The Princeton Guide to Evolution

Evolution of Eusociality

Laurent Keller and Michel Chapuisat

OUTLINE

1. Eusociality: A highly integrated form of social organization

2. What drives eusociality?

3. Working together

4. Intragroup conflicts and their resolution

Animal societies can reach very high levels of coordination and integration. In ants, bees, termites, and naked mole rats, hundreds of permanently nonreproducing workers help rear the offspring of a few fertile individuals, the queens and males. Societies with such a reproductive division of labor are called eusocial. The evolution of eusociality puzzled Darwin: How could workers pass on their characteristics to the next generation if they did not reproduce? W. D. Hamilton provided the answer in the 1960s with the concept of kin selection, the indirect transmission of genes through relatives, which occurs in stable associations of related individuals jointly exploiting and defending common resources. Despite the high level of cooperation characterizing eusocial societies, conflicts among individuals are still common, and sophisticated social mechanisms often contribute to maintaining social cohesion. Many eusocial species are extremely successful, because the coordinated and cooperative work of many individuals allows them to efficiently use and transform their environment while being robust to perturbations.

GLOSSARY

Cooperation. A collective action that benefits two or more individuals.

Eusociality. A social organization that includes reproductive division of labor, cooperative brood care, and overlap of generations.

Haplodiploidy. A sex determination system in which males are derived from unfertilized (haploid) eggs, and females from fertilized (diploid) eggs.

Kin Selection. Selection on genes for social traits that affect the fitness of relatives.

Philopatry. The tendency to stay close or return to the natal site.

Relatedness. The probability that individuals share identical alleles inherited from recent common ancestors.

Reproductive Altruism. An action by which an individual decreases its own reproduction to help one or several other individuals reproduce.

Self-organization. A spontaneous organization arising from local interactions among individuals, without central control.

1. EUSOCIALITY: A HIGHLY INTEGRATED FORM OF SOCIAL ORGANIZATION

Insect societies have long fascinated human beings, as illustrated by the Bible verse “Go to the ant, thou sluggard; consider her ways, and be wise” (Proverbs 6:6), which acknowledges the industrious nature of insect workers. The major organizing principle of insect societies is reproductive division of labor, whereby one or a few individuals (the queens) specialize in reproduction, whereas the others (the workers) participate in cooperative tasks such as building the nest, collecting food, rearing the young, and defending the colony. This division of labor may lead to amazing specializations, such as ants with a thickened and enlarged head that they use as an armored door to block the nest entrance, or “kamikaze” termites that explode to glue their opponents in toxic secretions.

The term eusociality refers to animal societies with a marked and permanent reproductive division of labor. In most of these societies, only one or very few individuals reproduce, although the number of potentially reproductive adults may number in the hundreds or thousands. In general, these societies also exhibit cooperative brood care and have overlapping generations. Eusociality was traditionally thought to occur only in insects; all species of ants and termites, as well as part of the bees and wasps, are eusocial. Colonies of termites are headed by one or a few queens and kings, and workers can be either males or females. In contrast, ants, wasps, and bees form matriarchal colonies in which all workers are females. In these colonies, there may be one or more queens, which usually mate once in early adulthood and store the sperm that they use throughout their lives.

Recently, eusociality was discovered in a small number of species belonging to diverse taxa of invertebrates, including parasitic flatworms, snapping shrimps, ambrosia beetles, gall-forming thrips, and gall-forming aphids. Moreover, two species of mammals were recognized to be eusocial, the naked mole rat and the Damaraland mole rat. The discovery of several taxa with social organizations similar to those of some social insects, and the realization that there is a continuum in the extent to which individuals abstain from reproducing, have led to the conclusion that it is somewhat difficult to classify species as being eusocial or not. For example, in many cooperatively breeding birds and mammals, some individuals forgo their own reproduction to help raise the offspring of others, but generally, this reproductive altruism is reversible and may cease if a breeding opportunity becomes available. This reversibility occurs in some wasps as well.

Despite some uncertainties about the transience or permanence of the reproductive division of labor, the number of truly eusocial species is relatively small, accounting for approximately 1 percent of the described animal species. However, eusocial species are ecologically very successful, being found in almost every type of terrestrial environment and making up a considerable proportion of the animal biomass of the earth—up to 50 percent in some tropical habitats. In many ecosystems, eusocial insects, ants in particular, are dominant organisms that play a crucial role as predators and pollinators, and in soil formation. This tremendous ecological success of eusocial insects is undoubtedly due to their social organization, based on large numbers of individuals cooperating in social groups, which provides multiple competitive advantages over solitary species.

2. WHAT DRIVES EUSOCIALITY?

An Apparent Paradox

The ecological success of eusocial species is based on an organization that seems to contradict the principle of natural selection. In the process of natural selection, genes conferring greater survival and reproduction increase in frequency over generations in a population, since individuals carrying these genes leave more descendants. This feedback loop explains the adaptation of organisms to their environment as alleles underlying more successful traits spread and increase in a population. Yet among many eusocial species, workers have particular morphology and physiology preventing them from reproducing. The paradox of why, in some animal societies, a proportion of the individuals in a population forgo reproduction to assist other group members did not escape the attention of Charles Darwin. In The Origin he noted that sterile workers of eusocial insects embodied “one special difficulty, which at first appeared to me insuperable, and actually fatal to my whole theory” (1859). Although inheritance mechanisms were not known at this time, Darwin drafted a solution to this apparent paradox, namely, that selection may operate not only at the level of the individual but also at the level of the family.

Kin Selection

In the mid-1960s, as a graduate student, William D. Hamilton resolved the paradox of reproductive altruism by showing that individuals can transmit copies of their own genes not only through their own reproduction but also by favoring the reproduction of kin, such as siblings or cousins. Kin share identical copies of genes inherited from their common ancestors in the same way a child possesses copies of paternal and maternal genes. Thus by helping their mother produce numerous fertile offspring (the males and the future queens), sterile workers have an excellent way of transmitting copies of their own genes to the next generation. This indirect selection on genes due to their effect on the fitness of relatives has been called kin selection.

Hamilton’s approach is based on the Darwinian principle of natural selection on genes, combined with Mendel’s genetics. The fundamental idea is to consider not only the direct reproductive output, or fitness, of individuals but their inclusive fitness, which includes the indirect effects of genes on the fitness of other individuals carrying copies of the same genes (see chapter III.4). A specific application of the general theory of inclusive fitness is to examine the conditions favoring the evolution of reproductive altruism through the mechanism of kin selection. Hamilton’s rule delineates when an individual transmits more copies of its own genes by behaving altruistically, that is, decreasing its own chances of survival and reproduction to help other individuals reproduce. In short, reproductive altruism is favored when the loss in the altruistic actor’s personal fitness is smaller than the gain in the personal fitness of the recipient of the altruistic act multiplied by the relatedness between the actor and the recipient. Personal fitness is an estimate of the number of descendants of an individual, relative to other individuals in the population, while the relatedness is the probability that the recipient carries copies of the actor’s genes, inherited from recent common ancestors.

A general description of Hamilton’s rule is that altruistic acts are more likely to be selected for when individuals are closely related and when the decrease in the actor’s personal fitness is relatively small compared with the increase in the recipient’s fitness. However, it is important to note that reproductive altruism does not require exceptionally high degrees of relatedness. It can evolve among distantly related individuals if the benefits for the recipient are high and the costs to the actor low.

The following simple example illustrates Hamilton’s rule. Imagine a gene that programs an individual to die so as to save relatives’ lives. One copy of the gene will be lost if the altruist dies, but the gene will increase in frequency in the population if, on average, the altruistic act saves the lives of more than two siblings (relatedness = 0.5), more than four nephews or nieces (relatedness = 0.25), or more than eight cousins (relatedness = 0.125). J.B.S. Haldane fully apprehended kin selection theory and Hamilton’s rule when he announced, having done some calculations on an envelope in a pub, that he would be ready to give his life to save two brothers or eight cousins.

There is unambiguous theoretical and empirical evidence that kin selection has been central to the evolution of eusociality and reproductive altruism by workers. First, for the problem to be resolved, all models accounting for the evolution of altruism explicitly or implicitly assume that altruistic individuals and recipients of help are related. Second, eusociality invariably evolves within groups of highly related individuals, such as one mother and her offspring. Well-marked reproductive division of labor is rare in societies where individuals are distantly related. There are a few ant species in which the relatedness between nest mates is indistinguishable from zero, but this low relatedness stems from an increase in queen number that occurred long after the evolution of morphological castes and reproductive division of labor. Workers of these ants generally have only vestigial (shrunken) ovaries, and it has been suggested that obligate sterility of workers prevented such societies from collapsing after the drop in relatedness. However, societies with very large numbers of queens are expected to be unstable in the long term, as there is no more selection for workers’ altruism.

Ecological and Life History Factors Favoring the Evolution of Eusociality

The evolution of eusociality depends on a combination of ecological, genetic, social, and life history factors that jointly determine whether the best option for a young individual is to stay in the group and sacrifice part or all of its direct reproduction to help others, or to leave the group to breed independently (see chapter VII.10). The payoffs of each strategy are determined by the benefits of helping in terms of group productivity, the genetic relatedness among group members, and the expected success of a young individual that attempts to reproduce solitarily. In turn, these three elements depend on ecological conditions—in particular, on the degree of ecological constraints on independent breeding—as well as on the ecology, life history, breeding system, and family structure of a species. For example, in some cooperatively breeding birds, the decision about whether to become a helper depends on territory availability. In other taxa, such as paper wasps, suitable nest sites are not limited, and females seem to associate because of the benefits conferred by sociality, particularly lower breeding failure.

An ecological factor that favors the evolution of eusociality is the coincidence of shelter and food. Thus, eusocial thrips, aphids, beetles, shrimps, termites, and mole rats live in cavities or burrows in which they obtain their food. Similarly, eusocial flatworms form colonies within their molluskan hosts. Such ways of life may promote sociality for several reasons. First, the high value of a habitat combining food and shelter may favor altruistic self-sacrifice for colony defense, leading to the evolution of a soldier caste. Second, this type of valuable habitat may select for philopatry and helping because the colony can be inherited by offspring, and some helpers may have a chance to replace breeders. Third, living in a confined habitat helps keep relatives in physical proximity and thereby creates opportunities for kin-selected reproductive altruism. Finally, because juveniles in such habitats are frequently self-sufficient with regard to food, they can devote themselves more directly, and at a younger age, to helping raise younger siblings.

High risks of mortality during the period of brood rearing may also promote eusociality, because sociality provides life insurance: if a cooperative breeder or a helper dies early, its investment is not lost, because other group members will finish rearing related brood. In contrast, a solitary breeder that dies before having raised its brood will have zero fitness. Hence, long development time and high mortality risk should favor helping and eusociality. These conditions occur in many species of Hymenoptera, such as ants, wasps, and bees, which have extended parental care and search for food outside nests.

The Role of the Family Structure

One factor that played a crucial role in facilitating the evolution of reproductive altruism is the type of family structure. Mother-daughter associations not only provide an opportunity for the offspring to help while they are still juveniles but also generate a genetic structure that favors the evolution of reproductive altruism.

Forty years ago, there was considerable discussion about whether eusociality evolved within groups composed of a mother and her offspring, or within groups composed of related individuals of the same generation, such as sisters. The asymmetry in relatedness occurring in mother-daughter associations should favor the evolution of eusociality. In such associations, daughters are on average as related to their mother’s offspring as to their own descendants (i.e., they share half their genes by recent common ancestry, r = 0.5), so they lose nothing by giving up direct reproduction in favor of the mother. By contrast, mothers are twice more related to their own offspring (r = 0.5) than to their daughters’ offspring (r = 0.25); they thus benefit from monopolizing reproduction. Therefore, a pronounced reproductive division of labor and monopolization of reproduction by mothers is expected to evolve and be stable in mother-daughter associations. In contrast, siblings are always more related to their own offspring (r = 0.5) than to those of their sisters (r = 0.25) and therefore should not easily forgo their reproduction to help their sisters.

The level of relatedness among the queen’s daughters also depends on the number of fathers—it decreases when the mother queen mates with multiple males. In a comparative study of 267 species of eusocial Hymenoptera, William O. H. Hughes and colleagues showed that monoandry (a queen mating with a single male) was ancestral on each of the occasions when eusociality evolved and that multiple mating evolved only after workers had lost reproductive totipotency. In fact, the breeding system of most eusocial species is conducive to lifetime monogamy. In ants, bees, and wasps, for example, the queens mate only during the mating flight, hence ensuring high relatedness of their offspring throughout their life. Similarly, termite colonies are typically initiated by a royal couple, which jointly produce all the colony offspring. Thus, the breeding system of species that became permanently eusocial was conducive to the formation of simple families in which offspring were highly related and thus more likely to forgo reproduction if this increased colony survival and productivity.

Caste Differentiation

Colonies of many eusocial insects (e.g., the honey bee, vespine wasps, and most ants and termites) contain distinct morphological castes: the queens are morphologically and physiologically specialized for reproduction, and the workers for other tasks such as foraging and brood care. The degree of polymorphism varies greatly. In some species such as allodapine bees, hover wasps, polistes wasps, and sweat bees, there is little morphological difference between queens and workers; the specialization is mostly behavioral. By contrast, many ants and termites are characterized by striking differences between queens and workers, with the latter sometimes having completely lost their ovaries and developed morphologies adapted for special tasks.

A broad comparison among social insects reveals an association between the queen/worker dimorphism and colony size. Morphological differences between queens and workers are generally absent or small in species forming small colonies, whereas the differences are well marked in species forming large colonies. Such an association can be explained by a relationship between colony size and the probability of workers to become replacement reproductives. This probability drastically decreases in larger colonies, with the effect that there is lower selection to retain reproductive ability. The only vertebrate species in which morphological castes have evolved is the naked mole rat. In this species, the lumbar vertebrae of the breeding female elongate after the onset of reproduction. Of all the vertebrate species, the naked mole rat forms the largest societies, with up to 300 individuals.

3. WORKING TOGETHER

Division of labor among workers plays a major role in the great ecological success of eusocial species. Analogously to somatic tissues in a multicellular organism, workers can specialize in various tasks (division of labor). Moreover, tasks can be divided in sequential actions performed by more than one individual (task partitioning). Hence, work can be performed collectively, with concurrent operations and synergistic interactions generating large benefits to the whole colony. In a feedback loop, the benefits of collective work further promote the evolution of reproductive altruism and contribute to stabilize eusociality. Over evolutionary time, selection has favored societies in which the work is organized in an efficient, robust, and flexible manner.

Efficiency, Robustness, and Flexibility

There are several ways by which division of labor can increase colony performance. First, the capacity to perform tasks concurrently often provides large advantages. In solitary species, a single individual conducts one task at a time, often in a specific sequence, and must complete a set of tasks to reproduce successfully. For example, a solitary sphecid wasp has to excavate a nest, find a prey item, sting it, bring it back to the nest, and then lay eggs. In contrast, eusocial species can conduct many tasks at the same time, seizing opportunities as they arise. The efficiency of eusocial species is further increased by the collective performance of tasks that would be out of reach of single individuals. For example, colonies of naked mole rats excavate burrows that can be more than 4 km in cumulative length. Six small ants can immobilize a large insect by seizing one leg each, scouts can recruit foragers to a rich food source, or nest temperature can be accurately controlled at all times. Finally, by repeating the same task in one area of the colony territory, for example, collecting food, feeding the brood, or guarding the nest entrance, workers can learn and become more efficient. They also minimize costs associated with traveling between tasks, and time lost in task switching.

Another feature of concurrent systems is robustness. The failure of one individual at one task does not compromise the whole enterprise. The redundancy of the system, with many individuals performing the same task and many concurrent production lines, makes it resistant to perturbations or catastrophic events. Finally, it is important to stress that workers do not usually work in a fixed and rigid way. They show behavioral flexibility, so that the number of workers engaged in each task can vary over time to match the needs of the colony and the changes in the environment.

Mechanisms Regulating the Division of Labor

Colonies face the complex challenge of dynamically allocating the correct number of workers to each task. Early researchers on division of labor considered that workers were rigidly programmed to perform only one task over long periods of their life, with task performance being determined by internal factors such as age, size, or morphology. Indeed, there is often a correlation between age and task in the social insects. Young individuals usually perform tasks within the colony, such as brood care or nest maintenance, while older individuals engage in outside, more risky jobs, such as foraging or colony defense. However, a fixed partitioning of tasks according to age or other internal factors gives little flexibility.

Despite physiological or age-related predispositions for certain tasks, workers are usually able to switch tasks according to needs. For example, if one behavioral caste is experimentally removed, nurses become foragers, or foragers switch to guards. However, task switching is likely to be costly and should occur only when necessary. More recently, researchers have considered that the colony is a self-organizing system in which a flexible division of labor arises from the independent actions and decisions of workers, without any central or hierarchical control. Several models in which division of labor emerges by self-organization have been proposed, based on spatial location, task encounter, or physiological threshold. An important class of models is based on response thresholds, with workers performing a task when a specific stimulus for this task exceeds their individual threshold. In the response threshold model, the task and stimulus are linked in a negative feedback loop that regulates the system, as performing the task decreases the stimulus for this particular task. Variation in response thresholds among individuals results in worker specialization; however, the system retains flexibility and self-adjusts to needs. For example, honey bee workers will start to fan to cool the hive if the temperature exceeds a given threshold. The ones with the lowest response threshold will start to fan first. By so doing they decrease the temperature, which may not reach the threshold of other workers under normal conditions. However, if the temperature continues to rise, other workers with higher response thresholds will start to fan. Hence, a subset of workers become task specialists because of small differences in threshold response, but all workers are able to perform the task if needed. Variation in response threshold can come from many sources, including genotypic differences or experience. A division of labor may emerge spontaneously when individuals with different response thresholds group together. Selection can then favor response threshold distributions that ensure the most efficient allocation of workers to tasks.

4. INTRAGROUP CONFLICTS AND THEIR RESOLUTION

Despite high levels of cooperation and apparent harmony, potential conflicts persist in colonies of eusocial species. Potential conflicts arise because, in contrast with cells of an organism, colony mates are not genetically identical (see chapter VII.9). Hence, kin selection predicts that individuals with partially divergent genetic interests may attempt to favor the propagation of their own genes, possibly to the detriment of their nest mates. Colony members can compete over direct reproduction or over allocation of colony resources to various relatives, and the potential conflict may translate into actual conflict or may remain unexpressed.

Conflict over who reproduces is common in many eusocial species. For example, dominance behavior and linear hierarchies frequently occur within small colonies of wasps, bees, and ants. Some potential conflicts are specific to the social Hymenoptera, which are male-haploid, female-diploid. Queens and workers may compete over the production of males and over the allocation of colony resources to males and females, respectively. These potential conflicts sometimes degenerate into open conflicts. In some ant species the queens and workers both try to influence the relative investment in females versus males. In these species, workers kill brothers to favor their more related sisters, while queens influence colony sex allocation by biasing the sex ratio of their eggs toward males.

Conflict Resolution

Within animal societies, the resolution of potential conflicts still results in a wide range of outcomes. The expression of conflict can range from high levels of actual conflict to its complete absence. Understanding how potential conflicts among individuals are resolved is important to comprehending the emergence of cooperation in social groups, the evolutionary transition toward eusociality, and the further increase in complexity of societies.

Several types of factors and mechanisms contribute to align the divergent interests of colony members, thereby favoring peaceful cooperation in cohesive social groups. A major factor is genetic homogeneity, which results in high and symmetrical degrees of relatedness among group members, thus reducing the area and magnitude of potential conflicts. Other important elements are the multiple benefits of group living, as compared with solitary breeding, as well as the costs of behaving selfishly. In short, solitary or selfish behaviors are most likely to be selected against when cooperation and division of labor provide large synergistic fitness benefits and when open conflicts decrease colony productivity.

Finally, multiple socially mediated mechanisms may contribute to restrain within-group selfishness. These social mechanisms may be based on pacific “social contracts,” such as leaving enough reproduction for each breeder to stay peacefully in the group. Social cohesion can also be enforced individually or collectively by direct actions against individuals that behave selfishly, in the form of aggression, coercion, or punishment. Proximately, reproductive altruism within eusocial groups may be socially enforced and may thus reach higher levels than the subordinates’ optimum. Power asymmetries, or unequal access to information, may tip the balance in favor of one party or another. In naked mole rats, the breeding female frequently attacks subordinate females and by so doing suppresses their reproductive attempts. In ants, bees, and wasps the workers often police one another: they suppress male-destined eggs laid by other workers, or ally against a sister that tries to overturn their mother. Overall, social processes such as coercion and policing appear to play a major role in preventing outbursts of conflicts within social groups and may thus be very important for the evolution and maintenance of eusociality.