The evolutionary consequences of learning under competition.

Area of Science:

• Evolutionary biology
• Behavioral ecology
• Game theory

Background:

• Learning is a widespread animal behavior crucial for individual adaptation and population variation.
• The role of social learning in driving evolutionary change remains poorly understood.
• Game theory models, like producer-scrounger and hawk-dove, are used to study social interactions.

Purpose of the Study:

• To investigate how reinforcement learning in social contexts influences evolutionary diversification.
• To model the evolution of biases in action valuation during resource competition.
• To explore the impact of learning on the evolution of abilities.

Main Methods:

• Developed game theoretical models of resource competition in small groups.
• Incorporated reinforcement learning mechanisms to control animal actions.
• Analyzed the evolution of subjective valuation biases and their effect on fitness.

Main Results:

• Biases in the subjective valuation of actions readily evolve through reinforcement learning.
• These biases can lead to disruptive selection on learning rules, potentially causing genetic polymorphisms.
• Learning can also drive disruptive selection on the ability to perform tasks.

Conclusions:

• Reinforcement learning in social contexts can be a significant driver of evolutionary diversification.
• Social learning dynamics can promote the evolution of novel traits and abilities within populations.
• This research highlights the interplay between learning, social behavior, and evolutionary processes.