Living in stable cohesive groups comes with a number of costs, some of which increase over broad ranges of group size, including risk of infection and competition for access to resources, such as mates, safe spatial positions in the group geometry, food, and supportive partners (Krause & Ruxton, Reference Krause and Ruxton2002). Dunbar provides a unique perspective on the problems faced by primates living in large groups. Here, we critically evaluate some of the premises of this framework in the light of the existing literature.

First, we propose scrutinizing the behavioral mechanisms suggested to alleviate the “stresses” associated with living in large groups, as the target article remains vague about the precise function of grooming and alliances. It is generally accepted that social strategies have evolved to mitigate the costs of group-living for at least some group members (Silk, Reference Silk2007; Snyder-Mackler et al., Reference Snyder-Mackler, Burger, Gaydosh, Belsky, Noppert, Campos, Bartolomucci, Yang, Aiello, O’Rand, Harris, Shively, Alberts and Tung2020; Thompson, Reference Thompson2019). Importantly, different types of adversities arising from increasing group size are countered by distinct affiliative social strategies. For instance, casting a wide social net of tolerant, but not particularly strong affiliative relationships, may help maintain a safe central spatial position, whereas forming a few strong, lasting affiliative ties with equal and reliable partners provides access to monopolizable resources, if these partners serve as allies in outward aggressive conflicts (Ostner & Schülke, Reference Ostner, Schülke, Naguib, Barrett, Healy, Podos, Simmons and Zuk2018). The two strategies are likely independent of each other. The strength of an individual’s strong ties and the number of weak ties it forms (relative to the average strength) are not correlated across group members in 15 groups of five primate species (Schülke et al., Reference Schülke, Anzà, Crockford, De Moor, Deschner, Fichtel, Gogarten, Kappeler, Manin, Müller-Klein, Prox, Sadoughi, Touitou, Wittig and Ostner2022). In our reading, Dunbar argues that groups can grow larger if individuals invest more in the strength and number of strong affiliative relationships, because these relationships serve as alliances, but he does not explain how these function. Under Dunbar’s central premise of purely social costs of group size we propose that alliances will provide two benefits. Strong affiliative relationships underlying alliances offer functional agonistic support when an individual is actually confronted aggressively, and consequently will also modify stressor appraisal. If an individual can count on a few reliable partners as allies, aggression is less likely to elicit a strong stress response, because escalation by the aggressor becomes less likely. Therefore, it is not (as proposed by Dunbar) the grooming received from a partner that buffers physiologically against within-group aggression, but the functional and perceived support provided by the strong relationships built and maintained by grooming and other affiliative behaviors.

Second, we question the proposal of an endorphin-driven infertility trap in larger groups driven solely by increased social stressors for several reasons. (i) The unified stress response originating in the hypothalamus in reaction to a diverse array of adversities produces CRH which is cleaved into ACTH triggering glucocorticoid release from the adrenal cortex and the precursor of beta-endorphin (Pilozzi, Carro, & Huang, Reference Pilozzi, Carro and Huang2021). Given the one-size-fits all nature of the stress response, the parallel endorphin and glucocorticoid release will not only be caused by heightened aggression in larger groups, but equally by increased feeding competition negatively affecting energy balance. Indeed, if primates are released from food scarcity by provisioning, group size often doubles as do reproductive rates which suggests ecological constraints overriding social stressors as the main predictor of fertility at least in these circumstances (Borries, Koenig, & Winkler, Reference Borries, Koenig and Winkler2001; Takahata et al., Reference Takahata, Suzuki, Agetsuma, Okayasu, Sugiura, Takahashi, Yamagiwa, Izawa, Furuichi, Hill, Maruhashi, Saito, Saito and Sprague1998). Similarly, constrained access to food alone may explain very fine-grained diminishing returns with increasing group size particularly for capital breeders or relaxed income breeders that store energy for the upcoming reproductive event before they conceive (Brockman & van Schaik, Reference Brockman, van Schaik, Brockman and van Schaik2005; Heesen et al., Reference Heesen, Rogahn, Ostner and Schülke2013; Jönsson, Reference Jönsson1997; Koenig & Borries, Reference Koenig and Borries2001). Thus, we disagree with the dismissal of ecological influences on group size related decreases in productivity. (ii) The target article remains unclear why social stressors, i.e. aggression rate, increase with larger group sizes (independent of increased feeding competition) and why any increase in aggression – even aggression among third parties - should lead to a stress response. The stress response is an evolved response to restore disrupted homeostasis and not a reaction to something that has no fitness consequences (Beehner & Bergman, Reference Beehner and Bergman2017; Brent et al., Reference Brent, Chang, Gariépy and Platt2014; Ostner & Schülke, Reference Ostner, Schülke, Naguib, Barrett, Healy, Podos, Simmons and Zuk2018). (iii) Related to the above, grooming itself seems an unlikely candidate to lift individuals out of an endorphin-caused “infertility trap” as grooming does not buffer, but instead enhances the endorphin response (Marchin & Dunbar, Reference Marchin and Dunbar2011; Pearce et al., Reference Pearce, Wlodarski, Machin and Dunbar2017).

Third, we caution against dismissing the possibility of phylogenetic inertia as a factor explaining the observed distribution of group sizes and associated adaptations, as several of the traits discussed clearly cluster taxonomically. We provide three examples: (i) the soft touch of grooming cannot be a key mechanism across primates, because large platyrrhines do not use grooming as their main affiliative behavior (Aureli & Schaffner, Reference Aureli and Schaffner2007) and lemurs do not groom softly, but groom at high stroke frequencies using their specialized tooth comb and tongue instead of their hands which also puts an energetic limit to their total grooming time (Rosenberger & Strasser, Reference Rosenberger and Strasser1985); (ii) changing diet to free time for grooming is not a flexible option for colobines. Foregut fermenters do not strive on diets rich in those fruits that Dunbar envisions (low in fiber and rich in carbohydrates), but suffer damage to their gut microbiome from associated changes in pH that may even hamper a return to a leafy diet (Goltenbolth, Reference Goltenbolth, Kloes and Lang1976). If colobines fall back on fruit when their preferred leaves are not available (Milich et al., Reference Milich, Stumpf, Chambers and Chapman2014), these fruits are unripe, fibrous and lack juicy flesh (Davies, Oates, & Dasilva, Reference Davies, Oates and Dasilva1999); (iii) females regularly engage in aggressive coalitions only in cercopithecines but not in lemurs, colobines, and most of the platyrrhines irrespective of group size which suggest phylogenetic constraints on the evolution of alliance formation for within-group conflict (Chapais, Reference Chapais1995; Nakamichi & Koyama, Reference Nakamichi and Koyama1997). In summary, critical pieces of the pathway linking changes in grooming behavior to increased group size are not free to vary between species which calls for an analysis strategy closely tracking the shifts in different behaviors and in group size across the primate phylogeny.