Chapter 12 Chapter 10 You tickle me, and I will ride on your head

We have studied interactions between survival machines belonging to the same species -- parental, sexual, and aggressive interactions.However, there seem to be some noteworthy aspects of animal interactions that are clearly not covered by the above three categories.The social habit of many animals is one aspect.Birds, insects, fish, whales, and even mammals living on the plains always gather in groups for food.Members of these collectives are usually of the same species, but there are exceptions.Zebras and wildebeests often mix together, and birds of different species can sometimes be seen gathering in flocks.

Living in groups can bring all sorts of benefits to a selfish individual.Here, I am not going to list them one by one, but I am going to talk about a few instructive examples.In it I will also recall some examples of overt altruistic behavior that I gave in Chapter 1, because I said that these examples are to be explained later.This necessarily involves a discussion of social insects; indeed, no discussion of animal altruism can be complete without mentioning social insects.Finally, in the miscellaneous content of this chapter, I will talk about the important concept of mutual altruistic behavior, that is, the principle of "what is convenient for others is convenient for oneself".

Animals must live together because their genes benefit more from the association of living in groups and pay less for it.When hunting in groups, hyenas can catch much larger beasts than when they are alone. Although they have to divide the food after catching wild animals, it still pays off for each selfish individual participating in the group hunting.It is presumably for similar reasons that certain spiders work together to weave a vast communal web.Emperor penguins huddle tightly together for warmth.This is because after being huddled together, the exposed body surface of each penguin is much smaller than when alone.When two fish are swimming in the water, if one swims behind the other while maintaining a certain slope, it can gain a hydrodynamic benefit from the turbulence created by the fish in front.This may be part of the reason fish swim together in groups.

How to use turbulence to reduce air resistance is also a trick familiar to cyclists.Birds may form a V-shape in flight for this reason as well.Since the first bird is at a disadvantage, the birds presumably compete to avoid this role.It is likely that they take turns serving as this involuntary navigator.This is a delayed mutual altruism, the form of altruism we discuss at the end of this chapter. Many of the possible benefits of living in groups have to do with avoiding predators.Hamilton brilliantly developed this theory in a treatise entitled The Geometry of the Selfish Herd.In order not to cause misunderstanding, I want to emphasize that by "herd of selfishness" he means "herd of selfish individuals".

Let's start again with a simple "pattern".Although the pattern is abstract, it can help us understand the real objective world.Imagine a group of animals of a certain species being hunted by a predator.The animal closest to the predator is often the first to be attacked.For predators, this strategy makes sense because it saves energy.But for prey animals, this strategy has an interesting consequence.That is, each of the fleeing animals is trying to avoid being in the closest position to the predator.If the animals spot the predator at a distance, they can simply run away.Even when a predator appears suddenly and silently, like a beast hiding in dense grass, each animal plays by ear and tries to avoid being in the closest position to the predator.We can imagine that there is a "danger zone" around every hunted animal.In this danger zone, the distance from any point to this animal is shorter than the distance from that point to any other animal.For example, if a group of animals being pursued moves in a regular geometric pattern spaced apart from each other, the danger zone for each animal (unless it happens to be on the edge) is roughly six-sided shape.If a predator happens to be lurking in the hexagonal danger zone of individual A, individual A may be eaten.Individuals on the fringes of the herd are especially vulnerable, because their danger zone is not a relatively small hexagon, but has an open end, and a wide area beyond the open end is their danger zone.

A sane individual will obviously try to minimize his danger zone.It especially tries to avoid being on the fringes of the herd.If it finds itself on the fringes, it takes immediate action and moves to the center. Unfortunately, there must always be a "person" on the edge, but as far as each individual is concerned, this "person" had better not be it!Thus, as a group of animals advances, individuals on the periphery keep moving toward the center.If the group of animals were originally loose or scattered, the result of this movement toward the center of the group would quickly cause them to huddle together.Even if the pattern we speak of begins without any tendency to gather together, and the hunted animals initially scatter randomly, selfish motives will prompt each individual to try to squeeze itself among the others in order to narrow their respective danger zones. .In this way, clusters form quickly and become denser and denser.

In real life, this tendency to gather is apparently limited by various resistances, otherwise the animals would inevitably scramble and become exhausted.Still, the pattern is interesting because it illustrates how even some extremely simple hypotheses can lead to the conclusion that animals tend to clump together.Some more complex patterns were proposed.Although these models have greater practical significance, the simpler models proposed by Hamilton are not detracted from them.The latter helps us study the phenomenon of animals clumping together. The selfish herd model itself does not allow for cooperative interactions.There is no altruism here, just each individual exploiting every other individual for self-interest.But in real life it often happens that individuals seem to be actively trying to protect their fellow group members from predators.

Speaking of this, I can't help but think of the alarm calls of birds.This kind of alarm call made other individuals flee for their lives, which indeed served as a warning.No one suggested that the alarming individual was "wanting to direct predator fire" on itself.It merely lets its partners know that a predator is present -- that is, alerts them.But at first glance, the behavior itself seems altruistic, since its effect is to draw the predator's attention to the caller.We can draw an indirect inference from a fact discovered by PR Marler.This alarm call in birds seems to have some desirable physical properties: It is often difficult for predators to detect where the call is coming from.Ask an acoustic engineer to design a sound that would be difficult for a predator to track down, and it would likely resemble the natural alarm call of many small singing birds.In nature, the formation of this alarm call must be the result of natural selection.We know what that means.This means that many individuals lost their lives because their alarm calls were not perfected.So there always seemed to be danger in sounding the alarm.The theory of the selfish gene would have to demonstrate that there is a convincing virtue to sounding the alarm that outweighs the attendant danger.

Actually it's not very difficult.In the past, it has been repeatedly pointed out that the alarm calls of birds are actually "incompatible" with Darwin's theory.The result is that in order to explain this phenomenon, it has become a kind of game for people to invent various reasons.So we face so many plausible explanations today that we are at a loss for what to say.Clearly, if some individuals in the flock are close relatives, the gene for the alarm call must thrive in the gene pool, since there is a high probability that some of the individuals saved will have it.Even if the individual who calls the alarm pays a high price for the altruistic behavior by attracting predators, it's worth it.

If you find this notion of kin selection unconvincing, there are other theories to choose from.Trivers offers five insightful ideas for the various avenues in which an individual who calls the police on his partner can gain self-interest.But here are two of my own, which I think are more convincing. I call the first idea the Cave theory.Cavey is originally Latin, meaning "beware".Today, elementary school students use this signal to warn other students when they see the teacher approaching.The theory applies to birds that employ a camouflage tactic, crouching motionless in undergrowth when faced with danger.Suppose there is a flock of these birds foraging in a field.At this time an eagle flew by in the distance.The eagle hadn't caught a glimpse of the flock yet, so it didn't fly straight over.But its sharp eyes may spot the flock at any moment, when it will swoop down and strike.If a small bird in the flock finds the eagle first, the rest of the birds have not found it yet.The quick-eyed little bird could have crouched down immediately and hid in the grass.But it did him no good by doing so, for his companions were still moving about, conspicuous and noisy.Any one of them might attract the eagle's attention, and the whole flock would be in danger.From purely selfish motives, the bird that spots the hawk should immediately hiss its mates to silence them at once, to reduce the chances of them inadvertently drawing the hawk into its vicinity.This is the best strategy for this little bird.

Another idea I'm going to talk about can be called the "never leave the team" theory.This theory applies to certain species of birds that fly away when they see a predator approaching, perhaps into a tree.Let us also imagine that one of a flock of foraging birds is the first to spot the predator.How should it act?It can fly away by itself without warning its mates.If so, it's about to become a loner, a member of a less conspicuous flock of birds.It is a well-known fact that hawks love to attack stray pigeons.Even if hawks do not have such hunting habits, there are many reasons that can be reasoned out that breaking away from the group might be a suicidal tactic.Even if its partner eventually flies away with it, the first individual to fly off the ground inevitably temporarily expands its own danger zone.Whether Hamilton's theory is correct or not, there are always some important advantages to living in groups of birds, otherwise the birds would not live in groups.Whatever these advantages may be, they are at least partially lost by the first bird to fly out of the flock.So what if the disciplined bird doesn't go AWOL?Perhaps it should rely on the cover that collective power can provide and continue its activities as if nothing had happened.But the risk is too great after all.Unobstructed and vulnerable to attack.It's much safer in the trees after all.Flying into the trees is indeed a good strategy, but make sure your partners act in unison.Only in this way will it not become a lonely bird separated from the flock, so that it will not lose the advantages provided by the collective, and at the same time it can get the benefits of flying to the trees and hiding.We see here again that what is gained by sounding the alarm is purely selfish gain.A somewhat similar theory has been proposed by EL Charnov and Krebs, who straightforwardly use the word "manipulation" to describe the influence the singing bird exerts on other birds.This behavior is far from pure, selfless altruism. On the surface, the above theories seem to contradict the statement that individuals who sound the alarm put themselves in danger.In fact there is no contradiction in it.If it does not call the police, it will put itself in greater danger.Some individuals have been killed by the alarm calls, especially those that are easily exposed to the source of the sound.Several other instances died due to failure to call the police.Why do birds make alarm calls when they are in danger?Many explanations have been proposed, Cave theory and the "never leave the team" theory are just two of them. How about the leaping Thomson's gazelle?I mentioned this phenomenon in the first chapter.This apparent altruistic suicide by the gazelle moved Ardley to assert that it could only be explained by group selection theory.This project poses a more serious challenge to the theory of the selfish gene.Birds' alarm calls are effective, but they are always careful not to signal their intentions when they signal.Not so with the gazelle's jump.They are posturing to the point of being annoying.It seems that the gazelle is sincerely trying to attract the predator's attention, sometimes almost as if it is teasing the predator.This phenomenon leads to a theory that is both intriguing and quite bold.N. Smythe first proposed the outline of this theory, but Zahavi was undoubtedly the one who developed it logically. We can formulate Zahavi's theory in this way.The key point of this theory is that the jumping behavior of gazelles is far from a signal for other gazelles to see, but for predators to see.Of course, other gazelles saw the jump, and it affected their behavior, but it was a by-product.Because the gazelle's jumping behavior was selected primarily as a signal to predators.This signal basically reads: "Look! How high I can jump! I'm obviously a strong gazelle, you can't catch me. You better be smart and catch my mates! They don't jump as fast as I do." Tall." In less anthropomorphic terms, the genes that make an individual jump high and stand out are less likely to be eaten by predators, which tend to select animals that appear to be easy prey.Many mammalian predators are particularly fond of hunting down old and frail animals. An individual animal that leaps vigorously displays its youthful vigor in a boastful way.According to this theory, such boasting is not altruistic at all.We can only say that this behavior is selfish, because its purpose is to tell the predator that it should go after other animals.In a sense, it's like a high jump competition to see who can jump the highest, and the loser is the predator's chosen target. Another example I have said to explore further is the suicide of bees.It almost certainly dies when it stings a honey marauder.Bees are nothing but a highly social insect.Others are wasps, ants and termites.The object of my discussion is social insects in general, not just the death squads of bees. The track record of social insects is well known, especially for their astonishingly close coordination and apparent altruism.Their suicidal stinging mission embodies a miracle of self-restraint.In the colony of the "honeypot" ants, there is a class of worker ants that do no other work and hang from the top of the nest all day, motionless.Their bellies protrude and are astonishingly large, like light bulbs, filled with food.Other worker ants use them as food banks.To us humans, such worker ants no longer exist as individuals; their individuality is clearly suppressed for the collective good.The social life of ants, bees or termites embodies a higher level of individuality.The food is distributed according to such strict standards that we can even say that they share a collective stomach.They communicate with each other through chemical signals or, in the case of bees, the well-known "dance".These means are so effective that the whole collective acts as if it were a unit, with its own nervous system and sense organs.They seem to be able to recognize and expel foreign invaders through the selectivity produced by the body's immune response system.Although bees are not "hot-blooded" animals, the rather high temperature inside the hive is regulated almost as precisely as the human body.Last but not least, this analogy can be extended to reproduction.In colonies of social insects, most individuals are sterile workers. The "germ line"—the continuous line of immortal genes—runs through a small number of individuals, the reproductive ones.They are similar to the germ cells in our testes and ovaries. The sterile workers are similar to our liver, muscle and nerve cells. Their suicidal behavior and other forms of altruistic or cooperative behavior are less surprising once we accept the fact that workers are sterile.The reason why the body of a normal animal is manipulated is to produce offspring and to raise other individuals with the same genes to ensure the survival of its genes.Suicide for the benefit of other individuals is not compatible with producing one's own offspring in the future.Thus, suicidal self-sacrificing behaviors rarely evolve.But worker bees never produce offspring of their own.All their energies are devoted to caring for relatives who are not their own offspring so as to preserve their genes.The death of a sterile worker bee affects its own genes as a leaf falling from a tree in autumn does to the genes of a tree. Talking about social insects makes one tempted to make mysteries, but in fact there is no need for it.But it's worth looking at how the theory of the selfish gene applies to social insects, especially how it can be used to explain the evolutionary origins of the remarkable phenomenon of worker sterility.Because this phenomenon seems to cause a series of problems. A colony of social insects is a large family, all members of which are usually born to one mother.Workers seldom or never reproduce, and are generally divided into several distinct classes, including small workers, large workers, soldiers, and some highly specialized classes such as "honeypot" ants.Fertile females are called queens, and fertile males are sometimes called drones or kings.In some of the higher societies, the reproductive females do nothing else, but they do an excellent job at producing offspring.They rely on workers to provide them with food and protection, and workers are also responsible for caring for the larvae.In some species of mother ants or termites, the queen is literally a gigantic egg-laying factory, her body hundreds of times larger than that of an ordinary worker, barely moving, and hardly resembling an insect in appearance.The queen is often cared for by the workers, who provide for the queen's daily needs, including providing food and transporting the queen's eggs to the collective nursery.If such an unusually large queen needs to leave the inner room, she has to ride on the backs of several groups of worker ants, and let them carry it out with dignity. In Chapter 7, I talked about the difference between procreation and parenting.I said that, in general, strategies that combine procreation and rearing can evolve.In Chapter 5, we saw that mixed, evolutionarily stable strategies can be divided into two types: either every individual in the population adopts a mixed strategy, so that individuals tend to combine reproduction and rearing judiciously, or the population Divide into two different types of individuals, which is what we originally imagined as a balanced situation between hawks and doves.It makes theoretical sense to achieve an evolutionarily stable balance between reproduction and rearing in the latter way.That is to say, the population can be divided into two parts: breeders and caregivers.But this state of evolutionary stability can only be maintained under the condition that the careee must be a close relative of the caregiver at least as closely as the caregiver's own offspring, if it had any. .Although in theory evolution could proceed in this direction, in practice it seems to be seen only in social insects. Individuals of social insects fall into two main groups: breeders and nurturers.Bearers are fertile males and females.The rearers are workers—sterile males and females in termites, sterile females in other social insects. These two groups of insects do not interfere with each other, so they can perform their tasks more efficiently.But the so-called effective here refers to who is effective? "What do workers actually gain from this?" This familiar question poses a challenge to Darwin's theory. Someone replied: "Nothing good".They believe that the queen is supreme, bossy and domineering, manipulating the workers through chemical processes to satisfy their selfish desires, and driving them to raise their many children.We saw Alexander's "parental manipulation" theory in Chapter 8, and what I said above is actually another formulation of this theory. A counter argument to this is that workers "cultivate" a fertile mother, driving the mother to increase her fecundity in order to replicate the worker's genes.The survival machines made by the queen are definitely not descendants of worker insects, but they are all close relatives of worker insects.Hamilton had a unique insight that, at least in colonies of mother ants, bees, and wasps, workers may in fact be more closely related to larvae than queens are to larvae!Hamilton, and later Trivers and Hale, proceeded with this view as a guide and finally achieved one of the most brilliant achievements in the theory of the selfish gene.Their reasoning goes like this. A group of insects called the Hymenoptera, which includes ants, bees and wasps, has a rather peculiar system of sex determination.Termites do not belong to this group and therefore do not have this characteristic.In a typical Hymenoptera nest there is only one mature queen.It flies out to mate once when it is young, and stores the sperm inside for ready access for the rest of its long life—a decade or more.It distributes sperm to its own eggs year after year, fertilizing the eggs as they pass through the fallopian tubes.But not all eggs are fertilized.Unfertilized eggs become males.Therefore, the male has no father, and each cell in his body has only one set of chromosomes (all from the mother) instead of two sets of chromosomes (one from the father and one from the mother) like in our body.Following the analogy in Chapter 3, a male Hymenoptera has only one copy of each "volume" in each of its cells, instead of the usual two. The Hymenoptera female, on the other hand, is normal because she has a father and has the usual two sets of chromosomes in each of her body cells.Whether a female becomes a worker or a queen depends not on her genes, but on how she grows.In other words, each female has a complete set of genes for becoming a queen and a complete set of genes for becoming a worker (or in other words, there are also several sets of genes that make it a worker insect, soldier insect, etc. of various professional levels). genes).Which set of genes plays a decisive role depends on its lifestyle, especially on the food it eats. Although the actual situation is very complicated, the basic situation is roughly the same.We don't know how this bizarre sexual reproductive system could have evolved.There is no doubt that this evolutionary phenomenon must have a cause.But we can only treat it tentatively as an incomprehensible phenomenon of the Hymenoptera, whatever the original reason, which disturbs the calculation of relatedness indices which we mentioned in Chapter 6. Set of simple methods.This shows that the sperm of male insects are not different from each other like our human sperm, but are exactly the same.Males have only one set of genes per somatic cell, not two.So each sperm has to receive the whole set of genes, not just a part -- 50 percent, so all the sperm in a particular male are identical.Now let us calculate the kinship index between the mother and offspring of this insect. If it is known that a male has gene A in his body, what is the probability that his mother also has this gene?The answer must be 100%, because the male has no father and all its genes come from its mother. Now suppose a female is known to have gene B in her body. There is a 50 per cent chance that her son will also have this gene, since he has only received half of his mother's genes.This statement sounds like a contradiction, but in fact there is no contradiction.Males get all their genes from their mothers, who pass on only half of their genes to their sons.The answer to this paradox is that males have only half the usual number of genes.So is the "true" kinship index between them 1/2 or 1?I don't think it's necessary to worry about this problem.Indices are nothing but units of measurement that people conceive to solve problems.If its application in a particular case presents us with difficulties, we simply abandon it and reapply to the fundamental principles.From the point of view of gene A in the female, the probability that her son also has this gene is 1/2.as many as its daughters.From the female's point of view, therefore, she is as closely related to her offspring as our human offspring are to their mother. But when we talk about sisters, things get complicated.Not only do siblings come from the same father, but the two sperm that impregnated their mothers are identical in every gene.Therefore, sisters are the same as identical twin sisters in terms of genes from the father.If there is gene A in a female, the gene must have come from either the father or the mother.If the gene comes from the mother, there is a fifty percent chance that its sister also has the gene.But if the gene comes from the father, the chances of its sister also having the gene are one hundred percent.Therefore, the kinship index between siblings of Hymenoptera insects is not 1/2 (normal sexual reproduction animals are 1/2), but 3/4. For this reason the Hymenoptera female is more closely related to her fellow sisters than she is to her own offspring.Hamilton saw this, even though he wasn't quite so forthright when he said it.He thought it quite possible that this peculiarly close kinship prompted the female to use her mother as an efficient machine for producing her sisters.This gene, which produces sisters for females, makes copies of itself more rapidly than genes that produce their own children directly.The sterility of the worker insects is thus formed.The true gregariousness of the Hymenoptera, and the consequent sterility of the workers, seems to have evolved independently more than eleven times, while in the rest of the animal kingdom it has evolved only once in the termites.Come to think of it, this is no accident. However, there is something odd here.If the workers are to succeed in using their mother as a machine for producing sisters, they must counteract the mother's natural tendency to produce an equal number of younger brothers for her.From the worker's point of view, the chance that any of its siblings has one of its genes is only 1/4.So it wouldn't necessarily be to the workers' advantage if females were able to produce the same number of fertile offspring, since then they wouldn't be able to reproduce their precious genes to the maximum. Trivers and Hare argue that workers must try to influence the sex ratio in favor of females.They applied Fisher's method of calculating optimality ratios (which we discussed in the previous chapter) to the special case of Hymenoptera and recalculated.The results showed that, for the mother, the optimal investment ratio was 1:1 as usual, but for the sisters, the optimal ratio was 3:1, favoring sisters over brothers.If you are a Hymenoptera female, the most efficient way for you to reproduce your own genes is not to reproduce yourself, but to have your mother produce reproductively viable sisters and brothers for you in a 3:1 ratio.But if you must reproduce yourself, it's in your genes' best interest to have an equal number of fertile sons and daughters. We have seen above that the difference between queens and workers is not genetic.As far as the genes of a female embryo are concerned, it can be both a worker and a queen, with the former "hoping" for a sex ratio of 3:1 and the latter "hoping" for a sex ratio of 1:1. What exactly does "hope" mean?It means that the genes in the queen reproduce themselves best if they have an equal proportion of fertile sons and daughters. But the same gene in a worker reproduces itself best if it influences the worker's mother to have more daughters.Note that there is no contradiction in this statement.For the gene must make the most of all the forces at its disposal.If the gene can affect the growth process of an individual that is sure to become a queen in the future, its best strategy to use this control power is a situation; and if it can affect the growth process of a worker individual, it uses that The best strategy for power is a different story. This means that how to use this reproductive machine has caused a conflict of interest for both parties.The queen "strives" to produce an equal proportion of males and females.The workers work to influence the sex ratios of these fertile offspring to a ratio of three females to one male.If our assumption about workers using the queen as a reproductive machine is correct, workers should be able to achieve a 3:1 male-to-female ratio.Otherwise, if the queen really has all the power and the workers are nothing more than the queen's slaves and compliant royal nursery "nanny" then we should be looking at a 1:1 ratio because this is the queen" Would love to "achieve a ratio.In such a peculiar struggle between generations, which side can win?This problem can be proved by experiments.Trivers and Hare have done this experiment with a large number of ant species. The sex ratio of interest is the ratio of reproductive males to females.They are large, winged mother ants.At regular intervals, they fly out of the anthills in groups to mate.Afterwards, the young queen may have to form another colony.To estimate sex ratios it is necessary to count these winged individuals.Be aware that in many species the reproductive males and females vary in size.This situation compounded the problem.As we have seen in the previous chapter, Fischer's method of calculating optimality ratios can only be applied strictly to the amount of investment in males and females, but not to the calculation of males and females. number.Trivers and Hale took this situation into account, so they weighed the ants during the experiment.They used twenty different ant species and calculated sex ratios based on the investment in fertile males and females.They found that the male-to-female ratio was convincingly close to a 3:1 ratio, confirming the theory that workers actually manipulate everything for their own benefit. In this way, among the several types of ants used as research objects, worker ants seem to "win" in this conflict of interests. This situation is not surprising, because the worker individual, as the guardian of the larvae, naturally enjoys more real power than the queen individual.Genes that try to manipulate the swarm through individual queens are no match for genes that manipulate the swarm through individual workers.Interestingly, under what special circumstances can the queen enjoy greater real power than the workers?Trivers and Hale found that it was possible to rigorously test the theory in a special case. We know that certain species of ants keep slaves.The worker ants of these slave-serving species either don't do any day-to-day work, or if they do, they're clumsy.They are good at hunting around for slaves. This kind of situation where the two armies confront each other and kill each other is only seen in humans and social insects.Among many ant species there is a special class called soldier ants.They have particularly hard and well-developed upper and lower jaws, which are sharp weapons for fighting.They attack other ant colonies exclusively for the benefit of their own group.Such raids aimed at capturing slaves were but a special form of their war effort.They attack the nest of another species, attempting to kill the defending worker or soldier ants of the other species, and eventually take away the unhatched larvae of the other species, which hatch in the predator's nest.They do not "know" that they have become slaves.They start to work according to the inherent neural program, and perform their duties exactly as if they were in their own caves.These slaves stay in the ant nest and take care of various daily tasks such as managing the ant nest, cleaning, collecting food, and caring for the larvae, while the worker ants or soldier ants who specialize in catching slaves continue to go out to capture more slaves. It is a good thing that these slaves are of course unaware that they are not related at all to the Queen and the larvae they tend.They unknowingly raise batch after batch of newly captured slave soldier ants.自然选择在影响奴隶物种的基因时，无疑有利于各种反奴隶制度的适应能力。不过，这些适应能力显然并不是十分有效的，因为奴隶制度是一种普遍现象。 从我们目前论题的观点来看，奴隶制度产生了一种有趣的后果。在捕捉奴隶的物种中，女王现在可以使性比率朝它"喜欢"的方向发展。这是因为它自己所生的子女，即那些专门捕捉奴隶的蚂蚁不再享有管理托儿所的实权。这种实权现在操在奴隶手中。这些奴隶"以为"它们在照顾自己的骨肉兄弟或姐妹。它们所做的大抵无异于它们本来在自己穴里也同样要做的一切，以实现它们希望达到的有利于姐妹的3：1比例。但专门掳掠奴隶的物种的女王能够采取种种反措施，成功地扭转这种趋势。对奴隶起作用的自然选择不能抵消这些反措施，因为这些奴隶同幼虫并无亲缘关系。 让我们举个例子来说明这种情况。假定在任何一个蚂蚁物种中，女王"试图"把雄性卵子加以伪装，使其闻起来象雌性的卵子。在正常情况下，自然选择对职蚁"识破"这种伪装的任何倾向都是有利的。我们可以设想一场进化上的斗争的情景，女王为实现其目的不断"改变其密码"，而职蚁不断进行"破译"。在这场斗争中，惟能通过有生殖能力的个体把自己的基因传递到后代体内的数量越多，谁就能取胜。我们在上面已经看到，在正常情况下，职蚁总是取胜的一方。但在一个豢养奴隶的物种中，女王可以改变其密码，而奴隶职蚁却不能发展破译的任何能力。这是因为在奴隶职蚁体内的任何一个"有破译能力"的基因并不存在于任何有生殖能力的个体体内，因此不能遗传下去。有生殖能力的个体全都是属于豢养奴隶的物种，它们同女王而不是同奴隶有亲缘关系。即使奴隶的基因有可能进入任何有生殖能力的个体体内，这些个体也是来自那些被掳掠的奴隶的老家。 因此，这些奴隶最多只能忙于对另一套密码进行破译！由于这个缘故，在一个豢养奴隶的物种中，女王因为可以随心所欲地变更其密码而稳操左券，绝对没有让任何有破译能力的基因进入下一代的风险。 从上面这段比较复杂的论证得出的结论是，我们应该估计到在豢养奴隶的物种中，繁殖有生殖能力的雌虫和雄虫的比率是1：1而不是3：1。只有在这种特殊情况下女王能够如愿以偿。这就是特里弗斯和黑尔得出的结论，尽管他们仅仅观察过两个豢养奴隶的物种。 我必须强调指出，我在上面是按照理想的方式进行叙述的。实际生活并非如此简单。譬如说，最为人所熟知的群居昆虫物种--蜜蜂--似乎是完全违反"常情"的。雄蜂的数量大大超过雌蜂，无论从职蜂或从蜂后的观点来看，这种现象都难以解释。汉密尔顿为了揭开这个谜，他提出了一个可能的答案。他指出，当一只女王飞离蜂房时，它总要带走一大群随从的职蜂，它们帮这只女王建立一个新的群体。这些职蜂从此不再返回老家，因此抚养这些职蜂的代价应该算是繁殖成本的一部分。这就是说，从蜂房每飞走一只女王就必须培育许多额外的职蜂来补缺。对这些额外职蜂所进行的投资应算作对有生殖能力的雌蜂的投资额的部分。在计算性比率的时候，这些额外的职蜂也应在天平上称分量，以求出雌蜂对雄蜂的比例。如果我们这样理解问题的话，这个理论毕竟还是站得住脚的。 这个精巧的理论还有另外一个更加棘手的问题需要解决。在一些物种中，年轻的女王飞出去交配时，与之交配的雄蜂可能不止一只。这意味着女王所生育的女儿之间的亲缘关系平均指数小于3/4，在一些极端的例子里，甚至可能接近1/4。有人把这种现象解释为女王借以打击职蜂的一种巧妙的手段！不过这种看法似乎不合逻辑。附带说一句，这似乎意味着女王飞出去交配时，职蜂应伴随在侧，只让女王交配一次。但这样做对这些职蜂本身的基因并没有任何好处--只有对下一代职蜂的基因有好处。每一只职蜂所"念念不忘"的是它自身的基因。有些职蜂本来是"愿意"伴随其母亲的，但它们没有这样的机会，因为它们当时还没有出生。一只飞出去交配的年轻女王是这一代职蜂的姐妹，不是它们的母亲。因此，这一代职蜂是站在女王这一边而不是站在下一代职蜂那一边的。下一代的职蜂是她们的侄女辈。好了，说到这里，我开始感到有点晕头转向。是结束这个话题的时候了。 我在描述膜翅目职虫对其母亲的行为时使用了"耕耘"的比喻。这块田地就是基因田。职虫利用它们的母亲来生产它们自身的基因的拷贝，因为这样比职虫自己从事这项工作更富有成效。源源不断的基因从这条生产流水线上生产出来，包装这些基因的就是称为有生殖能力的个体。这个"耕耘"的比喻不应与群居昆虫的另外一种可以称为耕耘的行为混为一谈。群居昆虫早就发现，在固定的地方耕种粮食作物比狩猎或搜集粮食有效得多。而人类在很久之后才发现这个真理。 譬如说，在美洲有好几个蚂蚁物种以及与这些物种完全无关的非洲白蚁都培植"菌类植物园"。最有名的是南美洲的"阳伞蚁"（parasol ants)。这种蚁的繁殖能力特别强。有人发现有的群体其个体竟超过两百万个之多。它们筑穴于地下，复杂的甬道和迥廊四通八达，深达十英尺以上，挖出的泥土多达四十吨。地下室内设有菌类种植园地。这种蚂蚁有意识地播种一种特殊品种的菌类。它们把树叶嚼碎，作为特殊的混合肥料进行施肥。这样，它们的职蚁不必直接搜寻粮食，只要搜集制肥用的树叶就行了。这种群体的阳伞蚁吃树叶的胃口大得惊人。 这样它们就成为一种主要的经济作物害虫。但树叶不是它们的食粮，而是它们的菌类的食粮。菌类成熟后它们收获食用，并用以饲养幼虫。菌类比蚂蚁的胃更能有效地消化吸收树叶里的物质。因此蚂蚁就是通过这样的过程而受益。菌类虽然被吃掉，但它们本身可能也得到好处，因为蚂蚁促使它们增殖，比它们自己的孢子分散机制更有效。而这些蚂蚁也为植物园"除草"，悉心照料，不让其他品种的菌类混迹其间。由于没有其他菌类与之竞争，蚂蚁自己培植的菌类得以繁殖。 我们可以说，在蚂蚁和菌类之间存在某种利他行为的相互关系。值得注意的是，在与这些蚂蚁完全无关的一些白蚁物种中，独立地形成了一种非常相似的培植菌类的制度。 蚂蚁有其自己的家畜和自己的农作物。蚜虫--绿蚜虫和类似的昆虫--善于吮吸植物中的汁液。它们非常灵巧地把叶脉中的汁液吮吸干净，但消化这种汁液的效率却远没有吸吮这种汁液的效率高，因此它们排泄出仍含有部分营养价值的液体。一滴一滴含糖丰富的"蜜汁"从蚜虫的后部分泌出来，速度非常之快，有时每个虫在一小时内就能分泌出超过其自身体重的蜜汁。在一般情况下，蜜汁象雨点一样洒落在地面上，简直和《旧约全书》里提到的天赐"灵粮"一样。但有好几个物种的蚂蚁会等在那里，准备截获蚜虫排出的食粮。有些蚂蚁会用触角或腿抚摩蚜虫的臀部来"挤奶"。蚜虫也作出积极的反应，有时故意不排出汁液，等到蚂蚁抚摩时才让汁液滴下。如果那只蚂蚁还没有准备好接受它的话，有时甚至把一滴汁液缩回体内。有人认为，一些蚜虫为了更好地吸引蚂蚁，其臀部经过演化已取得与蚂蚁脸部相象的外形，抚摩起来的感觉也和抚摩蚂蚁的脸部一样。蚜虫从这种关系中得到的好处显然是，保证安全，不受其天然敌人的攻击。象我们牧场里的乳牛一样，它们过着一种受到庇护的生活。由于蚜虫经常受到蚁群的照料。它已丧失其正常的自卫手段。有的蚂蚁把蚜虫的卵子带回地下蚁穴，妥为照顾，并饲养蚜虫的幼虫。最后，幼虫长大后又轻轻地把它们送到地面上受到蚁群保护的放牧场地。 不同物种成员之间的互利关系叫做共生现象。不同物种的成员往往能相互提供许多帮助，因为它们可以利用各自不同的"技能"为合作关系作出贡献。这种基本上的不对称性能够导致相互合作的进化上的稳定策略。蚜虫天生一副适宜于吮吸植物汁液的口器结构，但这种口器结构不利于自卫。蚂蚁不善于吮吸植物的汁液，但它们却善于战斗。照料和庇护蚜虫的蚂蚁基因在基因库中一贯处于有利地位。在蚜虫的基因库中，促进蚜虫与蚂蚁合作的基因也一贯处于有利地位。 互利的共生关系在动植物界中是一种普遍现象。地衣在表面上看起来同任何其他的植物个体一样。而事实上它却是在菌类和绿海藻之间的，而且相互关系密切的共生体。两者相依为命，弃他就不能生存。要是它们之间的共生关系再稍微密切那么一点的话，我们就不能再说地衣是由两种有机体组成的了。也许世界上存在一些我们还没有辨认出来的，由两个或多个有机体组成的共生体。说不定我们自己就是吧！我们体内的每个细胞里有许多称为线粒体的微粒。这些线粒体是化学工厂，负责提供我们所需的大部分能量。如果没有了线粒体，要不了几秒钟我们就要死亡。 最近有人提出这样的观点，认为线粒体原来是共生微生物，在进化的早期同我们这种类型的细胞就结合在一起。对我们体内细胞中的其他一些微粒，有人也提出了类似的看法。对诸如此类的革命性论点人们需要有一段认识的过程，但现在已到了认真考虑这种论点的时候了。我估计我们终将接受这样一个更加激进的论点：我们的每一个基因都是一个共生单位。我们自己就是庞大的共生基因的群体。当然现在还谈不上证实这种论点的"证据"，但正如我在上面几章中已试图说明的那样，我们对有性物种中基因如何活动的看法，本身其实就支持了这种论点。这个论点的另一个说法是：病毒可能就是脱离了象我们这种"群体"的基因。病毒纯由DNA（或与之相似的自我复制分子）所组成，外面裹着一层蛋白质。它们都是寄生的。这种说法认为，病毒是由逃离群体的"叛逆"基因演化而来，它们在今天通过空气直接从一个个体转到另一个个体，而不是借助于更寻常的载运工具--精子和卵子。假设这种论点是正确的，我们完全可以把自己看成是病毒的群体！有些病毒是共生的，它们相互合作，通过精子和卵子从一个个体转到另一个个体。这些都是普通的"基因"。其他一些是寄生的，它们通过一切可能的途径从一个个体转到另一个个体。如果寄生的DNA通过精子和卵子转到另一个个体，它也许就是我在第三章里提到的那种属于"佯谬"性质的多余的DNA。如果寄生的DNA通过空气或其他直接途径转到另一个个体，它就是我们通常所说的"病毒"。 但这些都是我们在以后要思考的问题。目前我们正在探讨的问题是发生在更高一级关系上的共生现象，即多细胞有机体之间的而不是它们内部的共生现象。共生现象这个字眼按照传统用法是指属不同物种的个体之间的联系关系（associations）。不过，我们既然已经避开了"物种利益"的进化观点，我们就没有理由认为属不同物种的个体之间的联系和属同一物种的个体之间的联系有什么不同。一般地说，如果各方从联系关系中获得的东西比付出的东西多，这种互利的联系关系是能够进化的。不管我们说的是同一群鬣狗中的个体，或者是完全不同的生物如蚂蚁和蚜虫，或者是蜜蜂和花朵，这一原则都普遍适用。事实上，要把确实是双向的互利关系和纯粹是单方面的利用区别开来可能是困难的。 如果联系的双方，如结合成地衣的两方，在提供有利于对方的东西的同时接受对方提供的有利于自身的东西，那我们对于这种互利的联系关系的进化在理论上就很容易想象了。但如果一方施惠于对方之后，对方却迟迟不报答，那就要发生问题。这是因为对方在接受恩惠之后可能会变卦，到时拒不报答。这个问题的解决办法是耐人寻味的，值得我们详细探讨。我认为，用一个假设的例子来说明问题是最好的办法。 假设有一种非常令人厌恶的蜱寄生在某一物种的小鸟身上，而这种蜱又带有某种危险的病菌。必须尽早消灭这些蜱。一般说来，小鸟用嘴梳理自己的羽毛时能够把蜱剔除掉。可是有一个鸟嘴达不到的地方--它的头顶。对我们人类来说这个问题很容易解决。一个个体可能接触不到自己的头顶，但请朋友代劳一下是毫不费事的。如果这个朋友以后也受到寄生虫的折磨，这时他就可以以德报德。事实上，在鸟类和哺乳类动物中，相互梳理整饰羽毛的行为是十分普遍的。 这种情况立刻产生一种直观的意义。个体之间作出相互方便的安排是一种明智的办法。任何具有自觉预见能力的人都能看到这一点。但我们已经学会，要对那些凭直觉看起来是明智的现象保持警觉。基因没有预见能力。对于相互帮助行为，或"相互利他行为"中，做好事与报答之间相隔一段时间这种现象，自私基因的理论能够解释吗？威廉斯在他1966年出版的书中扼要地讨论过这个问题，我在前面已经提到。他得出的结论和达尔文的一样，即延迟的相互利他行为在其个体能够相互识别并记忆的物种中是可以进化的。特里弗斯在1971年对这个问题作了进一步的探讨。但当他进行有关这方面的写作时，他还没有看到史密斯提出的有关进化上稳定策略的概念。如果他那时已经看到的话，我估计他是会加以利用的，因为这个概念很自然地表达了他的思想。他提到"俘虏的窘境"--博弈论中一个人们特别喜爱的难题，这说明他当时的思路和史密斯的已不谋而合。 假设B头上有一只寄生虫。A为它剔除掉。不久以后，A头上也有了寄生虫。A当然去找B，希望B也为它剔除掉，作为报答。结果B嗤之以鼻，掉头就走。B是个骗子。这种骗子接受了别人的恩惠，但不感恩图报，或者即使有所报答，但做得也很不够。和不分青红皂白的利他行为者相比，骗子的收获要大，因为它不花任何代价。当然，别人为我剔除掉危险的寄生虫是件大好事，而我为别人梳理整饰一下头部只不过是小事一桩，但毕竟也要付出一些代价，还是要花费一些宝贵的精力和时间。 假设种群中的个体采取两种策略中的任何一种。和史密斯所做的分析一样，我们所说的策略不是指有意识的策略，而是指由基因安排的无意识的行为程序。我们姑且把这两种策略分别称为傻瓜和骗子。傻瓜为任何人梳理整饰头部，不问对象只要对方需要。骗子接受傻瓜的利他行为，但却不为别人梳理整饰头部，即使别人以前为它整饰过也不报答。象鹰和鸽的例子那样，我们随意决定一些计算得失的分数。至于准确的价值是多少，那是无关紧要的，只要被整饰者得到的好处大于整饰者花费的代价就行。在寄生虫猖獗的情况下，一个傻瓜种群中的任何一个傻瓜都可以指望别人为它整饰的次数和它为别人整饰的次数大约相等。因此，在傻瓜种群中，任何一个傻瓜的平均得分是正数。事实上，这些傻瓜都干得很出色，傻瓜这个称号看来似乎对它们不太适合。现在假设种群中出现了一个骗子。 由于它是唯一的骗子手，它可以指望别人都为它效劳，而它从不报答别人给它的好处。它的平均得分因而比任何一个傻瓜都高。骗子基因在种群中开始扩散开来。傻瓜基因很快就要被挤掉。这是因为骗子总归胜过傻瓜，不管它们在种群中的比例如何。譬如说，种群里傻瓜和骗子各占一半，在这样的种群里，傻瓜和骗子的平均得分都低于全部由傻瓜组成的种群里任何一个个体。不过，骗子的境遇还是比傻瓜好些，因为骗子只管捞好处而从不付出任何代价，所不同的只是这些好处有时多些，有时少些而已。当种群中骗子所占的比例达到百分之九十时，所有个体的平均得分变得很低：不管骗子也好，傻瓜也好，它们很多因患蜱所带来的传染病而死亡。即使是这样，骗子还是比傻瓜合算。那怕整个种群濒于灭绝，傻瓜的情况永远不会比骗子好。因此，如果我们考虑的只限于这两种策略，没有什么东西能够阻止傻瓜的灭绝，而且整个种群大概也难逃覆灭的厄运。 现在让我们假设还有第三种称为斤斤计较者的策略。斤斤计较者愿意为没有打过交道的个体整饰。而且为它整饰过的个体，它更不忘记报答。可是哪个骗了它，它就要牢记在心，以后不肯再为这个骗子服务。在由斤斤计较者和傻瓜组成的种群中，前者和后者混在一起，难以分辨。两者都为别人做好事，两者的平均得分都同样高。在一个骗子占多数的种群中，一个孤单的斤斤计较者不能取得多大的成功。它会化掉很大的精力去为它遇到的大多数个体整饰一番--由于它愿意为从未打过交道的个体服务，它要等到它为每一个个体都服务过一次才能罢休。因为除它以外都是骗子，因此没有谁愿意为它服务，它也不会上第二次当。如果斤斤计较者少于骗子，斤斤计较者的基因就要灭绝。可是，斤斤计较者一旦能够使自己的队伍扩大到一定的比例，它们遇到自己人的机会就越来越大，甚至足以抵消它们为骗子效劳而浪费掉的精力。在达到这个临界比例之后，它们的平均得分就比骗子高，从而加速骗子的灭亡。当骗子尚未全部灭绝之前，它们灭亡的速度会缓慢下来，在一个相当长的时期内成为少数派。因为对已经为数很少的骗子来说，它们再度碰上同一个斤斤计较者的机会很小。因此，这个种群中对某一个骗子怀恨在心的个体是不多的。 我在描述这几种策略时好象给人以这样的印象：凭直觉就可以预见到情况会如何发展。其实，这一切并不是如此显而易见。为了避免出差错，我在计算机上摸拟了整个事物发展的过程，证实这种直觉是正确的。斤斤计较的策略证明是一种进化上稳定的策略，斤斤计较者优越于骗子或傻瓜，因为在斤斤计较者占多数的种群中，骗子或傻瓜都难以逞强。不过骗子也是ESS，因为在骗子占多数的种群中，斤斤计较者或傻瓜也难以逞强。一个种群可以处于这两个ESS中的任何一个状态。在较长的一个时期内，种群中的这两个ESS可能交替取得优势。按照得分的确切价值--用于模拟的假定价值当然是随意决定的--这两种稳定状态中的一种具有一个较大的"引力区"，因此这种稳定状态因而易于实现。值得注意的是，尽管一个骗子的种群可能比一个斤斤计较者的种群更易于灭绝，但这并不影响前者作为ESS所处的地位。如果一个种群所处的ESS地位最终还是驱使它走上灭绝的道路，那么抱歉得很，它舍此别无他途。 观看计算机进行模拟是很有意思的。模拟开始时傻瓜占大多数，斤斤计较者占少数，但正好在临界频率之上；骗子也属少数，与斤斤计较者的比例相仿。骗子对傻瓜进行的无情剥削首先在傻瓜种群中触发了剧烈的崩溃。骗子激增，随着最后一个傻瓜的死去而达到高峰。但骗子还要应付斤斤计较者。在傻瓜急剧减少时，斤斤计较者在日益取得优势的骗子的打击下也缓慢地减少，但仍能勉强地维持下去。在最后一个傻瓜死去之后。骗子不再能够跟以前一样那么随心所欲地进行自私的剥削。斤斤计较者在抗拒骗子剥削的情况下开始缓慢地增加，并逐渐取得稳步上升的势头。接着斤斤计较者突然激增，骗子从此处于劣势井逐渐接近灭绝的边缘。由于处于少数派的有利地位同时因而受到斤斤计较者怀恨的机会相对地减少，骗子这时得以苟延残喘。不过，骗子的覆灭是不可挽回的。它们最终慢慢地相继死去，留下斤斤计较者独占整个种群。说起来似乎有点自相矛盾，在最初阶段，傻瓜的存在实际上威胁到斤斤计较者的生存，因为傻瓜的存在带来了骗子的短暂的繁荣。 附带说一句，我在假设的例子中提到的不相互整饰的危险性并不是虚构的。处于隔离状态的老鼠往往在舌头舔不到的头部长出疮来。有一次试验表明，群居的老鼠没有这种毛病，因为它们相互舔对方的头部。为了证实相互利他行为的理论是正确的，我们可以进行有趣的试验，而老鼠又似乎是适合于这种试验的对象。 特里弗斯讨论过清洁工鱼（cleaner fish）的奇怪的共生现象。已知有五十个物种，其中包括小鱼和小虾，靠为其他物种的大鱼清除身上的寄生虫来维持生活。大鱼显然因为有人代劳，为它们做清洁工作而得到好处，而做清洁工的鱼虾同时可以从中获得大量食物。这样的关系就是共生关系。在许多情况下，大鱼张大嘴巴，让清洁工游入嘴内，为它们剔牙，然后通过鱼鳃游出，顺便把鱼鳃也打扫干净。有人认为，狡猾的大鱼完全可以等清洁工打扫完毕之后把它吞掉。不过在一般情况下，大鱼总是让清洁工游出，碰都不碰它一下。这显然是一种难能可贵的利他行为。因为大鱼平日吞食的小鱼小虾就和清洁工一样大小。 清洁工鱼具有特殊的条纹和特殊的舞姿，作为清洁工鱼的标记。大鱼往往不吃具有这种条纹的小鱼，也不吃以这样的舞姿接近它们的小鱼。相反，它们一动不动，象进入了昏睡状态一样，让清洁工无拘无束地打扫它们的外部和内部。出于自私基因的禀性，不择手段的骗子总是乘虚而入。有些物种的小鱼活象清洁工，也学会了清洁工的舞姿以便安全地接近大鱼。当大鱼进入它们预期的昏睡状态之后，骗子不是为大鱼清除寄生虫，而是咬掉一大块鱼鳍，掉头溜之大吉。但尽管骗子乘机捣乱，清洁工鱼和它们为之服务的大鱼之间的关系，一般地说，还是融洽的，稳定的。清洁工鱼的活动在珊瑚礁群落的日常生活中起着重要的作用。每一条清洁工鱼有其自己的领地。有人看见过一些大鱼象理发店里排队等候理发的顾客一样排着队伍，等候清洁工依次为它们搞清洁工作。这种坚持在固定地点活动的习性可能就是延迟的相互利他行为形成的原因。大鱼能够一再惠顾同一所"理发店"而不必每次都要寻找新的清洁工，因此，大鱼肯定感觉到这样做要比吃掉清洁工好处大。清洁工鱼本来都是些小鱼，因此这种情况是不难理解的。当然，模仿清洁工的骗子可能间接地危害到真正的清洁工的利益，因为这种欺骗行为产生了一些压力，迫使大鱼吃掉一些带有条纹的、具有清洁工那种舞姿的小鱼。真正的清洁工鱼坚持在固定地点营业，这样，它们的顾客就能找上门来，同时又可以避开骗子了。 当我们把相互利他行为的概念运用于我们自己的物种时，我们对这种概念可能产生的各种后果可以进行无穷无尽的耐人寻味的猜测。尽管我也很想谈谈自己的看法，可是我的想象力并不比你们强。我想还是让读者自己以此自娱吧！