Seasonal payoff variations and the evolution of cooperation in social dilemmas

TL;DR

This paper investigates how seasonal, periodic environmental changes influence the evolution of cooperation in social dilemmas, revealing that the frequency of payoff shifts and network structure critically affect cooperative outcomes.

Contribution

It provides a detailed analysis of how periodic environmental variations impact cooperation, highlighting the importance of change frequency and network topology in social dilemma dynamics.

Findings

Fast periodic changes between globally ordered games promote cooperation.

Slow periodic changes between global and local ordering games favor cooperation.

Network structure significantly influences cooperation, with lattices promoting local order and random graphs hindering it.

Abstract

Varying environmental conditions affect relations between interacting individuals in social dilemmas, thus affecting also the evolution of cooperation. Oftentimes these environmental variations are seasonal and can therefore be mathematically described as periodic changes. Accordingly, we here study how periodic shifts between different manifestations of social dilemmas affect cooperation. We observe a non-trivial interplay between the inherent spatiotemporal dynamics that characterizes the spreading of cooperation in a particular social dilemma type and the frequency of payoff changes. In particular, we show that periodic changes between two available games with global ordering best be fast, while periodic changes between global and local ordering games best be slow for cooperation to thrive. We also show that the frequency of periodic changes between two local ordering social dilemmas is irrelevant, because then the process is fast and simply the average cooperation level of the two is returned. The structure of the interaction network plays an important role too in that lattices promote local ordering, whilst random graphs hinder the formation of compact cooperative clusters. Conversely, for local ordering the regular structure of the interaction network is only marginally relevant as role-separating checkerboard patterns do not rely on long-range order.