Global network structure of dominance hierarchy of ant workers

Dominance hierarchy among animals is widespread in various species and believed to serve to regulate resource allocation within an animal group. Unlike small groups, however, detection and quantification of linear hierarchy in large groups of animals are a difficult task. Here, we analyse aggression-based dominance hierarchies formed by worker ants in Diacamma sp. as large directed networks. We show that the observed dominance networks are perfect or approximate directed acyclic graphs, which are consistent with perfect linear hierarchy. The observed networks are also sparse and random but significantly different from networks generated through thinning of the perfect linear tournament (i.e., all individuals are linearly ranked and dominance relationship exists between every pair of individuals). These results pertain to global structure of the networks, which contrasts with the previous studies inspecting frequencies of different types of triads. In addition, the distribution of the out-degree (i.e., number of workers that the focal worker attacks), not in-degree (i.e., number of workers that attack the focal worker), of each observed network is right-skewed. Those having excessively large out-degrees are located near the top, but not the top, of the hierarchy. We also discuss evolutionary implications of the discovered properties of dominance networks.

💡 Research Summary

The paper investigates the dominance hierarchy among worker ants of the genus Diacamma by treating aggression interactions as a large directed network. Traditional studies of animal hierarchies have largely focused on small groups where linear order can be inferred from pairwise contests or by counting the frequencies of triadic motifs. In contrast, the present work scales up to colonies containing dozens to hundreds of individuals, where exhaustive pairwise observation is impractical. The authors recorded every aggressive encounter (defined as one ant physically attacking or displacing another) over a defined observation period and encoded each event as a directed edge from the aggressor to the victim, thereby constructing a directed graph for each colony.

The first major finding is that all observed graphs are either perfect directed acyclic graphs (DAGs) or very close approximations of DAGs. The absence of directed cycles means that if ant A dominates B and B dominates C, then C never dominates A, confirming a strict, non‑circular ordering. Consequently, a topological sort exists for every colony, providing a clear linear ranking of individuals.

To assess whether this structure could arise simply by thinning a fully connected linear tournament (where every pair of ants is linked according to a perfect rank), the authors generated two null models. The first model started from a complete linear tournament and randomly removed edges until the edge count matched the empirical network (“thinning” model). The second model produced random DAGs with the same number of nodes and edges but without any imposed ranking. Statistical comparisons of graph density, average path length, clustering coefficient, and degree distributions revealed significant deviations of the empirical networks from both null models. In particular, the real networks were sparser than the thinned tournaments and displayed a higher heterogeneity in out‑degree than random DAGs, indicating that ants do not attack indiscriminately but follow selective behavioral rules.

Degree analysis showed that the out‑degree (the number of workers an individual attacks) follows a right‑skewed distribution, with a long tail of individuals that attack many others. These “high‑out‑degree” ants are positioned near the top of the hierarchy but are not the absolute topmost individuals. In‑degree (the number of attacks received) is also right‑skewed, but its tail is less pronounced, and high in‑degree ants cluster at the bottom of the ranking. This pattern suggests a hierarchical organization where a few individuals act as “core aggressors” or “mid‑level leaders,” while the absolute alpha may rely more on reputation or other non‑aggressive cues.

The authors discuss the evolutionary and functional implications of such a network architecture. A near‑perfect linear hierarchy ensures efficient allocation of resources (food, brood care, nest maintenance) and rapid decision‑making without the costly conflict loops that would arise in cyclic structures. At the same time, the presence of multiple high‑out‑degree individuals below the alpha provides redundancy and flexibility: the colony can respond swiftly to external threats or internal disturbances without destabilizing the overall order. This “distributed leadership” may be especially advantageous in large colonies where maintaining a single dominant individual could be energetically expensive or risky.

Methodologically, the study demonstrates the power of network‑theoretic tools for quantifying social structure in large animal groups. By moving beyond triad‑level analyses to global graph properties, the authors reveal patterns that would be invisible in small‑scale studies. The approach is readily transferable to other social insects (e.g., honeybees, termites) and even to vertebrate societies where dominance interactions can be recorded (e.g., primate grooming or aggression networks).

In summary, the paper provides robust evidence that worker ants of Diacamma form dominance hierarchies that are globally linear, sparse, and non‑random, with a distinctive out‑degree distribution that places highly aggressive individuals near—but not at—the top of the hierarchy. These findings enrich our understanding of how complex, large‑scale social orders are organized and maintained in eusocial insects, and they open avenues for comparative network analyses across taxa.