Cooperation in primates

Akihiro Eguchi


All primates are social whether they are solitary or gregarious, diurnal or nocturnal, and they cooperate in various different ways. Most of them form groups and develop patterned relationships within general areas over long periods of time. The grouping types vary depending on species and environmental factors, and the variations include multi-male and multi-female, uni-male and multi-female, uni-female and multi-male, uni-male and uni-female, or dispersed societies (Fuentes, 2007). Because humans live in groups, the grouping behavior may be unquestioningly accepted, but in fact, the reason why they live in groups is a fundamental question that should be asked from an evolutionary perspective. It is because, in a general sense, it would be preferred for animals to forage alone rather than having to aggressively compete for limited food resources among group members (Wittig & Boesch, 2003). Therefore, many researchers have been seeking an explanation of grouping behavior and how it overcomes the disadvantages of increased competition, and they have developed three concrete theories (kin-selection, reciprocal altruism, and mutualism) to explain the cause in past several decades.

In the earliest days of the development of the science of primatology, Wrangham (1980) proposed the inter-group feeding competition hypothesis that supposes that grouping is actually beneficial to cooperatively defend food resources by showing an evolutionarily stable strategy (ESS) over the competition. He described how food distribution determines the balance between within- and between-contest competitions based on the defensibility of resources, and the balance becomes the key for the grouping behavior. On the other hand, van Schaik (1983) argued that predation risk is the exclusive cause for grouping behavior and explained that living in groups enhances predation defense and overcomes the disadvantages by reducing the likelihood of any specific individual being attacked. This theory was called selfish herd and was mathematically explained by Hamilton (1971). Those antagonistic theories were actually both supported with data by Kumar (1995) in his study of lion-tailed macaques (Macaca silenus) and by Takahata et al. (1998) in their study of Japanese macaques (Macaca fuscata) respectively, and this hypothesis became the basis of the research to find out the driving force of the grouping behavior in primates.

Based on these two theories with further analysis, Sterck, Watts, and van Schaik (1997) developed a synthetic current socioecological model to describe a fundamental mechanism of group formation. This model explains that predation risk forces females to live in groups; however, group living leads to competition over limited food resources both within and between groups. Depending on the quantity, quality, and distribution of food, the group size of females is determined, and then the male distributions and competitive regime, which determines social relationships, follow. Even though there are some skeptical views toward this model because of inconsistencies in various data sources, such as no correlation between food distribution and competitive style in some populations (Menard, 2004; Thierry, 2008), this model became one of the main streams of research in primate ecology.

The social relationships developed based on those factors are complex, but they can be simplified into two basic types of interactions: competitive or cooperative. People usually use the word cooperation to refer to an action for a common purpose, mutually benefitting each other. However, according to Kappeler and van Schaik (2005), in an evolutionary biology, cooperation refers to actions or traits “characterized by costs to an actor and benefits to other nonspecifics” (p. v).  These cooperative interactions are actually more common than competitive interactions (Sussman & Garber, 2004) and emerge under a wide variety of contexts in different forms. Many researchers assume that these cooperative behaviors can be broadly explained by three different mechanisms: kin-selection, reciprocal altruism, and mutualism.

A large body of observational data has revealed that most primates usually prefer cooperation with kin than with non-kin and with close kin rather than with distant kin (Silk, 2005). For example, within dispersed and solitary foraging species, matrilineal related female mouse lemurs (Microcebus murinus) gather together to sleep for increasing predator defense, helping thermoregulation, and solving a shortage of suitable roots (Eberle & Kappeler, 2006). Similarly, even within a solitary species like orangutans (Pongo pygmaeus), a mother and a daughter tend to live closer together and they show considerable range overlap (Singleton & van Schaik, 2000).  In some other species that form matrilineal dominance hierarchies like rhesus macaques (Macaca mulatta) and Japanese macaques (Macaca fuscata), a daughter will inherit a rank from her mother (Widdig et al., 2006; Yamagiwa, 2008), and the higher the rank in the hierarchy they become, more the reproductive success they tend to have. Another interesting behavior observed within long-tailed macaques (Macaca fasciculari) was that mothers exaggerated their action in tool-use if their offspring are around, possibly to facilitate the learning of the action by their own infants (Masataka et al., 2009). This kind of kin-bias in behavior is called nepotism, and often observed in various different ways in primates.

According to Silk (2005), male primates are less likely to show nepotism possibly because of the difficulty of recognizing their kin. Experiments conducted by Rendall (2004) showed that males cannot distinguish unfamiliar kin from non-kin, and the distinction is solely based on cues like familiarity, mating history, age, etc. Additionally, females in many species try to confuse paternity through promiscuity as a counterstrategy for infanticide by males (Agrell, Wolff, & Ylönen, 1998), which results in males having trouble distinguishing their children from those of others.

However, evidence shows that infanticidal males in a population of blue monkeys (Cercopithecus mitis) still selectively target other unrelated infants (Cords & Fuller, 2010), and hanuman langurs (Presbytis entellus) selectively protect their infants from enemies (Borries et al., 1999). Also, even if the kin distinction between father and children is difficult, children are able to recognize their matrilineal relationship within their group. For example, cooperative behaviors of chimpanzees (Pan troglodytes) like grooming, co-foraging, and sharing foods tend to be among maternal kin (Mitani, 2009). These data show that regardless of the gender, most primates exhibit cooperative behaviors, which commonly involve kin-biases.

In spite of the large body of evidence of cooperative behaviors within primates, some types of self-sacrificing cooperative behavior contradict the theory of natural selection proposed by Darwin (1859); It is because although the law of natural selection indicates that each individual tries to maximize their fitness in the successive generation by outperforming other members in a species, the altruistic cooperative behavior may result in a loss of the fitness of an actor. In order to provide a reasonable explanation, Wynne-Edwards (1962) supported group selection theory that expanded the unit of the natural selection from the individual to the group. He believed that individuals in a group are willing to sacrifice themselves in order to maximize the overall fitness of the group.

However, this theory was strongly criticized by many researchers like Smith (1964) and Hamilton (1964), and in fact, this theory can easily be disproven by showing the strategy is not an ESS. Suppose in a cooperative group, a selfish individual appears as a result of a mutation. Based on natural selection, the selfish individual tries to maximize its own fitness thus passing more genes to the next generation. As time goes, the cooperative individuals will eventually be wiped out. Therefore, the group selection theory cannot explain the reason for cooperation.

Therefore, instead, Hamilton (1964) explained cooperative behavior by kin selection theory with an idea of inclusive fitness. Inclusive fitness is determined based not only on the actor’s own fitness but also on genetic relatedness with the recipient of his cooperation and the degree in which he can impact their fitness. Therefore, if the cost to the individual is less than the benefit of closely related kin, it tends to exhibit altruistic behaviors, thus resulting in kin-biases in behaviors. He described the theory in an inequality rB > C where r is relatedness, B is benefit to the recipient, and C is a cost for the actor. Therefore, this theory predicts that the cooperation is more common in kin, and any costly cooperative behaviors are limited to be only in close kin.

The theory of kin-selection successfully explained many cooperative behaviors; however, there are several skeptic views toward the mechanism as well. Silk (2005) pointed out that there is no reasonable way to estimate the change in fitness of individuals due to a certain action, and so it is not possible to precisely evaluate Hamilton’s rule. Additionally, many reports indicate that cooperation can often occur even among non-kin, so kin-selection cannot explain all cooperative behaviors of primates. For example, among blue monkeys (Cercopithecus mitis), there is no kin-preference in grooming behavior; however, there is a correlation between grooming and rank in agonistic dominance linear hierarchy (Rowell, Wilson, & Cords, 1991). Therefore, researchers believed that there should be some other factors that influence cooperative behavior rather than, or in addition to the kin selection.

One experiment conducted by Seyfarth and Cheney (1984) showed a hint for answering the question. It is known that vervet monkeys (Chlorocebus pygerythrus) use different vocal semi-language to share information or to signal alarm calls depending on different predators (Seyfarth, Cheney, & Marler, 1980). Seyfarth and Cheney (1984) found that vervet monkeys show a greater willingness to respond to vocal solicitations for aid from individuals who have previously groomed them. Similarly, long-tailed macaques (Macaca fascicularis) showed a willingness to help individuals who have groomed them before when agonistic competition occurs within a group (Hemelrijk, 1994). This type of evidence proposed that the cooperative behaviors like grooming or supporting in fight are not an altruistic but rather mutually beneficial.

Another example which cannot be explained by kin-selection alone is cooperative breeding, which is commonly observed within polyandry species like Calitrichids. Generally, males of most species do not provide any parental care to their children because the primary concern of males is to mate with as many females as possible. However, this case is an exception. Because the females usually bear twins, it is not possible for females to carry infants around while catching bugs to eat and producing highly condensed milk for their infants. Therefore, males and siblings in the group extensively help to raise those infants in exchange for living in a group to reduce predation risk. However, many studies show that there is no difference in the amount of help given to infants between kin and non-kin (Silk, 2005). Therefore, apparently there are many cooperative behaviors even within non-kin.

In order to explain this type of cooperation among non-kin, Trivers (1971) formalized a theory of reciprocal altruism, which supposes that the altruistic act is actually not intended to give away a favor to others but to invest for future reciprocity from the recipient of the favor. In other words, he assumed that the cost from altruistic cooperative behavior to non-kin individual will be complemented later by similar cooperative behavior from the recipient. Because in this way, both helper and recipient mutually benefit in overall, the cooperation can be now explained contingently to the evolutionary theory.

Axelrod and Hamilton (1981) explained the altruistic reciprocity hypothesis within the context of the Prisoner’s Dilemma in game theory, and they proposed an ESS, Tit-for-tat, to represent the reciprocal altruism. This strategy has only two simple rules to represent the reciprocity: showing a cooperative behavior in the first move, and then imitating whatever the opponent did from the preceding move. This strategy was later refined by adding an adaptive mechanism to deal with mutation with some degree of forgiveness for one time uncooperative act of a partner (Nowak, 1990). Since then, these types of theories derived from interdisciplinary field of study became popular, and much literature focuses on various alternative explanations (Trivers, 2006). These explanations provided many possible reasons for cooperation based on reciprocity in a social world.

Beside these theory-based discussions, de Waal and Bosnan (2005) discussed a reasonable and precise mechanism of the cost offset by later benefit within reciprocal altruism by focusing on the cognitive capability of various non-human primate species. They listed three different types of mechanisms of reciprocity based on different levels of cognitive demand. The least cognitive demanding mechanism they explained is symmetry-based reciprocity, which is based on general relationship influenced by factors like time spent together, relatedness, and gender. They believe mechanism serves the basis of cooperation, but most of the time this mechanism does not sufficiently determine the exact pattern of reciprocity.

Conversely, the most cognitive demanding mechanism they explained is calculated reciprocity, which is based on a strict scorekeeping of cost and benefit brought by each individual over a long time interval. For example, rhesus monkeys (Macaca mulatta) give favor to others depending on how much favor they are given recently by the each, and chimpanzees (Pan troglodytes) even seek payback from others depending on how much they are harmed by the individual (de Waal & Luttrell, 1988).  However, an experiment conducted by de Waal (2000) with capuchin monkeys (Cebus apella) found another mechanism, attitudinal reciprocity, which reflects an altruistic attitude toward other individuals of the group at the time. This type of altruistic attitude varies depending on different partners but does not require any scorekeeping; therefore, this mechanism is in between the symmetry-based reciprocity and the calculated reciprocity in terms of cognitive demand. Based on this type of categorization, it is important to take actual capabilities of species into account in order to figure out the exact mechanism they are using.

Mutualism is another variation of cooperation, in which two or more individuals engage in cooperative behaviors that aim to benefit all partners (Boesch, Boesch, & Vigilant, 2005).  This behavior is different from reciprocal altruism because all members act and benefit at the same time. Coalitions of primates can be an example of mutualism, which refer to cooperative fights against other individuals for various purposes including to protect from other stronger individuals, to compete for limited resources, and to acquire higher social rank (van Schaik, Pandit, & Vogel, 2005). However, because coalitions are costly behaviors that potentially result in serious injury, the question is how come this type of behavior has evolved.

One possible reason can be found in a classic theory of intergroup feeding competition proposed by Wrangham (1980), which is actually an ESS if resources are clumped in high-quality patches, and limited numbers of individuals exploit the resources. Suppose there is a fruit tree that can support only two individuals, but there are four individuals A, B, C, and D that forming a linear dominance hierarchy, in that order. A and B are likely to monopolize the resources on the tree. In order for C and D to obtain the resources, they need to cooperate to compete against A and B individually. Now, A and B need to cooperate to get the resources back again. Therefore, theoretically, this type of cooperation can naturally occur even if the behavior involves risk of injury, thereby making the choice of partner more important.

Many reports also indicate that the coalition partner choice is strongly influenced by their maternal relatedness. For example, the pattern of coalition formation is significantly influenced by the relatedness among rhesus macaques (Macaca mulatta). If two are maternal kin, they often affiliate and rarely agonistically compete against each other when compared to non-kin or paternal kin (Widdig et al., 2006). Chacma baboons (Papio hamadryas ursinus) also form close and supportive cooperative relationships within close kin, especially a mother and daughter (Silk et al., 2010). Therefore, this type of cooperative behavior is contingent to Hamilton's rule that indicates costly behavior is limited toward close kin (Hamilton, 1964).

Based on this kind of evidence, van Schaik, Pandit, and Vogel (2005) explained that competition over limited food resources among females results in strongly kin-biased coalitional formation, thus making most female primates a philopatry, the behavior to stay in their natal group. On the other hand, because resources for males are females, of which the availability is less stable than food availability, male-male coalitional patterns tends to be more complex involving dynamics of main three categories: conservative style to suppress threatening males by higher ranked males, bridging style to attack a male in middle rank by highly ranked males and lowly ranked males, and revolutionary style to overtake the rank by lowly ranked males, depending on situations (Chapais, 1995).

One problem with the mutualism was its vulnerability because in many types of mutualism, members can take advantage of the benefit by free riding without contributing anything. However, this problem seems to be well taken care of within any species which exhibits mutually cooperative behaviors. For example, savannah baboons (Papio cynocephalus), macaques, and vervet monkeys (Cercopithecus aethiops) as reported by Cheney, Seyfarth, and Smuts (1986), form long-term stable bonds by increasing opportunity to interact and build a strong trust in the relationships; at the same time, any free riding is continuously checked and excluded.

These three different explanations for cooperative behavior among primates are not mutually exclusive, but any combination would result in various cooperative behaviors. This is an important topic to study because even though most primate species including humans show some types of grouping behavior, for any animal it can actually be detrimental because they have to compete for limited resources. Years of research indicate that the main explanations for the cause of this behavior seems to be either kin-selection, altruistic reciprocity, or mutualism depending on the context. Kin-selection well explains the high likelihood of kin-biases in behaviors of primates. The altruistic reciprocity explains the cooperative behavior among non-kin. The mutualism explains behaviors like coalitions. Even though they were naturally forced to group together because of predation risk and food availability, each species developed various unique cooperative strategies to deal with situations.