The "Insect Prison Game" is a novel theoretical framework that synthesizes principles of evolutionary game theory with the behavioral ecology of eusocial and territorial insects. Unlike classical models such as the Prisoner’s Dilemma, which focus on binary cooperation versus defection, the Insect Prison Game introduces a tripartite strategic space: Escalate (Fight), Submit (Retreat), or Contain (Imprison). This paper defines the game’s payoff matrix based on empirical observations of ant raiding behavior, parasitic wasp host manipulation, and termite colony defense. We demonstrate that under conditions of resource scarcity and high relatedness, the "Contain" strategy becomes an evolutionarily stable state (ESS), leading to the formation of living prisons—functional but subjugated colonies. The model predicts that insect prisons emerge not as a pathology of conflict but as an optimal solution to the cost-benefit asymmetry of total annihilation.
| R \ D | Escalate | Submit | Contain | |-------|----------|--------|---------| | | (E_c, E_c) | (V, 0) | (V - C_c, -P) | | Submit | (0, V) | (V/2, V/2) | (0, V) | | Contain | (-P, V - C_c) | (V, 0) | (V/2 - M, V/2 - M) | insect prison game
The Insect Prison Game: A Model of Escalation, Cooperation, and Containment in Competitive Ecosystems The "Insect Prison Game" is a novel theoretical
3.3 Termite Colony Wars (Macrotermes bellicosus) In prolonged colony conflicts, termites sometimes block enemy soldiers into sealed chambers rather than killing them. These prisoners are not executed but starved or reabsorbed. This represents a "punishment" Containment strategy that deters future escalation without incurring direct combat costs. We demonstrate that under conditions of resource scarcity
3.1 Slave-Making Ants (Formica sanguinea) Empirical data show that F. sanguinea rarely kills defending F. fusca workers. Instead, they employ a "Contain" strategy: they raid pupae, bring them back, and the eclosing adults function as prison laborers. In IPG terms, Escalate (killing all defenders) yields short-term gain but loss of future labor. Contain yields long-term net benefit (V - M) > (V - C_c) when M is low.
[Generated for Academic Purposes] Journal: Journal of Theoretical Biology & Game Ecology (Hypothetical)
