Dominance Hierarchies: How Animal Societies Organise Rank
Topic: Animal Behaviour & Ethology Subject: Dominance, Social Rank, and the Pecking Order Last verified: 2026-09-23
Across many group-living species, from paper wasps and cichlid fish to hyenas and macaques, repeated contests between the same individuals settle into predictable patterns in which one animal consistently gives way to another. These patterns, known as dominance hierarchies, shape who eats first, who breeds, and how stressful daily life is. Since the Norwegian zoologist Thorleif Schjelderup-Ebbe described the "pecking order" of domestic hens in 1922, dominance has become one of the most widely used and most debated concepts in behavioural biology [Schjelderup-Ebbe 1922; Drews 1993]. Researchers have since built statistical tools to measure hierarchies, run experiments on how ranks form, tracked the lifelong fitness costs of low rank, and shown that social status can reach down to the level of immune-cell gene expression [Neumann et al. 2011; Strauss & Holekamp 2019; Snyder-Mackler et al. 2016].
This article covers what dominance is and how it is measured, how ranks form and are sometimes inherited, what rank means for survival, reproduction and health, and why one of the best-known examples, the "alpha wolf", turned out to be an artefact of captivity. It closes with the practical implications for managing animals in captivity and for conservation.
Defining and Measuring Dominance
Schjelderup-Ebbe's study of domestic chickens, published in 1922 in the Zeitschrift für Psychologie, recorded which hens pecked which within flocks and found that these relationships were consistent and could be arranged into an ordered list, the Hackliste or pecking list. It was the first detailed description of a dominance order in any animal, and the term "pecking order" entered both scientific and everyday language [Schjelderup-Ebbe 1922; Drews 1993].
Over the following decades, "dominance" came to be used in many incompatible ways. In an influential review, Carlos Drews examined thirteen published definitions. Some treated dominance as a trait of an individual and some as a property of a relationship; some were based on aggression, some on priority of access to resources, and some on theoretical constructs rather than observable behaviour [Drews 1993]. Drews argued that Schjelderup-Ebbe's original approach had the greatest descriptive value. He proposed a structural definition that remains widely used: dominance is "an attribute of the pattern of repeated, agonistic interactions between two individuals, characterized by a consistent outcome in favour of the same dyad member and a default yielding response of its opponent rather than escalation" [Drews 1993]. Under this definition, dominance status (dominant or subordinate) belongs to a pair of animals. Dominance rank (high or low) describes an individual's position in the group's hierarchy, so it depends on who else is in the group [Drews 1993].
This distinction has practical consequences. A hierarchy is inferred from many pairwise relationships, and those relationships need not line up neatly. In a perfectly linear hierarchy, every pair has a clear dominance relationship and all relationships are transitive: if A dominates B and B dominates C, then A dominates C. Real data sets usually contain pairs that were never seen interacting and pairs whose contests go both ways. Han de Vries developed an improved test of linearity based on Landau's classic index h. The test handles unknown and tied relationships by randomisation, giving the modified index h′ and a significance test for whether a group is more linear than chance would predict [de Vries 1995].
Linearity describes structure, but many questions require a score for each individual. Two approaches now dominate the field. David's score ranks individuals from a matrix of wins and losses, weighting each win by the strength of the opponent. Elo-rating, borrowed from chess, updates each individual's score after every interaction in sequence. The winner gains more points for an unexpected victory than for an expected one [Neumann et al. 2011]. Julia Neumann and colleagues validated Elo-rating for animal data. They showed that, because it is updated interaction by interaction, it can track rank changes over time instead of requiring observations to be pooled into fixed periods, and that it gives hierarchies consistent with established matrix-based methods [Neumann et al. 2011]. A later simulation study asked how much data these methods need. It found that the ratio of observed interactions to individuals required for a reliable hierarchy is surprisingly low, though it depends on how steep the hierarchy is. David's score and a randomised version of Elo-rating performed best unless hierarchies were extremely steep [Sánchez-Tójar et al. 2018].
How Ranks Form: Attributes, Experience and Social Dynamics
The most intuitive explanation for a hierarchy is that it reflects differences in individual attributes such as size, weaponry, age or fighting ability: the strongest animal wins and ends up on top. Ivan Chase examined this idea formally. He argued that if hierarchies were fully determined by individual attributes, attributes would have to predict contest outcomes much more reliably than empirical data suggest in order to produce the high frequency of linear hierarchies seen in real groups [Chase 1974]. Chase and colleagues later tested the idea experimentally with cichlid fish. Differences in attributes did play a significant part, but social interaction was necessary to produce the high proportion of groups with linear hierarchies. In other words, hierarchies are partly self-organising and are not simply a read-out of each animal's qualities [Chase et al. 2002].
One important source of this self-organisation is the effect of experience on later contests. In a wide range of animals, winning a contest increases the chance of winning the next one (the winner effect) and losing reduces it (the loser effect) [Hsu et al. 2006]. In a major review, Yuying Hsu, Ryan Earley and Larry Wolf concluded that losing reduces an animal's willingness to engage in contests, while winning increases its willingness to escalate. Consistent with the sequential assessment model of contest behaviour, prior experience matters less once a contest escalates to physical fighting [Hsu et al. 2006]. They also noted that the neuroendocrine literature on these effects is large but fragmented. No single mechanism has yet been shown to explain how experience is integrated into behaviour, and how multiple past experiences combine is virtually unknown for any species [Hsu et al. 2006]. Winner and loser effects therefore offer a plausible route by which small initial differences, or the chance outcome of an early contest, can be amplified into a stable rank order [Chase et al. 2002; Hsu et al. 2006].
Inherited Rank: Hyenas, Macaques and Baboons
In some species, rank depends less on individual fighting ability than on family. In many cercopithecine primates, including macaques, baboons and vervets, females acquire ranks adjacent to their mothers'. Members of a matriline hold neighbouring positions, producing a hierarchy of family lines [Holekamp & Smale 1991]. The pattern was first documented in Japanese macaques (Macaca fuscata). In 1958 Syunzo Kawamura described the "matriarchal social order" of the Minoo-B group, where offspring ranked according to their mothers' positions [Kawamura 1958]. The regularities of this system became known as "Kawamura's principles". Daughters rank just below their mothers, and within a matriline younger sisters typically come to outrank older ones, a pattern known as youngest ascendancy [Kawamura 1958; Holekamp & Smale 1991].
Kay Holekamp and Laura Smale showed that spotted hyenas (Crocuta crocuta) show a strikingly similar process, which they called maternal rank "inheritance". Young hyenas born into a clan come to occupy positions immediately below their mothers after a long period of social learning, with support from the mother and other kin during interactions [Holekamp & Smale 1991]. The quotation marks around "inheritance" matter: rank is not genetically inherited but socially acquired, through the support that high-ranking mothers can give.
Such societies are sometimes described as convention-based, because position in the hierarchy is gained largely independently of physical traits such as size or weaponry [Strauss & Holekamp 2019]. This creates an evolutionary puzzle. If rank strongly affects fitness, why do low-born individuals accept their place? Eli Strauss and Kay Holekamp analysed 27 years of data spanning five generations of wild spotted hyenas. Females who repeatedly formed coalitions with their top allies were more likely to improve their position in the hierarchy, suggesting that social alliances allow status to change within an otherwise stable, nepotistic system [Strauss & Holekamp 2019]. Using lifetime reproductive success as the fitness measure, they showed that these rank changes had major fitness consequences, and that small differences in rank could become amplified over multiple generations [Strauss & Holekamp 2019].
Fitness Consequences, Despotism and Insect Hierarchies
A central assumption of dominance research is that high rank brings benefits. Lee Ellis compiled findings from several hundred studies of nonhuman animals into a cross-species comparison. Outside the primates, the evidence largely supported the hypothesis that high-ranking males achieve greater reproductive success than subordinates [Ellis 1995]. For primates the picture was more mixed, which led to a phylogenetically controlled meta-analysis by Bonaventura Majolo and colleagues. Drawing on 94 published studies of 25 primate species, they found that dominant males had higher fecundity and mating success. Dominant females had higher infant survival to the first year, but there was no significant effect of dominance on female feeding success. The effect on female fecundity was not significant once phylogeny was controlled for, and the benefits of dominance for female fecundity were stronger in longer-lived species [Majolo et al. 2012]. Rank matters, but how much it matters varies with sex, species and the measure of fitness used.
How unequally benefits are shared also varies between societies. Sandra Vehrencamp's model of despotic versus egalitarian societies treated the dominant's share of group resources or reproduction as something it can bias in its own favour. That bias is limited because subordinates must gain enough from staying in the group, compared with leaving, for group living to persist [Vehrencamp 1983]. The model predicts steep, despotic inequality where subordinates have few alternatives, and more egalitarian arrangements where they can leave easily or where the group depends on their cooperation. This framework became a foundation for later work on reproductive sharing in cooperative breeders [Vehrencamp 1983].
Dominance hierarchies are not confined to vertebrates. In the paper wasp Polistes fuscatus, queens and workers form a linear dominance hierarchy that determines how food, work and reproduction are divided within the colony [Tibbetts 2002]. Elizabeth Tibbetts discovered that these wasps have highly variable yellow facial and abdominal markings. When she altered a wasp's markings with paint, it received more aggression from nestmates than control wasps painted without altering their pattern. That aggression declined as the new markings became familiar [Tibbetts 2002]. This was experimental evidence that individual recognition, long thought to require a vertebrate brain, supports a stable dominance hierarchy in an insect [Tibbetts 2002].
Rank, Stress and Health
The link between social status and health has made dominance research relevant well beyond ethology. In an influential review, Robert Sapolsky asked whether dominant or subordinate primates are more stressed. His answer was that it depends on the social organisation of the species and population [Sapolsky 2005]. Where dominants maintain their position through frequent harassment of subordinates and subordinates have little social support or means of escape, low rank tends to be associated with markers of chronic stress. Where hierarchies are unstable and dominant individuals must repeatedly defend their position, high rank can carry the greater burden [Sapolsky 2005]. Sapolsky reviewed how the stressful aspects of rank have adverse adrenocortical, cardiovascular, reproductive, immunological and neurobiological consequences, and considered how these findings apply to socioeconomic status and health in humans [Sapolsky 2005].
Observational studies struggle to separate cause from effect: does low rank cause poor health, or does poor health lead to low rank? Noah Snyder-Mackler and colleagues addressed this experimentally with 45 adult female rhesus macaques (Macaca mulatta) at the Yerkes National Primate Research Center. Because females introduced earlier into newly formed groups tend to achieve higher rank, the researchers could manipulate social status by controlling the order of introduction [Snyder-Mackler et al. 2016]. Status affected the immune system at several levels, from the numbers of immune cells to gene expression. Natural killer cells showed the greatest sensitivity to rank [Snyder-Mackler et al. 2016]. Low-status females showed stronger activation of an NF-κB-mediated proinflammatory pathway. When blood cells were challenged with bacterial lipopolysaccharide, low-status females mounted a stronger overall response, favouring antibacterial responses, while high status favoured antiviral responses [Snyder-Mackler et al. 2016]. The authors described the immune effects of status as plastic, which shows that rank can causally shape physiology rather than only reflect it [Snyder-Mackler et al. 2016].
Wolves and the "Alpha" Myth
Few ideas about animal hierarchies are as widely repeated as the "alpha wolf": a pack of rivals constantly contesting dominance and held in check by a top-ranking alpha male and alpha female. L. David Mech traced this view to research carried out mainly on captive wolves (Canis lupus). Those captive packs were usually assembled from unrelated animals from various sources, placed together and allowed to breed [Mech 1999]. Mech noted that the prevailing alpha view had been reinforced by many authors, including his own earlier work [Mech 1999]. Among such assemblages of unacquainted wolves, dominance hierarchies with alpha, beta and omega animals did form, and Mech accepted that those labels were probably appropriate in that setting [Mech 1999].
In the wild, Mech argued, the typical wolf pack is not an assemblage of rivals but a family: a breeding pair and their offspring of the previous one to three years, or sometimes two or three such families [Mech 1999]. Offspring usually disperse, most of them when 1–2 years old, so the only long-term members of most packs are the breeding pair [Mech 1999]. His conclusions were based on a literature review and 13 summers of observing a free-living pack on Ellesmere Island in the Canadian High Arctic. The so-called alpha male and alpha female of a natural pack, he concluded, are simply the parents. Dominance contests with other pack members are rare, if they occur at all, and during those 13 summers he saw none [Mech 1999]. The parents lead through a division of labour: the female predominates in pup care and defence, and the male predominates in foraging, food provisioning and the travel associated with them [Mech 1999].
Mech compared drawing conclusions about natural packs from captive assemblages to drawing conclusions about human family life from people in refugee camps. He suggested that calling a breeding wolf an "alpha" adds no more information than calling any parent an alpha, and proposed terms such as "breeding female" or simply "mother" [Mech 1999]. He reserved possible use of "alpha" for the relatively few large packs containing multiple litters, where older breeders may dominate younger ones. His central objection was less about the word than about "what the terminology falsely implies: a rigid, force-based dominance hierarchy" [Mech 1999]. The correction does not mean wolves lack dominance behaviour. Parents do dominate their offspring. But it shows how the social conditions of study animals can create the very patterns researchers go on to describe.
Conservation and Welfare Relevance
Knowledge of dominance hierarchies has direct applications wherever people manage animal groups. The wolf case shows that grouping unfamiliar, unrelated animals, as often happens in captivity, can produce social dynamics quite unlike those of natural groups [Mech 1999]. Experimental work shows that hierarchy formation depends on the history of interactions among group members as well as on their individual attributes [Chase et al. 2002; Hsu et al. 2006]. Adding, removing or regrouping animals can therefore reopen questions of rank that had been settled.
Because rank can affect stress physiology and immune function, the position an animal ends up in within a managed group is a welfare issue, not just a behavioural detail [Sapolsky 2005; Snyder-Mackler et al. 2016]. Sapolsky's comparative analysis suggests that the costs of subordination depend on features that managers can influence, such as the stability of the hierarchy and whether subordinates have social support or room to avoid dominants [Sapolsky 2005]. In species with nepotistic, kin-based hierarchies, such as many cercopithecine primates and spotted hyenas, rank is maintained through family support and alliances [Holekamp & Smale 1991; Strauss & Holekamp 2019]. Separating kin can therefore change the rank structure of an entire group, not just the status of the individuals moved.
The same principles apply to wild populations. In African elephants (Loxodonta africana), Karen McComb and colleagues used playback experiments in Amboseli National Park, Kenya, to show that family groups led by older matriarchs were better at distinguishing the contact calls of familiar individuals from those of less familiar ones. The social knowledge of the oldest female influenced the group as a whole, and families with older matriarchs had higher per-capita reproductive success [McComb et al. 2001]. The authors warned that removing older, experienced individuals, often targeted by hunters because of their size, could have serious consequences for populations of socially complex mammals [McComb et al. 2001]. A later study tested elephants that, decades earlier, had been separated from family members and translocated during culling operations. They performed poorly on tests of social knowledge. They failed to distinguish callers by familiarity, and unlike an undisturbed population they did not adjust their responses to calls simulating individuals of increasing age and social dominance [Shannon et al. 2013]. For translocation, culling and harvest policy, these findings suggest that the social structure of a population, including who leads and how dominance relations are understood, is a conservation asset in its own right and not only a matter of population numbers [McComb et al. 2001; Shannon et al. 2013].
References
[Chase 1974] Chase, I.D. (1974). Models of hierarchy formation in animal societies. Behavioral Science, 19(6), 374–382. https://doi.org/10.1002/bs.3830190604
[Chase et al. 2002] Chase, I.D., Tovey, C., Spangler-Martin, D. & Manfredonia, M. (2002). Individual differences versus social dynamics in the formation of animal dominance hierarchies. Proceedings of the National Academy of Sciences of the USA, 99(8), 5744–5749. https://doi.org/10.1073/pnas.082104199
[de Vries 1995] de Vries, H. (1995). An improved test of linearity in dominance hierarchies containing unknown or tied relationships. Animal Behaviour, 50(5), 1375–1389. https://doi.org/10.1016/0003-3472(95)80053-0
[Drews 1993] Drews, C. (1993). The concept and definition of dominance in animal behaviour. Behaviour, 125(3–4), 283–313. https://doi.org/10.1163/156853993X00290
[Ellis 1995] Ellis, L. (1995). Dominance and reproductive success among nonhuman animals: a cross-species comparison. Ethology and Sociobiology, 16(4), 257–333. https://doi.org/10.1016/0162-3095(95)00050-U
[Holekamp & Smale 1991] Holekamp, K.E. & Smale, L. (1991). Dominance acquisition during mammalian social development: the "inheritance" of maternal rank. American Zoologist, 31(2), 306–317. https://doi.org/10.1093/icb/31.2.306
[Hsu et al. 2006] Hsu, Y., Earley, R.L. & Wolf, L.L. (2006). Modulation of aggressive behaviour by fighting experience: mechanisms and contest outcomes. Biological Reviews, 81(1), 33–74. https://doi.org/10.1017/S146479310500686X
[Kawamura 1958] Kawamura, S. (1958). The matriarchal social order in the Minoo-B group: a study on the rank system of Japanese macaque. Primates, 1(2), 149–156. https://doi.org/10.1007/BF01813701
[Majolo et al. 2012] Majolo, B., Lehmann, J., de Bortoli Vizioli, A. & Schino, G. (2012). Fitness-related benefits of dominance in primates. American Journal of Physical Anthropology, 147(4), 652–660. https://doi.org/10.1002/ajpa.22031
[McComb et al. 2001] McComb, K., Moss, C., Durant, S.M., Baker, L. & Sayialel, S. (2001). Matriarchs as repositories of social knowledge in African elephants. Science, 292(5516), 491–494. https://doi.org/10.1126/science.1057895
[Mech 1999] Mech, L.D. (1999). Alpha status, dominance, and division of labor in wolf packs. Canadian Journal of Zoology, 77(8), 1196–1203. https://doi.org/10.1139/z99-099
[Neumann et al. 2011] Neumann, C., Duboscq, J., Dubuc, C., Ginting, A., Irwan, A.M., Agil, M., Widdig, A. & Engelhardt, A. (2011). Assessing dominance hierarchies: validation and advantages of progressive evaluation with Elo-rating. Animal Behaviour, 82(4), 911–921. https://doi.org/10.1016/j.anbehav.2011.07.016
[Sánchez-Tójar et al. 2018] Sánchez-Tójar, A., Schroeder, J. & Farine, D.R. (2018). A practical guide for inferring reliable dominance hierarchies and estimating their uncertainty. Journal of Animal Ecology, 87(3), 594–608. https://doi.org/10.1111/1365-2656.12776
[Sapolsky 2005] Sapolsky, R.M. (2005). The influence of social hierarchy on primate health. Science, 308(5722), 648–652. https://doi.org/10.1126/science.1106477
[Schjelderup-Ebbe 1922] Schjelderup-Ebbe, T. (1922). Beiträge zur Sozialpsychologie des Haushuhns. Zeitschrift für Psychologie, 88, 225–252. (No DOI; APA PsycInfo record 1926-07009-001.) https://psycnet.apa.org/record/1926-07009-001
[Shannon et al. 2013] Shannon, G., Slotow, R., Durant, S.M., Sayialel, K.N., Poole, J., Moss, C. & McComb, K. (2013). Effects of social disruption in elephants persist decades after culling. Frontiers in Zoology, 10, 62. https://doi.org/10.1186/1742-9994-10-62
[Snyder-Mackler et al. 2016] Snyder-Mackler, N., Sanz, J., Kohn, J.N., Brinkworth, J.F., Morrow, S., Shaver, A.O., Grenier, J.-C., Pique-Regi, R., Johnson, Z.P., Wilson, M.E., Barreiro, L.B. & Tung, J. (2016). Social status alters immune regulation and response to infection in macaques. Science, 354(6315), 1041–1045. https://doi.org/10.1126/science.aah3580
[Strauss & Holekamp 2019] Strauss, E.D. & Holekamp, K.E. (2019). Social alliances improve rank and fitness in convention-based societies. Proceedings of the National Academy of Sciences of the USA, 116(18), 8919–8924. https://doi.org/10.1073/pnas.1810384116
[Tibbetts 2002] Tibbetts, E.A. (2002). Visual signals of individual identity in the wasp Polistes fuscatus. Proceedings of the Royal Society B: Biological Sciences, 269(1499), 1423–1428. https://doi.org/10.1098/rspb.2002.2031
[Vehrencamp 1983] Vehrencamp, S.L. (1983). A model for the evolution of despotic versus egalitarian societies. Animal Behaviour, 31(3), 667–682. https://doi.org/10.1016/S0003-3472(83)80222-X