Integrating Diverse Evolutionary Patterns of Collective Animal Behaviours into a Unified Selfish Herd Model

The behavioural adaptation of group-living animals in predation has been commonly illustrated as a united resistance against predators. However, fewer investigations have focused on whether individual fitness always conforms to group benefit in evolution. This study highlights the impact of individual selection towards the evolution of collective patterns, treating prey as purely egoistic agents unconcerned with predator fitness. Utilising the game-theoretic framework and computational approaches, we identify evolutionarily stable strategies and collective patterns under a complete range of environmental risk distributions. Our findings reveal that collective motion in groups is stable only when leaders face greater danger than followers. Additionally, we observe that the selfish herd scenario, both with and without crowding effects, is consistent with our results. Surprisingly, collective motion can also be stable in less commonly mentioned scenarios, like when the group centre is at more risk than the border. These findings imply that certain natural animal behaviours may be driven solely by individual selection, with minimal influence from higher-level mechanisms, redefining our understanding of group dynamics in predation.