How can non-human primates inform evolutionary perspectives on female-biased kinship in humans?

1. Introduction

Male-biased social structures predominate in human cultures, leading anthropologists to puzzle over explanations for the relatively rare societies organized around maternal kin. By contrast, matrilineal kin relationships dominate the social structures of non-humans, including our closest evolutionary relatives, the primates. The following review contributes to a special issue on female-biased kinship [1] that attempts to address how different forms of kinship organization have arisen in human cultures, as well as how female-biased kinship organizations function within societies. To that end, a broader phylogenetic perspective is valuable for exploring the evolutionary roots of human kinship organization, and for assessing how similar patterns of social behaviour operate in non-human primates.

The following review examines the role of matrilineal kinship in non-human primates and is aimed at providing an informed discussion of how these systems compare with potential analogues of female-biased kinship organization in humans. I do not attempt to systematically or exhaustively address the variation in kinship and dispersal patterns among non-human primates. Instead, this review begins with an examination of broad similarities and differences in human and non-human primate societies that may affect kinship organization. Following this, I probe into well-studied systems that may be of most interest to human evolutionary ecologists as comparative referents. Three case study systems are of particular interest, each for different reasons. First, several primate species live in socially monogamous, cooperatively breeding groupings that are often cited for their apparent resemblance to human families. Most are also territorial, which provides a potential for a commodity to be transmitted through a lineage. A key question is whether these groups, like human families, exhibit a clear sex-biased pattern of inheritance, particularly with respect to territory and/or breeding status. Second, cercopithecine primates warrant attention for the particularly well-defined role that matrilineal kinship plays in the social relationships of females. These species provide information on the functional roles of matrilineal kin relationships in primates and allow us to assess how well kin relationships and the acquisition of social status conform to the anthropocentric concepts of descent and inheritance. Third, I examine the role of kinship in great apes which share the most recent common ancestry with humans. As the extant great apes exhibit substantial variation in group structure and varying extents of female dispersal, they provide a good opportunity to address how social organization constrains the formation of matrilineal kin relationships.

2. Human matriliny and the matrilineal ‘puzzle’

Matrilineal systems can be broadly described as those in which maternal kinship more strongly influences social organization than paternal kinship [2]. Fewer than 20 per cent of contemporary human societies are classified as matrilineal [3]. Thus, a major emphasis of cultural kinship theory has been explaining why matriliny should be so rare. Theoretical arguments frequently reference the ‘matrilineal puzzle' [4], which poses that matriliny and related forms of female-biased kinship organization should not be favoured under most socioecological conditions. In its original sociocultural conception, the puzzle emerges because, even in matrilineal systems, control of resources and authority is most often vested in males. This creates conflicts of interest between a woman's brother, who provides the resources to her family, and her husband, who would effectively control them [5]. In evolutionary terms, this translates into a problem of relatedness [6]. Males in matrilineal societies often (though not always) transmit their wealth to their sisters' offspring, who are less closely related to them than are their own offspring. This is only expected to be evolutionary advantageous under conditions of very high paternity uncertainty [7,8], which is indeed common in matrilineal groups [9]. Not all matrilineal societies work in this way. Some, like the Chewa of Malawi, feature direct inheritance by daughters [10]. This raises a second evolutionary problem, which is that under most conditions, wealth or status is expected to contribute more to reproductive success for males, the high variance sex, than for females [11,12]. Whereas a high-status, wealthy male may be able to acquire multiple wives, women's fertility has stricter constraints imposed by the time and energy costs of pregnancy and lactation.

To understand why matriliny is rare, scholars have recently focused on a second set of questions: what socioecological conditions favour it? Women in matrilineal societies most often reside near their relatives after marriage [13]. Thus, it is possible that female dispersal is a primary constraint on matriliny. Additionally, it has been noted that matriliny is most common among horticulturist groups and least common among those that rely heavily on pastoralism [13]. The acquisition of cattle apparently led to the loss of matriliny in many Bantu groups [14]. It is suggested that pastoralism may favour the adoption of patriliny because the inheritance of livestock (in contrast to land) has a proportionately larger effect on the reproductive success of males versus females [14–16]. These findings emphasize the importance of evaluating the roles of dispersal and inheritance in a comparative examination.

3. Unpacking ‘matriliny'

The concept of matriliny encompasses a suite of interrelated constituents that are commonly, but not inextricably, linked [2]. To avoid confusion, the term female-biased kinship organization is a more appropriate umbrella term [1]. In common use, matrilineal can refer either to systems of descent or of inheritance. Matrilineal descent refers to reckoning of kinship through maternal ancestors. Because humans usually recognize both maternal and paternal relatives, descent more specifically refers to social conventions that govern from which ancestral line one acquires their group identity, or more broadly, to which relatives one is more strongly aligned. Matrilineal inheritance refers to transmission of resources or other forms of capital across generations using the maternal line. As noted above, this may take different forms in human societies but ultimately results in family assets being transmitted for the benefit of their daughters' offspring. Matrilineal descent and/or inheritance often co-occur with matrilocality, a form of post-marital residence wherein a married couple resides in proximity with the woman's kin. Collapsing these constituent behaviours into a singular concept of matriliny has increased the difficulty of identifying cultural, ecological and evolutionary causes [1] and warrants particular attention when exploring their roots in non-humans.

Although kinship is a vital component of social organization in both humans and non-humans, scholars studying primate communities do not use the same terminology. There are some obvious reasons for this, most notably that many of the human kinship terms refer to cultural constructs or to processes linked to nuclear families. However, whether these key structural characteristics of human societies have functional analogues in primates is an empirical question worthy of closer examination. For example, post-marital residence is conceptually similar to primate philopatry, which identifies which sex breeds in its natal group. In practice, this is more often described by its inverse, sex-biased dispersal, which can refer to which sex is more likely to leave its natal group, or if both sexes disperse, which sex moves the farthest. Both human and primate forms of dispersal limit inbreeding and determine the access of individuals to kin, but primates that disperse must search for new breeding opportunities, while human post-marital residence, by definition, is contingent on the establishment (and prior negotiation) of a partnership [17]. Primates form differentiated relationships with kin that are plausibly based on descent. They also have potential commodities to be inherited, such as status or territories. One goal of this review is to further probe the equivalence of these concepts and the processes that govern them. For clarity, I will avoid the use of human kinship terminology when discussing the primate evidence, focusing instead on describing the patterns observed. Any use of ‘matrilineal' as it pertains to primates refers only to kinship (i.e. genetic relatedness via the maternal line).

4. Non-human primates as comparative models

The conditions that make female-biased kinship organization a puzzle in humans are largely absent in non-human primates. First, non-human primates lack the forms of accumulated material wealth that form the basis for lineal inheritance. Primates do have the potential for inheritance of fitness-enhancing currencies, such as territory, status or parental investment. Primates also frequently exhibit nepotistic biases in social relationships, exhibiting increased tolerance towards or cooperation with certain kin [18,19]. Second, promiscuous mating by females is a feature of many primate breeding systems [20,21], creating low paternity certainty. Accordingly, direct paternal care is limited or absent in many primates [22,23]. These conditions may place constraints on the ability to recognize paternal kin. Third, given the importance of aggressive competition for mating success among males of most primate species, male status is transient. Thus, the potential for patrilineal transmission is limited because a father cannot typically sustain his status long enough to influence that of his sons. It is, therefore, a basic prediction that kinship ties, investment, and any form of inheritance would be facilitated by a known relationship to ego's mother and maternal relatives that associate closely with her.

Unlike humans, primates generally do not maintain contact with kin networks after dispersal from natal groups. We should predict a tight correlation between the sex that disperses and the valence of kin relationships. However, there are important sex differences in how this operates. Where male-biased dispersal is the norm, as among the majority of cercopithecines, females form coalitions that are specifically rooted in their matrilineal relationships, and these can be multigenerational [24]. Where female-based dispersal is the norm, as among most platyrrhines and apes, the resulting high relatedness among cohorts of males is predicted to facilitate cooperative relationships [25]. In fact, genetic data often fails to support the premise of higher male than female relatedness in such groups [26]. Moreover, true kinship among male dyads plays a limited role, perhaps because constraints on kin recognition limit the types of known male kin that are available [24]. Strong relationships among some brothers are observed, but competitive strength or socioecological ‘competence’ appears to override kinship in selection of coalition partners [27]. Those kin alliances that do form are most commonly based on maternal relationships, as in chimpanzees (Pan troglodytes) [28] and hamadryas baboons (Papio hamadryas) [29].

Hypotheses about kin relationships among females lie at the root of primate socioecology. Based on the twin premises that females are limited by access to food resources and, therefore, should be distributed on the landscape in accordance with the distribution of those resources, Wrangham [30] observed that whenever it is possible to monopolize food patches, females should form cooperative relationships to more effectively defend patches and exclude others. These relationships should form most readily among close kin who secure inclusive fitness benefits, reducing the costs of sharing the contested resource. Wrangham argues that feeding competition thus selects for female philopatry and the formation of strongly affiliated matrilineal kin groups. While female-biased kinship is often presented as an evolutionary puzzle in contemporary humans, it should be evolutionarily favoured in other primates.

Despite some loose similarities, primate socioecological models adopt a fundamentally different perspective than that on which much of kinship theory in humans is based. While both involve strategies to control resources within groups of related individuals, the primate models focus much more on the active collaboration of individuals to govern current and future resource access versus human models that focus largely on the flows of resources across generations of resources accumulated in the past.

5. Primate ‘family' groups

While the majority of primate species live in some form of polygynous grouping, several species live in ‘socially monogamous' groups, comprising a breeding pair and dependent offspring. Such groups are most well described among the hylobatids (gibbons and siamangs), and several species of platyrrhines (marmosets and tamarins, titi monkeys, owl monkeys and sakis) [31]. While commonly used, the term ‘monogamous' is misleading or imprecise for a number of reasons, including extensive extra-pair paternity, facultative polyandry (some females pair with more than one male), and high partner turnover [32–35].

Acknowledging their complexities, socially monogamous primate species have several features that make them of potential interest. First, groups often centre on a pair-bonded male and female. The neurobiological mechanisms underlying pair-bonding in platyrrhines bear important similarities to those in humans, suggesting broader functional similarities [36,37]. Second, these species often exhibit extensive biparental care for offspring [22,23], a defining feature of the human pair-bond [38]. Third, groups are often characterized by extensive cooperation, not just among the bonded pair, but by older offspring helping to feed and care for younger siblings [39]. Finally, these groups are often territorial, with the availability of territory being the key constraint on breeding. These traits and others have elicited comparisons with human cooperative breeding [40] and set up a context in which one could imagine the vertical transmission of territory or breeding status through the family.

The co-occurrence of social monogamy and paternal care in many species led to the hypothesis that this form of grouping evolved for the purpose of providing biparental care where it is most needed [31,41]. However, there are also many socially monogamous species that do not engage in direct paternal care [42]. Several phylogenetic analyses support an alternative ecological cause: social monogamy results when females are dispersed over the habitat in such a way that males cannot effectively monopolize more than one [42–44]. This can result from food resources that are high in quality but sparsely distributed, such that females are better off maintaining solitary territories than cooperating with other females to defend patches. A key implication of this model is that the socially monogamous group can be seen as an extension of a single female who maintains exclusive breeding privileges on a territory. While making similar predictions, this perspective differs intuitively from the models often cited for birds and humans which use male defence of resources as their starting point [43].

Among the socially monogamous platyrrhines, many are well studied in captivity where groups are artificially constructed. Their small size and monomorphism have made it more difficult to study the dynamics of group formation and dispersal in the wild. Most data comprise anecdotal reports, thus it is problematic to make reliable conclusions about commonalities or differences between species, many of which are poorly studied. The available data suggest that in most groups and most species, breeding is restricted, via either behavioural or hormonal mechanisms, to the oldest female in a group [45,46]. For most individuals maturing within a group, there will be no opportunity to breed there. Some will be forced out of their group by agonism from the breeding pair, though older offspring are sometimes tolerated if they contribute by caring for the parents' dependent offspring [47]. Typically, both sexes must disperse to achieve breeding status [48,49]. They may immigrate into a group where one of the breeding pair has died or lodge a hostile takeover of a breeding position in another group [50–52]. However, in rare cases when inheritance of the breeding position has been observed, it has been by a daughter of the former breeding female [50,53,54], approximating a matrilineal system.

Hylobatids exhibit some similarities to the socially monogamous platyrrhines. Groups comprise a socially bonded pair and associated immatures; some multi-male groups occur; the ‘pair' defends a territory; males participate extensively in parental care (siamangs, but not gibbons); the social bond is also a preferential mating partnership, though extra-pair copulations and paternities do occur [55]. Like the socially monogamous platyrrhines, replacement of breeding individuals largely occurs via immigration from outside groups. This includes not only replacement of deceased individuals and forcible takeovers but also voluntary abandonment of partners [56,57]. The frequency of ‘divorce' results in a relatively high rate of turnover among breeding pairs, a fact which has led researchers to reject the label ‘nuclear families', as many immatures may be unrelated to at least one of the adults with whom they reside. In both siamangs (Symphalangus syndactylus) and white-handed gibbons (Hylobates lar), genetic evidence suggests shorter dispersal distances for males versus females, such that males are more likely to move into adjacent, or even overlapping, territories to their natal territory [58,59]. Replacement from within a group is rare. All known cases involve males, though it is unclear whether or how they were related to either of the original pair [56,60,61]. Whether any kind of resource inheritance occurs in hylobatids is, therefore, ‘up in the air’, but there are some suggestions that the natal ‘parents' may facilitate territorial acquisition by offspring not only by tolerating them but perhaps by expanding and then conceding their original territory [61].

In summary, there are critiques about the conceptualization of monogamy in non-human primates, based primarily on arguments that these species exhibit significant departures from true social and sexual monogamy and vary in social organization between species or populations. The same could be said of humans, thus these critiques do not necessarily invalidate efforts to make comparisons. In the two major taxa of socially monogamous primates, males and females form social and sexual bonds to raise young and/or guard a territory. While these breeding units bear similarities to aspects of human family groups, they do not similarly result in clear systems of descent or inheritance, matrilineal or otherwise. Instead, territorial access and breeding priority are most often transmitted horizontally, as immigrating individuals successively replace members of the breeding partnership. Bequeathal, wherein parents voluntarily abandon a territory to the benefit of their offspring [62], has not been observed in these or other primates. Cases of apparent inheritance do exist, by daughters in platyrrhines and by sons in hylobatids, but it is difficult to rule out the possibility that young adults simply take the closest available breeding opportunity, occasionally in their own groups. A closer look at socioecological conditions in these species suggest that such small groups are favoured precisely because the habitats cannot support multiple breeding individuals in one location, thus are essentially incompatible with multigenerational family structures. Selection may instead favour distribution of kin into new territories, taking advantage of the first possible opportunities to acquire independent breeding status.

6. Matrilineal kinship in cercopithecines

The most obvious examples of matrilineal kinship in primates occur in the cercopithecines, a clade of highly social Old World monkey species, including the baboons (Papio spp.) and macaques (Macaca spp.). This is a diverse group of species that, with rare exception, exhibit female philopatry and male dispersal [63]. Females form multigenerational, matrilineal kin networks characterized by close affiliation, tolerance, and cooperation within kin groups, and competition for resources between kin groups. The strongest and most enduring affiliative bonds emerge among the closest maternal relatives (i.e. sisters and mother–daughter dyads) [64–66]. These relationships take shape early in life, thus an individual's concept of kin most likely originates by reference to those that affiliate most closely with its mother [67,68].

Matrilines are a key scaffold for the development of status relationships within cercopithecine groups [69], a pattern that has also been observed in distantly related white-faced capuchins (Cebus capucinus) [70], ring-tailed lemurs (Lemur catta) [71] and spotted hyenas (Crocuta crocuta) [72,73]. In many cercopithecines, females inherit their mothers' ranks following strict rules, the prototypical pattern being one in which a female's daughters earn ranks immediately below hers, and the youngest daughter ascends to the highest rank among her siblings [74]. Ranks are evident and are recognized long before females are capable of competing individually [75]. By virtue of this system, those who are closely related are also close in rank, but maternal relatedness more strongly drives affiliation between dyads than does rank similarity [64]. Secondary to their relationships with maternal kin, females of some species also associate preferentially with paternal relatives [65,76–81]. By contrast, where it has been examined, adolescent males do not appear to align with their matrilineal relatives, even during serious intragroup conflicts [82,83], and social ranks of immature males are not necessarily dependent on those of their mothers [84,85].

Playback experiments on chacma baboons (Papio hamadryas ursinus) and vervets (Chlorocebus pygerythrus) have shown that individuals not only recognize the vocalizations of their own kin but can recognize the kin relationships of others [86,87]. For example, baboons respond to recorded distress calls by attending to a close maternal relative of the caller [88] and will accept reconciliation signals from a close maternal relative of a recent aggressor in substitute for those from the aggressor herself [89]. Similarly, cercopithecines understand not only their own dominance relationships, but those between other individuals [90,91]. This extends to a hierarchical understanding of how dominance relationships are embedded within kin groups, such that rank differences are conceptualized differently between matrilines than within them [92]. In these primates, kinship has more than a simple egocentric significance but has attributes of ‘membership' that are important criteria of human descent systems.

Both dominance status and the presence of strong social bonds with kin have important fitness effects in cercopithecines [93,94]. Dominant females and those with strongly bonded social partners have longer lifespans, higher offspring survival and faster offspring growth and maturation [95–97]. This can be self-reinforcing, as stronger matrilines grow their coalition disproportionally, though this is not inevitable as dominant matrilines can be overthrown [82] or can split [98].

Thus, cercopithecines exhibit matrilineal inheritance of a fitness-enhancing commodity across multiple generations and with broad recognition of matrilineal identity throughout the group. This system would appear to rely on female philopatry, as well as on strong within-group competition over resources. However, in a number of elegant experiments, Chapais [75,99,100] demonstrated that rank, while inherited, is not possessed by a female, but is dependent on active maintenance through kin-based alliances. Rank reversals can easily be engineered by manipulating the availability of kin, and then only once those kin have actively supported one another. If for some reason a female loses access to her relatives, she quickly loses her rank to others with stronger support systems. Rank is not so much passed down in accordance with the rules of descent as it is shared by related individuals with mutual interests [24].

7. Our closest relatives

In the search for the evolutionary origins of human social structure, the other hominids are essential referents owing to their close phylogenetic relationship. This is an awkward comparison because, despite many biological similarities, the great apes exhibit substantial variation in social structures. For this particular comparison, the critical issue is that none of the living great ape species fit the conventional definition of being female philopatric, thus we may expect matrilineal relationships to be limited. Without making any assumptions about potential homologies between great apes and humans, we can use these examples to draw inferences about the socioecological constraints on kin relationships in closely related organisms.

Orangutans (Pongo spp.) exhibit a dispersed social organization characterized by a limited fission–fusion system [101,102]. Individuals spend the majority of their time alone, and while they do form small, temporary aggregations, the concept of an orangutan ‘group' is elusive. Both sexes disperse to ranges that differ from those of their mothers. However, dispersal is male-biased in that females are often found living in ranges that are adjacent to or overlapping those of their mothers, while males disperse over longer distances [103–106]. Despite limited contact, females do maintain matrilineal kin relationships. They are more likely to associate at all with maternal kin than with other females, and when they do so, they do so for longer durations [107]. While adult females rarely engage in direct affiliative contact with one another, close maternal kin are more likely to feed at close proximity and tolerate, or even encourage, their infants playing together [107].

Most gorilla (Gorilla spp.) breeding groups are one-male polygynous units, though mountain gorillas (Gorilla beringei beringei) are also commonly found in cohesive multi-male, multi-female groups [108–110]. Both sexes usually disperse, though with important differences. The modal pattern is for males to leave the group to join a bachelor group, or to become solitary, eventually attempting to form their own unit by luring females away from other males [111]. This can lead to dispersal distances for some males that greatly exceed those of females [112,113]. Among mountain gorillas, however, only about half of males emigrate [114,115], while others breed in their natal group, or succeed the dominant male through death or takeover [116,117]. There is no evidence for patrilineal succession. It is rare for co-breeding males to be father and son, or otherwise closely related [118]. Furthermore, while both unimale and multi-male groups have high degrees of paternity skew, true paternity does not appear to predict relationships between males and their offspring [119]. In western gorillas (Gorilla graueri), nearly all males emigrate, but in at least some populations, males from nearby groups share higher than expected relatedness and unusual tolerance in intergroup interactions [120] (but see [121]). It is not yet known how matrilineal versus patrilineal kinship influences the formation of these dispersed male networks. Female gorillas nearly always leave their groups at maturity and may transfer multiple times during their lives. As with males, dispersal need not eliminate interactions with kin. A genetic examination found that approximately half of females in a western gorilla population had close female relatives in their groups, a phenomenon that seems to result from co-dispersal of siblings [122]. While the likelihood of co-dispersal for paternal siblings is high (high reproductive skew leads to shared paternity within age cohorts), among mountain gorillas, those females most likely to affiliate and support one another are maternal kin [123].

Chimpanzee groups comprise ‘fission–fusion' social structures, with classic male philopatry, wherein males exclusively remain in their natal groups for their lifetimes. Mirroring the female-philopatric cercopithecines, male chimpanzees form strong, enduring bonds with one another [124], and as these bonds form the basis for cooperation in coalitionary aggression, males accrue reproductive benefits from their alliances [125,126]. While male philopatry is often thought to promote cooperation among related males, male bonds are not formed strictly on the basis of kinship [28,127,128]. Some brothers form strong bonds, others become rivals, and the majority of bonds are formed among distantly related males. When bonds are made among brothers, they are more strongly based on maternal rather than paternal relationships [28]. Although males may live with their mothers for several decades, relationships between adult males and their mothers are not consistently strong [129]. Male chimpanzees in at least one study community show an unexpectedly strong fidelity to their mothers' foraging ranges, returning there while ranging alone, even long after their mothers have died [130].

Female chimpanzees are less gregarious than males, and while frequently in spatial association with other females, they show low rates of active affiliation with one another. Female dispersal at maturity is the norm. Unlike in western gorillas, genetic data confirms that female dispersal really does have the effect of separating chimpanzee females from close kin [131]. However, females occasionally stay to breed in their natal communities. While these cases are unusual, females who remain in their natal groups readily maintain matrilineal relationships not unlike those seen in female philopatric species. Maternally related female dyads, and particularly mother–daughter dyads, exhibit very strong affinities, demonstrated in range overlap, proximity and grooming [131–133]. When both have infants, adult mother and daughter pairs are nearly inseparable, and make for particularly well-developed multigenerational play groups. In one exceptional example, a female at Gombe, herself a natal female, ‘adopted' her daughter's first infant, nursing and carrying her alongside her own, despite the continued presence of the biological mother [134]. Mother–daughter dyads have also formed offensive coalitions to attack other mothers and their infants [135,136].

The famous ‘F' family at Gombe provides a striking illustration that, whether formally reckoned or not, matrilines can have a strong influence on chimpanzee communities [137]. Female Flo, one of the first chimpanzees identified by Jane Goodall in 1962, launched what might validly be considered a dynasty. Already estimated to be in her 40s, Flo was known to have had five offspring. Her son Figan became alpha male, while daughter Fifi stayed in the community and became alpha female. Fifi gave birth to nine offspring, making her the most successful known wild chimpanzee mother (3–4 offspring is typical). Three of Fifi's sons became alpha males, one siring at least eight offspring [137,138]. Fifi's eldest daughter, Fanni, also stayed in her natal community and has given birth to at least eight offspring. Fanni's eldest son has recently become alpha male. Thus, over the past 50+ years, members of the F matriline, through high fertility and monopolization of the power structure, have come to form a large and influential proportion of the Kasekela community. The mechanisms for this success are not clear and could be genetic, but researchers attribute at least some of the F family's success to strong, supportive familial bonds [137].

Bonobos (Pan paniscus), chimpanzees' sister species, share homologous aspects of social organization—male philopatry and fission–fusion dynamics—with surprising differences in social behaviour. Key among these is that, despite being the dispersing sex, females maintain a high degree of affiliation and bonding with other females [139–141]. Genetic evidence confirms that these relationships are not based on kinship [142]. Females' social bonds provide a mechanism for coalitionary support which is often used to thwart harassment by males [143–145]. Unlike in chimpanzees, where all males are dominant to all females, female bonobos can match or outrank males and frequently occupy the alpha position in a community [139,144,146]. In addition to their support for other females, females maintain intense social bonds with their adult sons and render support for them in agonistic interactions [147]. Remarkably, mothers influence their sons' mating success, both by providing active agonistic support and by facilitating close spatial proximity to oestrous females [147,148]. A mother's presence, as well as her rank, may also influence her son's rank in the community ([149,150], but see [147]).

Among the great apes, matrilineal kinship should be predicted to be of little importance because all species exhibit some degree of female dispersal. Most female great apes exhibit weak tendencies to affiliate with one another. However, their kin relationships hold some interesting surprises. In bonobos, where female social relationships are the most well developed, they are not rooted in kinship. By contrast, in those species characterized by weaker female relationships, females go out of their way to maintain associations with the few female maternal kin that they have access to. Even where high paternity skew and male philopatry occur, paternal kinship has little influence on relationships among males. Thus, while great apes lack the ubiquitous matrilineal kin structures found in cercopithecines, there is strong evidence that individuals favour lifelong relationships with close maternal kin, even subverting constraints of dispersal to do so. Strong matriarchs appear to be able to exert considerable influence, direct and indirect, on offspring success and on community structures.

8. Discussion

While primate societies vary considerably, and no primate species provides a direct analogue to humans, the case study systems examined here can allow for some important conclusions. Kin relationships in non-human primates are consistently organized through the maternal line. This bias is remarkably persistent. Individuals form preferential affiliations with maternal kin even when those relationships are not readily available, and under conditions of both high and low paternity certainty. Even relationships among males, though often not based on relatedness at all, are more often reckoned through maternal versus paternal kinship. While there is evidence that at least some primates can recognize close patrilineal relationships, these tend to take a back seat to matrilineal ones. There are a few species in which fathers form strong relationships with sons [24,151], but there are essentially no strong cases of primates that structure broader social relationships on patrilines.

While dispersal may constrain the availability of kin, it does not appear to reduce the drive to cooperate with maternal relatives, which often materializes in unexpected ways. The contrasts in social organization between cercopithecines and the great apes imply that the benefits of bonding among female kin, particularly for feeding competition sensu Wrangham [30], have been subsumed by other social and ecological factors favouring male philopatry in hominids. The persistent insinuation of matrilineal kinship into these systems suggests that benefits for individuals may yet remain and apply in a broader range of contexts. Thus, even if we assume that early hominin ancestors may not have been female-bonded, per se, it is reasonable to conclude that matrilineal biases were conserved, ready to emerge and strengthen whenever other constraints were eased. Similarly, even within strongly patrilineal human societies, maternal kin may continue to exert both socially and biologically significant influences [152,153].

There are important limitations on the expression of female-biased kinship in primates which suggest fundamental differences from those systems in humans. The concepts of ‘inheritance' and ‘descent' are not fully realized among the primates. On the one hand, affiliation with matrilineal kin networks bears important similarities to human matrilineal descent. That is, individuals are most likely to identify their kin with specific reference to their mothers, and matrilineal relationships are salient to how individuals conceptualize or classify other individuals in the group. On the other hand, maternal kin relationships are usually narrowly comprised as strong affinities with the closest relatives—one's mother, offspring and maternal siblings. This may be owing to the limitations on kin recognition mechanisms, diminishing returns on inclusive fitness, or the enhanced potential for competition within extended kin networks.

The common patrilineal inheritance system is thought to be favoured when wealth (e.g. land, livestock) can be accumulated in suitable quantities that a male heir can achieve disproportionate reproductive success [13,14]. Such a condition probably does not exist in non-human primates, whose resources, if worth defending, are often distributed across areas that are difficult to defend [154]. Large territories require shared defence by multiple males (or females). When single males engage in intergroup agonism, such as in langurs [155] and gorillas [156], it is to defend access to mates. Harem leaders may indirectly defend resources [157], but probably only ephemerally, and short tenures prevent them from any kind of long-term claim to a group or territory. Here, I considered whether socially monogamous primates pose a potential exception, as paternity certainties are comparatively high and mated pairs control access to territories that are critical for breeding success. Yet, turnover also appears to be high within these species, territories are small, and inheritance from within groups rarely occurs. In summary, to the extent that territories or resources are ‘owned' by non-human primates, these systems are not conducive to patrilineal inheritance, nor to inheritance by individuals at all. However, when territories are defended collectively by the philopatric sex, such as by male chimpanzees, the results might be interpreted as collective inheritance by their descendants. In such cases, inheritance could occur even without specific recognition of descent, i.e. males need not know who their sons are, but these sons would inherit the fruits of their labour. Additionally, persistent kin bias and differential reproductive success may affect the genetic structures of territorial groups in ways that are more similar to inheritance than they might appear. For example, red howler monkey (Alouatta seniculus), groups are initially formed by cohorts of unrelated females, but via selective recruitment of daughters and expulsion of rivals, a territory can, over many generations, be controlled by the descendants of a single female [158].

While accumulated wealth has important impacts on the dynamics of human social organization, ownership and inheritance of material wealth are not universal human characteristics and were unlikely to be salient for much of our species' history. Recent cross-cultural surveys emphasize the extent to which inequalities in embodied wealth (e.g. somatic condition, knowledge) and relational wealth (e.g. social networks) are inherited within small-scale horticulturalist and foraging societies [159–161]. These analyses highlight foundational processes of familial inheritance that plausibly have roots in our primate ancestors, and if so, should parsimoniously follow matrilineal relationships. While a handful of primate studies have examined whether fitness traits in primates are heritable [162–165], the roles of non-genetic modes of inheritance (e.g. transmission of ecological knowledge, parental investment) have not been elucidated. For example, sex-biased parental investment [11,166] is conceptually linked with lineal forms of inheritance, but adaptive differences in parental investment have been notoriously difficult to demonstrate in non-human primates [167,168].

The rigid hierarchical systems of cercopithecines suggest that descent governs the inheritance of rank through matrilines across multiple generations. However, the proximate maintenance of these hierarchies indicates that ranks are not merely acquired but require consistent reinforcement by alliances among living kin with shared competitive interests. In these systems, kinship among adults may serve largely as a efficient vehicle for reciprocity or mutualism and may thus function in similar ways between and within generations [169]. If so, the increased importance of the vertical structuring of kin support from older generations of adults to younger ones [170–174] may have been a critical pathway to the robust inheritance systems observed in humans.

There are sound reasons to conclude that the architecture of human kinship organization has layers of complexity that do not have analogues in our phylogenetic relatives. How this affects biological outcomes is an open question that can only be addressed with efforts to translate across disciplines. Few primate studies have addressed vertical functions of kinship, particularly the significance of relationships between adult individuals and either parents and whether non-genetic components of fitness might be inherited. Studies on other long-lived, socially complex species, such as elephants and orca, have revealed the importance of older individuals as reservoirs of ecological knowledge that benefit their descendants [175–177], and that there are socially inherited components of social networks [178,179]; matrilineal kin groups seem to be particularly important to these dynamics. The human literature, in focusing on nominal kinship structures and intergenerational resource flows, has less often addressed the qualitative dynamics of cooperative interactions within kin groups, particularly within generations, as primatologists have. In both disciplines, inferences about analogy of kinship structures and the significance of matriliny would be enhanced by a better understanding of their biological consequences on populations and lineage success. It is clear that socioecological and cultural ‘rules' are often broken and that the underlying genetic structure of populations does not always align with the behaviours observed. This probably leads us to underestimation of the importance of maternal kinship. Concerted efforts for cross-species examination will allow us to evaluate whether human cultural rules have simply reinforced the more evolutionarily ancient foundations of kinship or have transformed them.

Data accessibility

This article has no additional data.

Competing interests

I declare I have no competing interests.

Funding

I received no funding for this study.