phenotypic differences, innate strategy choice tended to become fixed.
• The study identified specific conditions promoting the widespread evolution of learned behaviors.

Conclusions:

• Learning can readily evolve and become fixed in the entire population under biologically plausible conditions, particularly with environmental variability.
• Phenotypic asymmetry plays a crucial role in facilitating the evolution of learned strategy choice, especially in less dynamic environments.
• The findings challenge previous theoretical limitations and highlight the adaptive potential of learned behaviors in evolutionary game dynamics.