In less than 200 years, perceptions about the microbial world have sustained a transcendental shift: turning away from unsocial and disease-causing, to gregarious and beneficial across a series of niches and hosts, and as a link between all areas of life (Berg et al., 2020). This insight comes after the recognition that the majority of microorganisms are non-pathogenic (Casadevall and Pirofski, 2000) and that they may have a role in co-evolutionary relationships with their hosts, often resulting in mutually beneficial interactions in terms of fitness (Guerrero et al., 2013). In this sense several studies have contributed to our current understanding of how microbiota (including its genome—the microbiome) can influence the host’s health by disentangling its relationships host-related physiological, immunological, neurological, and developmental processes (Cryan and Dinan, 2012; Gerber, 2014; Stilling et al., 2014; Dinan and Cryan, 2017a,b; Rowland et al., 2018; Zheng et al., 2020). Microbiota is now recognized as a highly dynamic organ in space and time (Gerber, 2014; Donaldson et al., 2016; Uhr et al., 2019; Ji et al., 2020) modified by the dynamics of its microbial communities and numerous environmental forces, rather than solely influenced by host’s genetics and/or phylogeny (Archie and Theis, 2011; Gonzalez et al., 2011; Yatsunenko et al., 2012; Rothschild et al., 2018; Amato et al., 2019; Scepanovic et al., 2019).

Traditional approaches for the study of the relationship between the microbiome and social behavior have focused on the patterns and speed of pathogen transmission, however, contemporary approaches are based on a careful consideration of the extension, intensity, and ubiquity of microbial exchange between organisms and have supported the proposal that endosymbionts exchange could in fact be an “underappreciated” benefit of social interactions (Lombardo, 2008). Due to its role in inter-host transmission of microbes (Browne et al., 2017; Robinson et al., 2019; Sarkar et al., 2020), cumulative empirical studies suggest the influence of social behavior patterns on the characteristics of the microbiome and vice versa (Lombardo, 2008; Archie and Theis, 2011; Ezenwa et al., 2012; Montiel-Castro et al., 2013; Stilling et al., 2014; Archie and Tung, 2015; Johnson and Foster, 2018; Münger et al., 2018; Sylvia and Demas, 2018; Sherwin et al., 2019; Sarkar et al., 2020). These observations have led to the suggestion of the “social microbiome” concept (Sarkar et al., 2020). Defined as “the collective microbial metacommunity of an animal social group or social network” (Sarkar et al., 2020), this concept reflects the idea that the benefits provided by social life are highly intertwined with the paths of microbial transmission, which potentially arise as interactions amongst members of a given group. Thus, the composition of such microbial metacommunity may be influenced by social complexity, “in which individuals frequently interact in many different contexts with many different individuals, and often repeatedly interact with many of the same individuals in networks over time” (Freeberg et al., 2012).

Considering these definitions, our review is focused on unraveling the promising association between social complexity and the gut microbiota composition, in terms of diversity and similarity of microbial communities exhibited and shared among members of a social group. We start by proposing a general framework for exploring the possibility that gut microbiota interchange through social behavior could have important adaptive benefits to individuals within social groups. Then, we expand on the possibility that throughout their decision-making processes, subjects may also control over the selection of microbial communities (i.e., and not only genes or individuals) by expressing certain social behaviors and choice of particular social partners. Next, we use the framework of social complexity to propose several predictions describing how gut microbial diversity and/or similarity may vary in accordance with defining characteristics of social complexity in primate societies. Being highly gregarious organisms living in societies where an individual’s successful survival and reproduction depend on a network of complex social interactions (Maestripieri, 2010), primates are a suitable group to infer connections between social behavior and gut microbiota and, thus, we review primatological studies that offer evidence of the link between both variables.

Primate societies vary largely in size and composition, presumably due to socioecological conditions (Mitani et al., 2012; Strier, 2017), and are characterized by the establishment of strong and long-lasting social relationships (Seyfarth and Cheney, 2012a,b). As the number of these relationships increases with group size, social recognition and group management becomes more cognitively demanding processes, which has been evolutionary inferred from data on neocortex ratios (Dunbar, 1998). Hence, our next section tests the possibility of a quantitative association between social group size, neocortex ratio and gut microbial diversity across primate species. Notwithstanding the number of studies describing mechanisms by which microbiota interchange may influence individual fitness (Ezenwa et al., 2012), we suggest that the potential mechanisms that endosymbionts could employ to transfer between different hosts has not yet been thoroughly described. Thus, the final section of this review underlines the necessity that, to be a “collective metacommunity” (Sarkar et al., 2020), the social microbiome must either be horizontally transmitted across members of a group or environmentally acquired. Therefore, we suggest some of the potential biochemical and molecular mechanisms that gut bacteria could employ to tolerate the aerobic conditions of external environments and thus colonize other individuals’ internal environments.

According to socioecological theory and the social brain hypothesis, group life is an adaptation to solve social problems while group size and social structure—the degree to which the sexes are related, the patterning of affiliative and agonistic interactions between individuals, etc. (Hinde, 1976)—seem to be constrained by individual capacity to process (i.e., recognize, remember, and manage) both social and ecological information to sustain and monitor social relationships (Cunningham and Janson, 2007; Dunbar and Shultz, 2017; Tremblay et al., 2017). Therefore, it is assumed that primates’ behavior is influenced by the characteristics of the socioecological environment in which they live (Janson, 2000; Thierry, 2008; Dunbar and Shultz, 2017) and thus, individuals must frequently make social (e.g., with whom to rest, forage, play, etcetera) and non-social (e.g., when and where to rest or forage) decisions to cooperate or compete with conspecifics, aiming to balance the costs and benefits associated with group living (Conradt and Roper, 2005; Barrett et al., 2007; Majolo and Huang, 2017).

Group life often involves conflicts of interest because individuals should make shared or unshared decisions for accessing to limited resources (mainly food and mates) (Koenig, 2002; Conradt and Roper, 2009) while coping with predation pressure and infection risk by steady exposure to pathogenic agents (Janson and Goldsmith, 1995; Altizer et al., 2003; Nunn and Altizer, 2006; Ezenwa et al., 2016a). Indeed, both direct (e.g., grooming) and indirect (e.g., overlapping range use) social behavior may increase the vulnerability to infectious and non-infectious diseases but also enhance individual resistance to pathogen infection (Altizer et al., 2003; Nunn and Altizer, 2006; Ezenwa et al., 2016a) and impact positively on individual reproductive success, longevity, and survival (Kappeler et al., 2015; Ostner and Schülke, 2018). Another interesting possibility regarding microbial exposure is that within-group cooperation and in-group microbial exchange actually limit the extent of an infection. Pathogens in a certain geographic area could be a selective pressure leading to assortative sociality, out-group avoidance, and limited dispersal (Fincher and Thornhill, 2008). These three factors could easily lead to enhanced within-group sociality, increasing cooperation with known group members and establishing strong behavioral frontiers limiting between-group microbial exchange. In other words, such purported frontiers could lead to preferential within-group social interactions and therefore, shape social structure. This pattern may be evident across different human cultures, where greater inter-individual microbial homogeneity has been reported, for example, in subjects with more collectively driven social values across cultures (Fincher et al., 2008), in tightly knit rural social structures in India (Das et al., 2018), and in people living traditionally in Nigeria (Ayeni et al., 2018). Moreover, it would appear as if the changes in lifestyle of indigenous communities upon exposure to industrialization and westernization imply losing crucial socio-ecological relationships with their environment (Schnorr et al., 2014). Also, the benefits of a richer, more diverse, and high metabolizing microbiota (Obregon-Tito et al., 2015) could be lost and replaced by one associated with diets consisting of high-sugar and-protein (Sánchez-Quinto et al., 2020). This could even include the unfortunate loss of beneficial bacteria carrying previously unknown functional antibiotic resistance genes harbored in the gut microbiota of isolated human populations such as the Yanomami in the Amazon (Clemente et al., 2015).

Kin selection–or inclusive fitness theory (Hamilton, 1964)—and multilevel—or group–selection are two complementary theories proposed in evolutionary biology for explaining the evolution of social behaviors (Marshall, 2011; Kramer and Meunier, 2016) regarding the ubiquitous nature of cooperation occurring both among simple microorganisms and within highly complex societies (Mehdiabadi et al., 2006; Kramer and Meunier, 2016). In this regard, kin selection theory claims that interacting individuals tend to help each other in accordance with the degree of relatedness between them, whereas multilevel selection theory proposes that selection acts both directly on individuals and at multiple levels of biological organization, including cells and/or groups of subjects (Hamilton, 1964; Kramer and Meunier, 2016). Both theories require positive assortment of (genetically) similar individuals for cooperative behaviors to evolve (Kramer and Meunier, 2016). Nevertheless, kinship plays a limited role in structuring social relationships (Chapais and Berman, 2004; Langergraber et al., 2009). Cooperative behaviors among unrelated individuals are usually based on reciprocity although the phenomenon is rare, so interactions between non-kin are likely to be maintained by mutualism or manipulative tactics involving coercion or inducement (Clutton-Brock, 2009). The evolution of cooperative behaviors should consider the strength of selection, the heritability of the group and individual level traits, and the genetic correlation between them (Goodnight, 2005). However, given different examples of adaptability provided by studies of symbiotic relationships with microorganisms (Li et al., 2008; Ezenwa et al., 2012; Cani et al., 2019; Gurevich et al., 2020), it is highly unlikely that natural selection could be applying any selective mechanisms only to genotypes or phenotypes of individuals.

Derived from theoretical approaches proposed to link host’s microbiota and sociality, an alternative and complementary reasoning to understand the evolution of social behavior considers that individuals throughout their decision-making processes, may control the expression of certain social behaviors for selecting microbes within other members of their social groups, and not only the individuals or the genes within them. Nonetheless, such selective control could result costly for individuals, who must consider the ubiquity and diversity of both “good” and “bad” microorganisms (Achtman and Wagner, 2008; Godon et al., 2016) and, consequently, taking into account that microbial exchange between members of a social network and its physical environment may be a continuous process (Sarkar et al., 2020). In this context, behavioral effects of the microbiota could readily arise as a by-product of natural selection on microorganisms within the host and natural selection on hosts to depend upon their symbionts (Johnson and Foster, 2018). These arguments open the questions of how individuals operate the purported microbial selection and whether microbes—or microbial communities—participate in their own selection by modifying the host’s behavior.

The complexity of primate societies can be described based on the temporal variation of four distinct and interrelated components: (i) social organization, referring to group size, age-sex composition and the degree of spatial cohesion; (ii) social structure, describing the content, quality, and patterning of social relationships emerging from repeated interactions among conspecifics; (iii) mating system, involving the average number of males and females with whom each individual sustains mating interactions; and (iv) care system, providing information on who cares for dependent young as well as cooperative breeding in species where it plays a role (Kappeler and van Schaik, 2002; Kappeler, 2019). Conventionally, variation in group size is the most common measure of social complexity because it influences different aspects of the other components; however, all four components should be considered a coherent whole to understand the constraints and flexibility of primate social behavior (Kappeler et al., 2013) and define social complexity in a systematic manner (Kappeler, 2019).

According to Hinde’s framework, social systems—as a result of their complexity—have emergent properties that influence individual behavior and strategies (Hinde, 1976), which may imply either direct interaction among conspecifics (i.e., physical contact) or even mere spatial association (i.e., individuals in proximity or using the same places in different times). These social pathways may spread pathogenic agents causing infectious diseases among hosts (Altizer et al., 2003; Nunn and Altizer, 2006; Ezenwa et al., 2016b) as well as promote the exchange of mutualistic and commensal endosymbionts that may defend a given host from pathogens (Lombardo, 2008; Costello et al., 2012; Ezenwa et al., 2012). For instance, large aggregations of semi-social and eusocial insects are likely to attract pathogenic microorganisms, however, increments in group size are accompanied by increased antimicrobial strength (Turnbull et al., 2011).

Against this background, we expect that any social component could be linked to the structure of gut microbiota. In this sense, gut microbiota composition may vary accordingly to, for instance, spatiotemporal proximity patterns among conspecifics, with the number and duration of physical contacts because of the nature and quality of their multiple social interactions, or with the extent of overlap in their spatial range use. Following this reasoning, we formulate a series of predictions about how gut microbiota diversity and/or similarity may vary in relation to variation in social complexity and hopefully help understanding the potential key role of sociality on the horizontal microbial transmission between conspecific hosts.

One of the main conclusions derived from the “Social Brain Hypothesis” (Dunbar, 1998) is that across species’ evolution, cognitive capacities indexed by brain size or neocortex size have increased under the pressure of supplying the cognitive resources required to follow and participate in a greater diversity of possible behavioral interactions taking place in large group sizes (Dunbar and Shultz, 2021). However, comparative evidence suggests that greater cognitive capacities may also produce reductions in the number of social partners, associated to intense and even pair-bonded social relationships (Dunbar and Shultz, 2007). Furthermore, stronger social partner selection based on greater cognitive ‘power’ and within-group microbial exchange may serve as part of the behavioral immune system, limiting the range of interactions found within social groups (Fincher and Thornhill, 2008). In this context, if greater neocortex size is positively correlated with social group size (i.e., a simple measure of social complexity) across different species and the social microbiome increases its diversity in proportion to increments in social group size, then we can also predict that increased cognitive capacity to select social partners will reduce both group size and the diversity of the social microbiome.

Thus, to test for variations in microbial diversity, based on group size and neocortex size, we developed a metanalysis based on reported values from different studies, including the following variables: species’ specific Average group size, Neocortex size, and Body weight (Supplementary Table 1 and Supplementary Information) as predictors of Maximum number of observed microbial features (measured as the maximum number of operational taxonomic units, OTUs, or amplicon sequence variants, ASVs). Given that body weight may in turn correlate with gut length (Hounnou et al., 2002) (i.e., larger guts could provide more extensive areas for a greater diversity of gut microbial communities to establish) and with larger brain size, we included body weight as another predictor in this analysis. Therefore, through multiple linear regressions, we detected significant linear relationships (Methods described in Supplementary Information). The significant model (R² = 0.42; Adj R² = 0.40; F(3,30) = 20.70, P < 0.001), found that species’ Average group size was the single significant predictor (Table 1) of Maximum observed features. This analysis suggests, as expected, that larger group sizes, i.e., the simplest indicator of a group’s social complexity, predict a greater microbial diversity in the primate gut microbiota (Figure 3).

Metacommunity theory applied to host-microbial systems take into account the feedback between the host-as-patch and its microbial communities (Miller et al., 2018). As part of this feedback, the microbiome affects the condition of the host through, for instance, behavioral changes, and the patch (host) can ‘manipulate’ the composition of the microbial community owing to its dynamic’s nature (Miller et al., 2018). This implies, on the one hand, the understanding of the role that the microbiome could play in the evolution of social behavior. For now, mathematical models suggest that the microbiome is key to explain the evolution of complex social behaviors, such as altruism, cooperation, and parental care (Lewin-Epstein et al., 2017; Gurevich et al., 2020; Lewin-Epstein and Hadany, 2020). On the other hand, several hosts forming a social group can modify their social microbiome, as well as microbial transmission routes (Miller et al., 2018), by making shared or unshared decisions for accessing to resources (e.g., beneficial endosymbionts). In turn, these effects may affect metacommunity coexistence processes, such as competition-colonization trade-off in a patch dynamic system (Miller et al., 2018). Therefore, talking about the ‘social microbiome’ inherently refers to successful microbial transmission between partners or conspecifics by social processes at different organizational levels (Sarkar et al., 2020). Nonetheless, for this to occur it is crucial to be aware of the specific molecular and behavioral mechanisms allowing commensal gut bacteria to be transferred from one host to another, for instance, through a fecal-oral route of transmission (Browne et al., 2017). This transmission route requires that gut microorganisms survive and persist in stressful conditions (external and along a new gastrointestinal tract) while remaining viable to efficiently colonize the gut of other hosts.

Most often, gut microbes are deposited in the external environment as feces, which implies changes in bacterial metabolism and/or cellular architecture to maximize survival, avoiding the toxic effects of external conditions such as atmospheric oxygen, ultraviolet radiation, adverse temperatures and scarce or limited nutrients (Browne et al., 2017). Atmospheric oxygen availability is often mentioned as a detrimental factor for DNA and proteins of anaerobic bacteria, although many gut anaerobe bacteria show varying levels of tolerance (from minutes—Roseburia spp.—to hours—Bifidobacterium spp—or even days—E. coli—) being able to maintain metabolism using oxygen as the terminal electron acceptor (García-Bayona et al., 2020). For instance, Bacteroides species are metabolically and energetically robust and continue growing at 0.10 to 0.14 oxygen concentrations (Baughn and Malamy, 2004; García-Bayona et al., 2020), whereas Faecalibacterium pausnitzii relies on an extracellular flavin-thiol electron shuttle to survive in the presence of oxygen (Khan et al., 2012). Akkermansia muciniphila uses the cytochrome bd complex coupled to an unidentified NADH dehydrogenase to use oxygen as a final electron acceptor, which shifts metabolic process toward a higher acetate-to-propionate ratio, resulting in more ATP and NADH, eventually leading to a slightly increased yield and growth rate (Ouwerkerk et al., 2016).

Another strategy to preserve microbial DNA integrity, usually found in phylogenetically-diverse and human-associated pathogens (one that may be widespread in gut microbiota), is a viable, but non-culturable state which is a form of bacterial dormancy involving decrements in metabolic activity and the generation of a strengthened cell wall achieved by modifications to the peptidoglycan structure (Browne et al., 2017). A similar strategy to cope with environmental stress is the development of resilient, metabolically dormant structures known as spores, with integrity and fecundity conditions maintained by binding of DNA to small acid-soluble proteins in the spore core, in turn surrounded by several endurable protein layers with low permeability and high levels of peptidoglycan (Browne et al., 2017). Many spore-forming gut bacteria, such as members of Firmicutes (Browne et al., 2021), are known to promote homeostasis through the induction of immunomodulatory regulatory T cells, potentially representing up to 30% of the microbial abundance in the gut (Browne et al., 2017). However, spore-forming bacteria are less abundant within individuals but more prevalent in the human population compared to non-spore-forming bacteria (Hildebrand et al., 2021). Spore formation is associated, at least in gut Firmicutes, with features of host-adaptation such as genome reduction and specialized metabolic capabilities (Seedorf et al., 2014; Browne et al., 2021; Hildebrand et al., 2021). This suggests that at least for gut Firmicutes, host adaptation is an evolutionary trade-off between transmission and colonization abundance (Browne et al., 2021).

Because of the gregarious nature of bacteria, they are commonly found forming highly structured multispecies communities, known as microbial biofilms (Stoodley et al., 2002; Nadell et al., 2008) where individual survival strategies may co-occur. From the perspective of the principles established by sociobiology, social behavior is common among microorganisms (Crespi, 2001; Foster, 2010) and beneficial microbes are capable of cooperating with the host’s intrinsic immune capabilities to diminish pathogen colonization (Sassone-Corsi and Raffatellu, 2015). In biofilms, cooperation occurs at a local level, yet strong conflicts can still arise among multiple species and strains. Moreover, spontaneous mutation generates conflicts even within bacterial biofilms initiated by genetically identical cells (Nadell et al., 2008). Therefore, biofilm formation may be conceived as a result from a balance between bacteria cooperation and competition (Nadell et al., 2008).

Biofilm’s bacteria populations rearrange nucleotides, sugar, amino acids, and energy metabolism (Caro-Astorga et al., 2020). These modifications to metabolic arrangement coexist with spore formation, adhesion, motility, synthesis of extracellular polymers, activation of the ROS detoxification, machinery, and production of metabolites (Nadell et al., 2008; Pascoe et al., 2015; Caro-Astorga et al., 2020). Motility, adherence, and biofilm formation are behaviors that are often regulated by quorum sensing, a form of cell-to-cell communication in which small chemical signals, known as autoinducers, are secreted and accumulate as bacteria multiply (Thompson et al., 2016). Moreover, structural adaptations and interrelationships are made possible by the expression of sets of genes that result in segregated subpopulations with different phenotypes and physiological properties (Nadell et al., 2008; Pascoe et al., 2015; Ch’ng et al., 2019). Therefore, the characteristics of bacteria’s biofilms often differ as culture condition changed (Nadell et al., 2008) and biofilm formation represents a protected growth-mode allowing microorganisms to survive in hostile environments and dispersing into new niches (Hall-Stoodley et al., 2004; Pascoe et al., 2015). Such colonization involves a passage through the gastrointestinal tract, the establishment of a niche in the intestinal environment, the use of available nutrients, and replication to levels ensuring stability and survival (Browne et al., 2017).

Essentially, the transmission, and subsequent colonization of oral microbes from phylogenetically diverse microbial taxa into the gut may be a constant process in healthy individuals. Approximately one in three oral microbial cells pass through the digestive tract to settle in the gut, accounting for at least 2% of the classifiable microbial abundance in feces (Schmidt et al., 2019). Although a given host may select commensal bacteria by modulation of the intestinal environment by genetic (e.g., expression of fucosyltransferase 2 gene) and immunological (e.g., PSA signals through TLR2 directly on Foxp3⁺ regulatory T cells) responses (Coyne et al., 2008; Browne et al., 2017), beneficial microorganisms find an assortment of harsh environment conditions during their passage through gastrointestinal tract. These include digestive enzymes, acid pH in the stomach, defensins, and high concentrations of bile salts in the intestine (González-Rodríguez et al., 2013). Nonetheless, Bifidobacteria survive the gastrointestinal transit using stress response mechanisms (such as overproduction of subunit of the F0-F1-ATPase, production of exopolysaccharides and efflux pumps) and specific molecules (extracellular proteins like BopA) involved in adhesion and colonization factors, as well as by taking advantage of specific energy recruitment pathways (González-Rodríguez et al., 2013).

During intestinal colonization the regulation of surface architecture appears to be critical for establishing the commensal association of Bacteroides fragilis with its mammalian host (Liu et al., 2008). This microorganism presents a class of polysaccharide utilization loci (commensal colonization factors, ccf) that is highly induced during gut colonization and bacterial growth in mucin O-glycans (Lee et al., 2013; Donaldson et al., 2018). Furthermore, such ccf locus regulates the expression of specific capsular polysaccharides to attract IgA binding allowing for mucosal colonization (Donaldson et al., 2018). Lastly, one quorum sensing molecule, Autoinducer-2, upregulates the adherence to epithelial cells in Actinobacillus pleuropneumoniae and increases the expression of motility genes for several strains of E. coli. Therefore, Autoinducer-2, which is produced by multiple bacterial species (e.g., Bacteroides spp., Ruminococcus spp., Lactobacillus spp.) likely regulate bacterial behavior and community dynamics in the microbiota (Thompson et al., 2016). In summary, both host and gut microbes are supplied with selective and/or specific mechanisms, simultaneously operating to avoid pathogen spread and encourage successful transmission and colonization of beneficial endosymbionts. Therefore, these mechanisms must also be considered as part of the feedback between the host-as-patch and its microbial communities.

There is increasing evidence of the pivotal role that “good” and “bad” microorganisms play on host health. However, only the reality of the COVID-19 pandemic, starting more than two years ago, has helped us to conceive the actual implications of microbial transmission mediated by social contact within a highly connected world. Understanding this process has become a race for solutions that help to maintain or improve our physiological, immunological, and neurological health, which may be understood as the ‘equilibrium state’ between beneficial and detrimental forces. On the one hand, research on socially microbial transmission is focused on avoiding or slowing the spread of infectious diseases caused by pathogens. On the other hand, studies suggest that social transmission of commensal microorganisms should be fostered to preserve their health-related benefits. In this context, we propose a tight connection between gut microbiota and social complexity within primate groups. We outline predictions concerning how social organization, social structure, as well as mating and care systems may shape the gut microbiota composition at group level. Primate complexity is extensive and while some components receive more attention, others are not appropriately attended yet, like parental care, which seems to be a key aspect for immune and neuronal development in early-life when gut microorganisms play important priming—programming roles. More research is needed to understand the interplay between gut microbial communities and social behavior. An important conclusion is that, since social complexity represents an emergent property of individual social interactions, health may be favored by improving our own social (microbiota-exchange) connections network. Nonetheless, the recent experience suggests extreme care and analyses of the conditions regarding how certain microorganisms may be transmitted by local social behavior (at group level) with effects scaling up to a global order (at population level), with possibly serious public health consequences. Therefore, we consider that a profound and better understanding of the interplay between transmission of microorganisms and social interactions is necessary to distinguish between beneficial and pathogenic capabilities of socio-microbial exchange.

BP-G and AM-C organized the database, performed the statistical analysis, and wrote the first draft of the manuscript. All authors contributed to the conception and design of the study, manuscript revision, read, and approved the submitted version.