The Princeton Guide to Evolution

VII.9

Cooperation and Conflict: Microbes to Humans

Joan E. Strassmann and David C. Queller

OUTLINE

1. What is cooperation and why is it so important?

2. Fraternal and egalitarian cooperation

3. Fraternal cooperation is explained by kin selection

4. Egalitarian cooperation requires direct benefits

5. Conflict and control of conflict in fraternal cooperative systems

6. Conflict and control of conflict in egalitarian cooperative systems

7. Organismality results from high cooperation and low conflict

Cooperative interactions characterize all life, giving us spectacular multicellular organisms like kelp and kangaroos; complex societies like army ants, and hyenas; and extensive cooperative networks, like pollinators and their plants. Fraternal cooperation among related, like entities explains multicellularity and social insects. Egalitarian cooperation among different entities explains pollination, cleaning stations, and bacteria-insect symbioses. For cooperation to flourish, exploitation must be controlled; when it is, organismality results.

GLOSSARY

Cheating. Behavior that benefits oneself at a cost to others under circumstances within an otherwise-cooperative framework.

Cooperation. An interaction that benefits the recipient and the actor either directly or indirectly.

Direct Benefit. A benefit that accrues to the actor’s personal fitness.

Egalitarian Cooperation. Cooperation between unrelated individuals of the same or different species; the payoff to the actor must be direct.

Fraternal Cooperation. Cooperation between relatives; payoff to the actor can be direct or indirect.

Hamilton’s Rule: rB – C > 0, where r is the relatedness between the altruist and the beneficiary, B is the increased fitness benefit to the beneficiary, and C is the cost in lost fitness to the altruist.

Inclusive Fitness. The fitness of an individual that includes all ways it has increased its genetic representation in the next generation, from rearing progeny to the share it contributed to nondescendant kin.

Indirect Benefit. A benefit of actions of the actor toward nondescendant kin, who tend to share the actor’s genes.

Mutualism. Cooperation between nonrelatives of the same or different species; often, different goods or services are exchanged.

Organism. A living unit with high cooperation and very low conflict among its parts; an adapted unit that is not much disrupted by conflict at lower levels, nor subsumed into adaptation at higher levels.

1. WHAT IS COOPERATION AND WHY IS IT SO IMPORTANT?

Cooperation in an evolutionary sense is defined as an action performed by an actor that benefits its recipient. For cooperation to be evolutionarily stable (see chapter VII.3 on game theory), it should also benefit the genes that underlie the action, as by benefiting the actor directly or by benefiting relatives of the actor. A ground squirrel that sounds an alarm when a coyote is spotted benefits relatives in hearing range, causing them to run for shelter. A social amoeba joins with others to form a motile slug that can travel farther and lift off the ground before forming spores, though in the process some cells die by becoming the supporting stalk and therefore never reproduce. A sterile worker bee helps the queen rear sisters and brothers. That same worker bee takes nectar and pollen from flowers, in return pollinating the ova. A bacterium produces light inside a squid, so the squid casts no shadow in the moonlight when seen from below. Termites are able to eat cellulose only with the aid of their bacterial gut inhabitants. These are just a few examples of evolved cooperative interactions.

Cooperation may be the most underestimated process in the evolution of life, one that affects nearly every topic in this book, from speciation to phylogenetics to adaptation. One reason for its importance is that it is often easier to acquire a new capability by allying with another that already has the capability, as compared with the slow and less reliable process of accumulating mutations that provide the trait de novo, as Nancy Moran has shown so elegantly in the mutualism between sap-sucking insects like aphids and sharpshooters and their bacterial symbionts. The bobtail squid can more easily evolve the ability to house and feed luminescent bacteria than become luminescent itself. Termites cannot easily evolve the ability to digest cellulose that appears to be so easy for their spirochete gut bacteria.

Partly for this reason, cooperation can be enormously successful. The cooperative engulfing of a blue-green alga by a eukaryotic cell produced the green plants on which so much of life depends. Cooperation is the basis of the phenomenal success of the social insects (see chapter VII.13 on eusociality) and is at the heart of the major transitions in life, as highlighted by John Maynard Smith and Eörs Sathmáry. Cooperation is the source of eukaryote cells, multicellularity, and organisms themselves.

Cooperation has been fixed in highly successful alliances, but the lack of current variability in many species makes these less than optimal for studying how cooperation came to be. The most fruitful study thus involves organisms where cooperative alliances are still plastic, subject to measurable pushes and pulls as the conflicting interests of different parties surge and are then quelled. Such organisms include social insects (see chapter VII.13) and vertebrates (see chapter VII.10), which were important in the early development of the theory, and extend to the powerful experimental systems of social microbes and within-genome alliances studied more recently.

Here we divide cooperation into two natural kinds. We then discuss benefits, costs, conflict, and how conflict is controlled. We hope the curious reader will push back the darkness in some new corners of our cooperative world.

2. FRATERNAL AND EGALITARIAN COOPERATION

Fungus-growing ants provide an illustration of the two major types of cooperation. Atta ant colonies contain millions of workers that methodically strip the tropical forest for leaves to feed their fungus. The ant workers occur in several fixed forms, or castes, all working to rear the offspring of a single large queen. At the same time, the ants cooperate with a fungus, a little package of which is carried by a new queen when she starts her colony. In addition to dispersing the fungus, the ants feed it the leaves they harvest and, in turn, eat parts of the fungus.

Cooperation can be either fraternal or egalitarian, according to whether the cooperators are the same kind of entity or are different entities, according to David Queller. Cooperation among the ants themselves is fraternal cooperation, based on shared genes between cooperators. Cooperation between the ants and the fungus they grow is egalitarian cooperation that requires that each party benefit. These terms are taken from the last two components of the French Revolution cry “Liberté, égalité, fraternité!” (the first, liberté, could be viewed as the noncooperative, solitary option). Fraternal cooperation results in alliances of like individuals, including the same molecules in compartments, organelles in cells, cells in multicellular individuals, and individuals in colonies or societies. Egalitarian cooperation results in alliances of different kinds, including different molecules in compartments, different genes in chromosomes, different organelles in cells, and cooperation among different species.

Fraternal cooperation involves cooperation among relatives and so need not pay back directly to the actor’s phenotype. Thus, it is the only kind of cooperation that extends to true altruism. The genes causing the action proliferate because they are also present in the recipient. This indirect benefit is explained by William D. Hamilton’s kin selection theory. Such cooperation is exhibited by social insects, birds with helpers, wolf packs, and multicellularity; many other kinds of cooperation are found among entities that share genes.

Egalitarian cooperation involves cooperative acts in which both parties benefit directly and includes all cooperation between different species, including cleaner fish and their clients, leaches and their blood-digesting bacteria, plants and fungal mycorrhizae, and plants with their pollinators. Unrelated individuals of the same species also cooperate only under conditions favorable to each, so the evolution of this kind of cooperation has more in common with egalitarian than fraternal cooperation, so we treat it under that heading. This category includes males and females in sexual relationships.

3. FRATERNAL COOPERATION IS EXPLAINED BY KIN SELECTION

Like any other evolved adaptation, cooperation must increase the frequency of the genes that cause it. Unlike other evolved adaptations, this may seem to be a challenge, because cooperation involves an action that costs the actor and benefits another. Genes for fraternal cooperation can spread because of their benefits to identical genes occurring in kin. Fraternal cooperation is favored under conditions specified by Hamilton’s rule, rB – C > 0, where r is genetic relatedness between donor and beneficiary, B is benefit to beneficiary in terms of increased progeny, and C is cost to donor in terms of lost progeny. Genetic relatedness measures shared genes above average frequencies and is usually due to pedigree connections (giving help to partners with average allele frequencies will not change frequencies). This genetic relatedness among cooperators is often estimated using variable Mendelian markers, like DNA microsatellites.

Fraternal cooperation includes true altruism, which is behavior that reduces the actor’s direct fitness, because nondescendant kin can pass on the genes that underlie the actions. Many of the cooperators we think of first fall into this category. Social insect colonies are based on families. Helpers at the nest in mammals and birds are usually older progeny, related to the brood they rear.

Costs and benefits can be measured in terms of progeny lost and nondescendant kin gained by following a particular strategy. These values are often measured by comparing different strategies. For example, a wasp foundress that nests alone and produces a certain number of progeny could be viewed as a stand-in for a wasp foundress that helps a sister instead of reproducing herself, to provide an estimate of what she might have produced. Comparing the two could give an estimate of the benefit of cooperation.

An important advantage in social insects is division of labor, which means that different participants do different things and may even have different forms. For example, some ant species have workers that are adapted as foragers, and others as soldiers, the latter being larger, with large mandibles for biting. However, the initial advantages were not based on these derived forms.

Predators are a strong initial selective force for sociality. The two main lifestyles in social insects—life insurance and fortress defense—are thought to have evolved to protect against predators and their effects. Life insurance means that adult cooperators can take over half-raised young and finish the job if some adults die. This lifestyle is most likely to apply in species in which adult lifetimes are short, and offspring dependency is long, as in ants, bees, and wasps. Fortress defense means cooperators can take advantage of safe, defensible places for nesting, particularly inside edible resources. Termites, naked mole rats, and social shrimp may be eusocial because of fortress defense.

When there are advantages, the actual trigger for cooperation can be very simple. The sweat bee Halictus rubicundus that Jeremy Field and colleagues study in the United Kingdom is solitary in Belfast but has females that remain to help their mother in Sussex, where the growing season is long enough for the firstborn to rear subsequent young. When Field moved females from the more northern, solitary population to the warmer climate in Sussex, some of the females remained with their mother in a cooperative alliance, making the fraternal transition from solitary to social breeding.

Another kind of fraternal cooperation is multicellularity, which has arisen many times and seems to be an easy evolutionary step, if David Kirk’s evidence from the volvocine algae is any indication. Multicellular organisms usually have a single-cell bottleneck at the beginning of development, which is crucial, because this bottleneck causes the cells in the multicellular organism to be genetically identical. This means that cells specializing into somatic functions are not in conflict with the gonad cells.

A major advantage of multicellularity is the division of labor among many different cell types, but when the first cells came together they were probably not already specialized, so some other advantages might have been involved. Studies of the early stages of multicellularity in the volvocine algae have begun to examine these advantages. Chlamydomonas reinhardtii is a single-celled alga with two flagella that allow it to move. In its multicellular relative Volvox carteri, somatic cells numbering around 2000 have flagella, and the 16 or so gonadal cells do not, a division of labor that allows it to swim and reproduce at the same time. But the initial advantages of multicellularity might be sought in intermediate forms, like Gonium species, with relatively few largely undifferentiated cells in the multicellular body. Still, it is not entirely clear what initial advantages multicellularity confers. High on the list of candidates are increased motility and reduced predation, both advantages of larger size, but the continued presence of both single-celled and multicellular species argues against a single best strategy.

4. EGALITARIAN COOPERATION REQUIRES DIRECT BENEFITS

Cooperation among unrelated individuals must provide a direct benefit to the actor if it is to evolve. Exactly how this process works is the subject of mutualism theory, discussed in the section on conflict. For example, a female meerkat might join a troop of unrelated individuals and help rear the babies if she has some chance of reproducing in the group in the future. A male and a female northern mockingbird cooperate to rear their progeny, as do many other organisms with biparental care. Cleaner fish eat the parasites and dead cells off the larger fish that come to the cleaning station.

Division of labor often provides advantages from the very beginning of egalitarian cooperation. Since the partners are typically different in egalitarian cooperation, as with males and females in sexual reproduction, it is easy to imagine that they have different talents that their partners can use. Egalitarian cooperation easily provides benefits to each partner, because what is easy for one may be difficult or impossible for the other. Flowering plants provide a reward to their pollinators, who, usually incidentally, carry pollen to waiting, stationary ovules. Animals rely on bacteria for food digestion, providing an environment for growth in return. When the food is highly specialized (e.g., plant sap, blood, or cellulose), only one or two bacteria species may take on the digestive task. The symbionts of some animals, like corals, make their food from sunlight. Plants rely on cooperative relationships to extract nutrients from the soil, as with mycorrhizae. Some bacteria fix nitrogen in exchange for plant-produced carbon. In each of these examples, each partner provides something that is relatively easy for it to manufacture, in exchange for something that would be difficult or impossible for it to do. Among the most spectacular examples of these sorts of bargains are those that form eukaryote cells, with mitochondria providing energy conversion and, in plants, chloroplasts converting carbon dioxide to sugar.

In a powerful demonstration of how easily mutualism can evolve, William Harcombe took a strain of Escherichia coli that could not synthesize methionine and mixed it in a lactose environment with a strain of Salmonella enterica he had engineered with the capability of producing methionine as a waste product. At the beginning of the experiment neither could grow, because the Salmonella did not produce enough methionine for the E. coli to grow, so it did not produce enough sugar for the Salmonella. But in a structured plate environment, cooperation evolved that benefited both species. Increased methionine production was costly for Salmonella, so it evolved only in a private interaction with E. coli under a fixed-surface environment that ensured that benefiting neighbors of the other species would create additional benefits to self. In a liquid environment, where the benefits were dispersed more globally, cooperation did not evolve.

5. CONFLICT AND CONTROL OF CONFLICT IN FRATERNAL COOPERATIVE SYSTEMS

Fraternal cooperation requires genetic relatedness among participants, but this does not mean that all interactions among kin are cooperative. The problem of conflict for fraternal cooperators may seem small, since they are related. When the parties share genes, actions that excessively reduce the fitness of one individual will reduce that of the actor also. But if relatedness is less than one, or complete clonality, the interests of two parties are not identical, so there can be conflicts. Even within families, there are many conflicts of interest, and how these play out make up some of the richest stories of fraternal cooperation.

How is fraternal conflict controlled? This is a major question in the study of cooperation within families. When relatedness alone does not eliminate conflicts, outcomes are decided in large part by the relative power of the parties. For example, even when helping pays, it is still generally better to receive altruism than to give it, so there can be conflict over who helps and who receives help. In some cases this potential conflict is reduced because of asymmetries. Age is an important asymmetry, as in the case of a parent helping a child. Parents are often in a position to provide aid to offspring, while offspring may be incapable, at some ages, of helping parents. In other cases, power can determine who helps whom, with the stronger party forcing the weaker into the helping role. In Polistes wasps, one foundress usually takes the queen role and lays all the eggs and her sisters help her, taking on the risky foraging tasks. The one taking the queen role is usually the first to begin the nest in the springtime of temperate latitudes. She is typically the largest, which may be why she emerges from hibernation earlier and begins nesting activities sooner.

Once who helps whom is decided, fraternal cooperators may further disagree on the amount of aid given by the altruist. For example, a baby may demand more from its parent than that parent is selected to give if this imposes costs on the parent’s other progeny, because a parent is typically related to all its babies by one-half, while the baby is related to itself by one and to its full siblings by one-half. Thus, the baby will favor more investment in itself and less in its siblings. The famous parent-offspring conflict first described by Robert Trivers ensues.

Parent-offspring conflict does not have to involve how much the parent gives but can also involve sex ratios. In social ants, bees, and wasps, workers are more related to their sisters than to their brothers in colonies with a single once-mated queen because of haplodiploid sex determination. Consequently, workers will favor a more female-biased sex ratio among the brood they rear than the queen will favor. This interesting example is covered in more detail in the eusociality chapter (VII.13). The resolution of these and many other within-family conflicts of interest have provided some of the best tests and supports of kin selection theory.

Whereas some conflicts are resolved through relative individual power, others involve the power of a collective. Francis Ratnieks called the group enforcement of common good policing. He demonstrated that in honey bees, workers suppress other workers from laying eggs, at least in the presence of the queen. Policing has been demonstrated to be important in controlling many kinds of conflict.

6. CONFLICT AND CONTROL OF CONFLICT IN EGALITARIAN COOPERATIVE SYSTEMS

Conflict is potentially even stronger in egalitarian cooperation, in which neither partner has a genetic stake in the other; there can be a persistent evolutionary push and pull between cooperating parties. We begin with the topic of sexual reproduction. Before a new zygote is formed, in most eukaryotes, the diploid genome is divided in half in a normally cooperative meiosis. The fairness of meiosis is egalitarian cooperation, even though it occurs in a single individual, because some genes do not end up in the egg. This fairness of meiosis is sometimes defeated by meiotic drive, the name for the process that causes one allele to always make it into the progeny.

In sexual reproduction, the gametes from male and female fuse in another egalitarian process, at least for nuclear genes, because half come from the father and half from the mother, in most organisms. But mitochondria (and chloroplasts in plants) are usually inherited entirely through the female. Nuclear and mitochondrial genes cooperate in the adaptive function of the eukaryote cell, but the conflicts arising from differences in inheritance are not entirely resolved. For example, mitochondrial genes may cause male sterility in plants, in order to produce more seeds that transmit mitochondria, while nuclear genes act to restore male fertility. Similarly, cytoplasmic parasites, like Wolbachia, can bias sexual reproduction toward females who transmit the Wolbachia. But genes in the same individual are largely cooperative, and live or die with the individual.

After sexual reproduction, most organisms release the progeny into the world to fend for themselves, as seeds or eggs. But in some organisms, parents greatly increase the chance of progeny survival by caring for them (see chapter VII.8), which creates another arena for egalitarian cooperation. Though the parents are both related to the progeny, they are unrelated to each other and so will disagree on how much each should give the young. In some groups, like mammals, one sex has evolved special abilities for caring (milk production in females). In others, either parent can care, which generates a rich area of research into the specifics of such care. Confidence that one is actually the parent is a factor affecting which individual gives more care. This is usually the female, since her confidence of being the mother is greater than the male’s confidence of being the father, at least in organisms with internal fertilization.

There are many other fascinating examples of conflict in egalitarian systems. Toby Kiers and collaborators showed that when soybean rhizobia were prevented from fixing nitrogen by being isolated in a nitrogen-free atmosphere, the plants cut the amount of carbon they allocated to those nodules. Figs have evolved a complex relationship with their pollinating wasps, which enter the fig, lay their eggs, and either actively or passively pollinate the flowers within the fig. In the more basal species with passive pollination, the wasps simply encounter abundant pollen in their natal fig and transport the pollen by chance. In the more derived species, the wasps seek out the pollen-producing flowers in their natal fig, carry the pollen with them, and actively pollinate the flowers in the fig they choose for their eggs. Clearly, the latter form is a tighter mutualism, for the fig is dependent on an act the wasp would not necessarily perform. Jander and Herre found that the actively pollinated species had sanctions against wasps that did not pollinate sufficient flowers: those fruit were simply dropped from the tree and not allowed to ripen, killing the wasps inside.

Control of cheating in egalitarian relationships like those just described is based on how partners are kept to their end of the bargain. These controls take two general forms, called, somewhat confusingly, partner choice and partner fidelity feedback. Under partner choice, underperformers can be punished. The legume-rhizobium and the fig-wasp examples are examples. The plants reject poorly performing bacterial nodules, or wasps that do not provide sufficient benefits.

Under partner fidelity feedback, the fates of the partners can be so completely commingled that sanctions are rare, for they would hurt both partners. The eukaryotic cell is such a case; with rare exceptions any harm that either mitochondrion or host cell does to the other feeds back as harm to itself. Many phloem-feeding insects rely on bacteria to digest their sugary food and to produce essential vitamins. The aphids and their vertically transmitted Buchnera bacteria reproduce through the same pathway and are utterly dependent on each other; neither can do much to gain at the other’s expense. In general, cotransmission makes partner fidelity feedback strong.

7. ORGANISMALITY RESULTS FROM HIGH COOPERATION AND LOW CONFLICT

Fraternal and egalitarian cooperation alike can bring formerly separate entities together into alliances of varying degrees. Most of the alliances that are highly cooperative, with conflict at lower levels thoroughly controlled, are called organisms. This is the level at which adaptations are most common, and these adapted bundles compete with other adapted bundles. In an earlier paper(Queller and Strassmann 2009) we explored the consequences of taking high cooperation and low conflict as the definition of organismality and argued that other definitions of the organism cannot be consistently applied. Under our cooperation-based definition, widely recognized organisms such as whales and sequoia trees retain their organismality. To the organism list we might add the aphid-Buchnera symbiosis, Dictyostelium fruiting bodies, honey bee colonies, anglerfish mates, lichens, and some fig-wasp symbioses, to name a few. As with any other definition of organism, there may be gray areas, but this can be an advantage, because there is much to be learned about pattern and process of cooperative alliances at the borders of organismality. Cooperation is not just an activity engaged in by a few special organisms—it is how all organisms came to be in the first place.