The effects of social rank and payoff structure on the evolution of group hunting

Abstract

Group hunting is common among social carnivores, and mechanisms that promote this behavior are a central topic in evolutionary biology. Increased prey capture success and decreased losses from competitors are often invoked as factors promoting group hunting. However, many animal societies have linear dominance hierarchies where access to critical resources is determined by social rank, and group-hunting rewards are shared unequally. Despite this inequality, animals in such societies cooperate to hunt and defend resources. Game theoretic models predict that rank and relative rewards from group hunting vs. solitary hunting affect which hunting strategies will evolve. These predictions are partially supported by empirical work, but data needed to test these predictions are difficult to obtain in natural systems. We use digital evolution to test how social rank and tolerance by dominants of subordinates feeding while sharing spoils from group hunting influence which hunting strategies evolve in digital organisms. We created a computer-simulated world to reflect social and hunting dynamics of spotted hyenas (Crocuta crocuta). We found that group hunting increased as tolerance increased and as the relative payoff from group hunting increased. Also, top-ranking agents were more likely to group hunt than lower-ranking agents under despotic sharing conditions. These results provide insights into mechanisms that may promote cooperation in animal societies structured by dominance hierarchies.

Fig 1. Setup of groups for hunting games and possible offspring agent locations.

(A) N1 (black squares) is the location of a single agent. Grey squares represent the neighboring agents that N1 will play the 5 hunting games with during each world update. (B) The locations (grey squares) labeled with * show the possible offspring locations for a parent at location N1 (black square), as we assume offspring remain closer to their parent than do unrelated hyenas. Offspring replace the agent at their location, effectively killing off that agent.

Fig 2. Per capita payoffs under different experimental conditions.

Payoff scheme agents received for group hunting and solo hunting under different tolerance conditions (when the group-hunting payoff is 2 times the per capita payoff from solo hunting). Agents hunted in groups with four other agents and each agent decided whether to group hunt (i.e., hunt for large prey and share the reward with other agents that choose to group hunt) or solo hunt (i.e., hunt for small prey and the agent gets a smaller reward that is not shared). If an agent chose to group hunt, the payoff it received depended on its relative rank among other group hunters, the tolerance value, and how many other agents chose to group hunt. Agents received equal shares of the group-hunting payoff when tolerance = 1. As tolerance decreased, the payoff was increasingly skewed by dominance rank, with the highest-ranking agent receiving a greater proportion of the payoff. Agents received the same payoff from solo hunting regardless or rank under all tolerance conditions.

Fig 3. Group-hunting rates under varying tolerance conditions.

The average rate of group hunting in the agent population across different tolerance levels when the group-hunting payoff is (A): 2 times the solo-hunt payoff per capita and (B): 1.2 times the solo-hunt payoff per capita. Tolerance determined how equally the reward from group hunting was shared (i.e., 1 = shared equally, 0.64 = each agent gets 64% of the reward as the agent with the rank immediately above it). Each point represents the average group-hunting rate of the agents in their respective relative rank positions within hunting subgroups (e.g., rank 1 = highest ranking agent in the subgroup). Each tolerance condition was run for 1,000,000 updates (sampled every 1,000 updates), plots represent the mean of the last 100,000 updates (n = 101 samples) averaged across 200 replicates per condition. Bars represent ±2 standard errors of the mean group-hunting rates among agents of each relative rank across each tolerance condition.