Can amoebae be loyal?
The very question may make you twist up your face in skepticism. Surely this is a case of anthropomorphism at its most absurd. After all, loyalty is the sort of quality we like to think distinguishes us from animals – not to mention microbes. Loyalty demands steadfast dedication in the face of temptation. It requires us to make noble sacrifices, to give up selfish interests for something beyond ourselves – whether it be a nation, a church, a principle, a friend. “I will follow thee/To the last gasp with truth and loyalty,” says the faithful servant Adam to Orlando, the hero of Shakespeare’s As You Like It. How could an amoeba display such nobility of spirit?
Fair enough. But consider this: would you twist up your face if someone said his dog was loyal?
Somehow that doesn’t sound so absurd. People call their dogs loyal all the time, and the word seems to fit. Dogs are faithful to their owners, not trotting away with the first stranger to give them a biscuit. They will even make sacrifices for their owners, as testified by the cases of dogs that rescue their owners from floods and fires. They meet the criteria for loyalty, or at least a nonhuman version of it.
No one would argue that in some ways human loyalty is unique. But as dogs show us, some of the behavior that underlies our concept of loyalty is not limited to our own species. Biologists have discovered examples of this nonhuman loyalty in many species, including amoebae. By studying the evolution of loyalty in these organisms, scientists are gathering clues to the origin of loyalty in our own species.
The amoeba Dictyostelium discoides lives in forest soils, where for the most part it leads the life of a rugged individualist. Each amoeba prowls through the soil, searching for bacteria which it engulfs and digests. After gorging itself sufficiently, a single Dictyostelium divides in two, and the new pair go their separate, bacteria-devouring ways. But if the Dictyostelium in a stamp-size plot of soil should eat their surroundings clean, they send each other alarm signals. They then use the signals to steer toward their neighbors, and as many as a million amoebae converge in a swirling mound. The mound itself begins to act as if it were a single organism. It stretches out into a bullet-shaped slug the size of a sand grain, slithers up toward the surface of the soil, probes specks of dirt, and turns around when it hits a dead end. Its movements are slow – it needs a day to travel an inch – but the deliberateness of the movements eerily evokes an it rather than a they.
After several hours, the Dictyostelium slug goes through another change. The back end catches up with the tip, and the slug turns into a blob. About 20 percent of the cells move to the top of the blob and produce a slender stalk. In order to keep the stalk from flopping over, these cells must produce rigid bundles of cellulose. Unfortunately, this cellulose also tears apart the amoebae that make it. The remaining amoebae in the blob then take advantage of the suicide of their slugmates. They slide up to the top and form a globe. Each amoeba in the globe covers itself in a cellulose coat and becomes a dormant spore. In this form the colony will wait until something – a drop of rainwater, a passing worm, the foot of a bird – picks up the spores and takes them to a bacteria-rich place where they can emerge from their shells and start their lives over.
The individual amoebae forming the stalk make the ultimate sacrifice so that other Dictyostelium may live and perhaps reproduce. These stalk-formers are not marked for death when they are born. When the amoebae mix together and the slug takes shape, the individuals that wind up in the front end of the slug will be the ones that form the stalk. In other words, they get a losing ticket in the Dictyostelium lottery. Aside from their rotten luck, they are indistinguishable from the amoebae that will survive as spores.
It is remarkable that stalk-forming amoebae should remain loyal to their fellow amoebae. Why should they willingly join a group of other amoebae when their loyalty will end in its and their death? Why shouldn’t amoebae just stay away from the group and try to tough it out on their own? Of course, just joining a group is not a guarantee of loyalty. It’s not hard to imagine amoebae finding a way to avoid the lottery of death. Actually, we don’t even have to imagine them: scientists have discovered that some Dictyostelium will cheat their fellow amoebae, thanks to genes that ensure that they will form spores rather than stalks.
The puzzle of loyal amoebae is, at its foundation, a puzzle about evolution. In each generation, the members of a population will vary in all sorts of ways – in their size, in their shape, and in their behavior. Depending on the environment in which the population lives, some of these variations will give certain members an edge when it comes to surviving and reproducing. Genes that make successful variations possible will become more common, while the unsuccessful genes will become less common.
Imagine that a Dictyostelium divides in two, and one of its offspring undergoes a mutation that makes it disloyal. It escapes the stalk lottery, and is guaranteed to become a spore. Over generations, its descendants would become more common because none of them have to die making a stalk. Its cheating gene would become more common in the population as a result. Other individuals might also mutate into cheaters on their own, and their offspring would thrive as well. Meanwhile, genes that promote loyalty would become less common. It might be possible for Dictyostelium to continue organizing slugs and stalks if only a small fraction of amoebae cheated. But in time natural selection could produce so many disloyal amoebae that a slug would fail to produce a stalk, dooming the spores to death. As plausible as this scenario may be, scientists don’t see it happening in the real world. Dictyostelium is thriving happily in forests around the world. Clearly this disastrous burst of disloyalty has not evolved. Why not?
This question has dominated the work of many evolutionary biologists in recent decades. It applies not just to Dictyostelium, but to any species in which an individual’s behavior benefits a group, even if its behavior reduces its own chances of reproductive success. Loyalty is a fascinating enigma. The first major step toward solving this puzzle came in the 1960s. Evolutionary biologists William Hamilton and George Williams recognized that families may encourage loyalty. In most animal species, siblings share about half of their genes. That’s a far higher proportion than two unrelated animals. Being loyal to your family is not such a big sacrifice from an evolutionary point of view, because even if you don’t get to reproduce, your sibling may. And some of your genes will be carried by your nephews and nieces.
“Kin selection,” as the idea is known, helps shed light on many animal societies – particularly those of insects such as bees and ants. These insects have very strange societies, in which, typically, each colony is ruled by a single queen. While the queen produces thousands of offspring, her daughters are sterile and take care of their younger siblings. It turns out that these sterile daughters can pass on their genes more effectively by caring for their own siblings than by trying to lay eggs on their own. But this loyalty only exists while the queen lives. Once she dies, sister honeybees will fight to the death to become the new queen.
Kin selection can’t tell the whole story, though. For one thing, loyalty can be found even among organisms that aren’t closely related to one another. Dictyostelium is a case in point. The amoebae swarming together in a patch of soil may be only distantly related. Scientists suspect that amoebae join forces with strangers because they can form larger slugs. A larger slug can move farther and faster, possibly raising the odds that its spores will be able to reach fertile ground elsewhere. In this case, loyalty can only survive if disloyalty loses its appeal. Dictyostelium offers a couple of examples of how this happens. Once amoebae become destined to develop into stalk cells, they still need to receive signals from neighboring cells to complete their development. You could well imagine that if a mutant amoeba became deaf to these signals it could avoid its fate as a dead stalk cell and become a spore instead.
Recently David Queller and Joan Strassmann of Rice University and their colleagues experimentally created these deaf amoebae by knocking out the gene Dictyostelium needs to receive the development signal. (The gene is known as dimA.) The scientists mixed the dimA mutants with ordinary amoebae that were still able to receive the signal and turn into stalk cells. As they expected, the deaf amoebae did not become stalk cells. Instead, they prepared to become spores.
But when the researchers allowed these colonies to develop completely, they got a surprise. Most of the deaf amoebae failed to get into the ball of spores at the top of the stalk. Queller and Strassmann don’t yet know exactly why deaf amoebae can’t become spores as well as ordinary ones. But what is clear is that dimA must have more than one role. In some cases, it acts as a signal that tells an amoeba to become a stalk cell. But in cells that are destined to become spores, it must also have some essential role in their development. It’s common for genes to play different roles, and Queller and Strassman’s research suggests it may pose a major obstacle to the evolution of disloyalty. The advantages a disloyal amoeba gains by losing one of dimA’s functions are wiped out by its losing another, equally important one.
Disloyalty can also lose its luster if it becomes impossible to conceal from others. In another experiment, Queller and Strassmann discovered that some mutant Dictyostelium cheat if they lose a gene called csA. Normally csA produces a sticky protein on the surface of amoebae. The csA mutants, by contrast, are slippery. When amoebae form a slug, these slippery mutants slide back to the rear, where they will have a good chance of becoming spores rather than stalk cells. The problem for a csA cheater is that this same sticky protein serves as a badge of loyalty. When individual Dictyostelium start moving toward one another in the soil, they recognize their neighbors by their csA badge. This sticky protein allows two Dictyostelium to glue themselves together and continue searching for other amoebae with the same badge. Disloyal amoebae don’t have the csA badge, and so they are shunned. While disloyalty may yield an advantage once you’re in a group, if it prevents you from getting in, you’re out of luck.
Dictyostelium and other microbes have, over the past few years, become lab rats for studying loyalty and other sorts of social behavior. Scientists such as Queller and Strassmann have revealed a complexity to the social lives of these tiny organisms that was, until now, unappreciated. And because their societies can flourish in a petri dish – rather than a rain forest or a deep-sea vent – they can be easily studied. What’s more, the lessons that microbes teach scientists about the evolution of loyalty can shed light on the social lives of more complex organisms, such as porcupines or plovers. Or even people.
Obviously, the differences between humans and amoebae are huge. But both Homo sapiens and Dictyostelium live in groups, and the reproductive success of an individual from either species is tied up with how he, she, or it behaves with other individuals. Evolutionary biologists, anthropologists, and other researchers have been studying the biological roots of human loyalty, and found that the parallels between humans and amoebae are striking.
In the twenty-first century, humans may be loyal to nations, religions, or even brands. But these loyalties have emerged very recently in the evolution of our lineage. For most of the past two million years, research has shown that humans and their hominid ancestors have lived in the same basic social arrangements. Males and females probably formed long-term bonds in order to raise children. They used tools made of stone or wood to get their food – most likely a combination of tubers, fruit, and meat. A hominid did not try to go it alone, however. Instead, hominids probably got food cooperatively and shared it not only with their immediate families, but with people they were not directly related to.
How big were these groups of hominids? Robin Dunbar of the University of Liverpool has pioneered an innovative way to answer that question – by looking at the size and shape of fossil brains. Among primate species, there’s a pretty tight relationship between the size of some parts of the brain and the average group size. Dunbar believes that living in big groups requires a big brain to process the extra social information – who’s your friend, your enemy, who’s on top, who’s jockeying for a takeover, and so on. The size of hominid brain cases from two million years ago suggests that our ancestors were living in groups of about a hundred. The brain size of living humans suggests that the natural group size for our species today is only 150. It turns out that this is about the size of groups living in hunting-foraging societies today. Sociologists have asked people how many acquaintances they would feel comfortable asking for a favor. The answer turns out to be 135. Many similar results suggest that our brains only allow us to get to know about 150 people particularly well.
While people today may be loyal to abstractions such as truth or democracy, the earliest sort of loyalty was probably to these 150 people. Relatives probably inspired the most loyalty. Parents and children developed long-term bonds, based on intense feelings of love and dependence. These feelings – mediated by certain hormones and brain circuits – were essential for the survival of human children. Many animals, including species of fish and frogs, abandon their offspring as soon as they’re born. Most mammals extend the care of their young until they are weaned and then drive them away. But human children need more care than that, in part because our species has such a big brain.
Ounce for ounce, the brain demands twelve times more calories than muscle. And human brains grow at a spectacular rate throughout childhood. As a result, children need a lot of calories every day. Making matters worse, a weaned three-year-old human can’t just wander off to the nearest tree and start munching on leaves. Two million years ago, hominids were already using tools to obtain meat and other sorts of energy-rich food – food that only adults could get. Children had to depend on their parents. This situation probably favored the evolution of loyalty to one’s family. Of course, it also sometimes led to disloyalty – a male, for example, might abandon his family to take another mate.
Still, loyalty seems to have taken root in our species, even across the generations. Biologists have long been intrigued by the fact that human females undergo menopause in their forties, leaving them with decades of life left when they can’t have children. In chimpanzees, our closest living relatives, females are fertile almost until they die of old age. Kin selection may be the answer to this puzzle. Given the demands of rearing hominid children, evolution may have favored females who became infertile but could help raise their grandchildren. On balance, that strategy may have resulted in more of a female’s genes getting passed down through the generations. And studies on contemporary hunter-foragers have shown that grandmothers do indeed work hard to provide food for their grandchildren. What’s more, children in these groups who lose their grandmothers also become on average less healthy.
Our kind of loyalty to family isn’t so unusual among animals. What is unusual is our loyalty to those beyond our family. In hunter-forager societies, people will share hard-earned food with other members of their band, despite the fact that they are not related. This sort of loyalty gets codified into rules that parents teach their children, and that societies expect individual members to follow.
From an evolutionary point of view, this sort of loyalty in humans can be as puzzling as loyalty in amoebae. Say that you’re a hominid who feels a deep and abiding loyalty to your band. You give food to other families with no expectation of something in return. You will even risk your own life to rescue a fellow hunter from an attacking lion. Now imagine that another member of your band feels no compunction about being disloyal to you. He betrays your trust, steals your food, and pretends he doesn’t hear your cries for help. In this scenario, loyalty would seem to be the losing strategy. And if genes influence the degree of loyalty in people, then disloyalty would seem to be destined to spread and loyalty to disappear.
A number of scientists have distilled this conflict down to its basic elements, producing mathematical models of human societies. They draw much of their inspiration for these models from game theory, which turns behavior into a simple game in which players can choose different strategies. In such a model, a hominid might be loyal and cooperative, even if that cooperation reduced his personal gain. On the other hand, another hominid might be disloyal, taking advantage of the trust of others. These models can incorporate many other strategies as well, such as being provisionally loyal and turning disloyal if you encounter disloyalty in someone else. It’s possible to analyze these models mathematically, to see how they change over time. Does loyalty win out in the end, or is it overwhelmed by deception and selfishness?
In the 1970s, researchers proposed that loyalty might be fostered by something they called reciprocal altruism. In other words, people help nonrelatives, but only in the expectation of getting help in return. If someone doesn’t reciprocate, you don’t help them any more. The human brain seems particularly well-adapted to this strategy. We have powerful memories that allow us to keep a running tally of who’s been good to us and who hasn’t.
Unfortunately, models based on reciprocal altruism alone don’t do a good job of replicating loyal societies. Just a few cheaters can destroy widespread trust. And it’s not just mathematical models that show this to be the case. Psychologists have enlisted volunteers to play games that are based on models of reciprocal altruism. People are given a fund of play money, some of which they can put into a common account. The psychologists later divide the money among all the participants, even those who don’t give anything. Some people will act like free-riders, not giving any money of their own but taking from the common account. As soon as that happens, loyal players stop acting selflessly.
What does it take for loyalty to survive? Employing the power of punishment, reputation, and shunning can help. If people are punished for being disloyal, cheating becomes a much less attractive strategy. But it’s not a perfect solution. Even if a person is cooperating with the group, he or she may decide not to put in the effort necessary to punish cheaters, leaving that task to others. If only a few people are willing to punish cheaters, the cheaters win. Researchers have found they can fix this flaw by adding some new rules to the game. When players are deciding how to act toward one another, they can find out how other players have acted in the past. With this information, people tend to help those who contribute to the common good and not those who are disloyal. Loyalty can become even more common if people also shun people who don’t help punish cheaters.
These games, it seems, aren’t mere abstractions. They actually capture something real about how our brains work. Swiss researchers recently ran an experiment to see what happened in people’s brains when they punished others. The researchers had pairs of volunteers play a game together. (We can call the players Alice and Bill.) Alice and Bill each got ten money units. (Let’s call them dollars.) Alice could choose whether to hand over her money to Bill or hold onto it. If she sent the money to Bill, the experimenters quadrupled her ten dollars to forty. Now Bill had fifty dollars and Alice had nothing.
Bill could either send nothing back to Alice or send half of his money. The game had three outcomes. If both Alice and Bill cooperated, they both ended up with twenty-five dollars – over twice what they started with. But if Alice trusted Bill and he cheated her, Bill ended up with five times more, and Alice was broke. And if Alice didn’t trust Bill, they both ended up with the ten dollars that they started with.
If Bill cheated on Alice, the scientists gave her a chance to punish him. She could inflict up to twenty “punishment points” – each point equaling a dollar taken away from Bill’s winnings. But she had to pay a dollar herself for every point. The researchers monitored the brain activity of volunteers playing the role of Alice as they decided whether to punish or not. (They did so with PET scans, which can track radioactive tracers injected into the bloodstream.) The researchers then compared the brain activity in different versions of the game. In one version, for example, Alice could only inflict symbolic punishment points – Bill would be told she was angry with him, but her punishment didn’t affect his winnings. In another version, Alice could take away some of Bill’s money, but didn’t have to give up any of her own.
The researchers found that forcing Bill to pay for his transgression made Alice feel good. The human brain, like that of other mammals, has a circuit of neurons called the reward pathway, which produces pleasant feelings in response to rewards. In animals such as rats and monkeys, the reward pathway becomes active when the animal unexpectedly discovers food. When a human being thinks about a romantic partner or winning at gambling, the reward pathway switches on as well. And the Swiss researchers found that punishing cheaters also causes it to switch on. Significantly, the symbolic punishment didn’t create anything close to the response that the real punishment did. The volunteers apparently were anticipating the satisfaction they’d feel when their punishments were carried out. Just as tellingly, volunteers whose reward pathways responded more strongly were more willing to pay out of their own pocket to make the other players suffer for cheating.
What’s particularly fascinating about these results is how they mirror the behavior of amoebae. Loyalty has evolved in Dictyostelium, it seems, because disloyalty has been made unprofitable. Amoebae can recognize loyalty in others, for example, by recognizing proteins on cell surfaces, and shun cheaters. By contrast, we humans use our brains to identify the intentions of others, and shun the disloyal. But the outcome is apparently the same: cheaters suffer and loyalty is allowed to flourish.
Of course, how loyalty came to be attached to nations, churches, and other modern institutions is another story – one of cultures and history. But it turns out that the preface to that story stretches back millions of years.
Copyright 2005 Carl Zimmer