快猫短视频

IF YOU鈥橵E spent hours in the past few weeks battling through the mall, armed
only with a roll of wrapping paper and a shopping bag, you are probably
wondering why we go to such lengths to buy gifts for each other at this time of
year. This鈥攐r something like it鈥攊s a question that has also
intrigued biologists since Darwin鈥檚 day. Because natural selection favours
individuals who do well against their competitors, you would expect all living
things to be inherently selfish. But while our seasonal generosity is unique,
the animal world is bursting with examples of selfless giving that seem to
transcend basic survival instincts.

Many animals spend precious time grooming one another. Some will act as
lookouts for their group. Others help older adults to raise their young, rather
than reproducing themselves. 鈥淭he most extreme examples of altruism are
certainly among the social insects,鈥� says Richard Dawkins of Oxford University.
鈥淟ook at the soldier ants, who dig their jaws into a marauder, and never let go.
Or the honeypot ants who fill their bodies with sugar and act as living food
larders, to store food for other members of the colony.鈥�

Over the years biologists have come to realise that things are often not what
they seem. Sometimes helping other individuals, or even the group, is the best
way to ensure your own wellbeing. Is that why you bought that dishwasher for
your wife? Other times, seeming selflessness can be explained as an evolutionary
urge to promote the survival of relatives with whom we share some of our selfish
genes. This might account for the woolly socks you sent your brother in Alaska.
But there鈥檚 another type of altruism that is a bit more difficult to explain:
where the benefits of giving are not obvious and there鈥檚 no family connection.
Why exactly did you give that copy of Cheating Monkeys and Citizen Bees
to the workmate with whom everyone knows you don鈥檛 really see eye-to-eye?
Perhaps it鈥檚 because you expected something in return. Recent studies suggest
that not only is reciprocity a source of social cohesion, it also helps to
explain the escalating generosity that seems to take hold of us at
Christmas.

It is three decades since William Hamilton from Oxford University suggested
the idea of kin selection and built the foundation for the selfish gene theory.
He realised that you could understand why altruism is common within families if
you imagined that natural selection is working on an animal鈥檚 genes, rather than
the animal itself. The genes that make an individual likely to promote the
survival of its kin will proliferate down the generations because both animals
carry copies of the very same genes. But why would someone be nice to a
non-relative?

Prisoner鈥檚 Dilemma

In 1971, a biologist named Robert Trivers came up with an answer. He
suggested that animals might do favours for unrelated individuals because they
expect the favour to be returned in the future. He called the concept reciprocal
altruism. If reciprocation is to be a convincing explanation for altruism in
animals, you need to show that it can benefit individual animals in the long
term, even if they run a risk of meeting a cheat who may not return the favour.
And biologists have found a clever way to do this, without involving any real
animals at all. They use computers to simulate animals that continually play
games with one another in which they can choose whether or not to cooperate. The
games these cyber-creatures play are based on a concept known as the Prisoner鈥檚
Dilemma, which was originally developed to model social and military
conflicts.

The Prisoner鈥檚 Dilemma encapsulates the risk of reciprocation perfectly,
because the reward you get depends on what your partner chooses to do. If both
players cooperate, they each get an equal reward. But if the first player
cooperates and the second one sells out, the cheat receives the highest possible
reward, and the first player is fobbed off with a sucker鈥檚 payoff, which is the
lowest reward available. The temptation to cheat, then, is strong, and an
individual who cooperates runs a risk of losing spectacularly. On the other
hand, if both players cheat, they each receive a low payoff. The dilemma is that
the individual payoff is greatest if one player defects while the other does
not, but the average payoff is higher for both players if they cooperate.

So what is the best strategy? If we were dealing with humans, we might expect
two players, or indeed the entire population of players, to formally agree that
everyone cooperates all the time, and that people who cheated would be punished.
Which in fact is how civilised society is supposed to work. But can a similar
thing occur in the animal world, where signing contracts is not an option?

Generation game

We can use the Prisoner鈥檚 Dilemma to model natural selection if the payoff
represents fitness鈥攖hat is an animal鈥檚 ability to reproduce. In the model,
animals follow various strategies such as always cheat or always cooperate.
Those that tend to win a high payoff get to reproduce more than those with a
less successful strategy, so that the relative frequency of successful
strategies increases. If you run the game for a large number of 鈥済enerations鈥�,
allowing the cyber animals to play each other over and over again, eventually
the population should reach a stable balance where the most successful strategy
comes to dominate. And if that dominant strategy turns out to be a cooperative
one, then you have shown that cooperation can evolve in the best interests of
the individual animal.

The first paper showing that altruism could evolve was published in 1981 by
Hamilton and Robert Axelrod from the University of Michigan. It was based on a
strategy known as tit-for-tat. Individuals adopting this strategy play the
tactic their opponent played in the previous round. So assuming all the players
start off being generous, a whole population of tit-for-tat players simply
cooperate all the time. If a cheat arrives, that player will win the highest
prize for cheating against a cooperator once, after which it does consistently
badly, because tit-for-tat retaliates. More recently, there have been many
models that elaborate on tit-for-tat, making it more forgiving, or incorporating
mistakes. 鈥淭it-for-tat explains pretty well what has gone on,鈥� says Hamilton.
鈥淭he latest models which build on it just make the process seem even easier than
we thought.鈥�

The next question is obvious: do animals actually behave in this way? And
surprisingly, 17 years on, there is still no evidence that any species is
actually caught in a Prisoner鈥檚 Dilemma in its natural environment, and very
little evidence that animals follow a tit-for-tat type strategy.

Clever experiments with sticklebacks and guppies show that these fish use
tit-for-tat to decide on a partner when they risk their lives to inspect an
approaching predator (see 鈥淔ish United鈥�). But most cases of reciprocal altruism
we see in the wild are more complex. For example, a female vampire bat who has
recently eaten will regurgitate a meal to save a nest mate from starvation. She
will only do this for another bat with whom she has been a frequent roost mate,
and she is more likely to donate blood to a bat that has helped her in the past.
But a frequent roost mate could also be a relative, so it is difficult to know
whether kin selection or reciprocity explain such a sacrifice. And the more
complex the social structure under scrutiny, the more advanced the systems
of reciprocity. Primates, for example, barter one type of favour for another of
a different kind. 鈥淯nrelated vervet monkeys seem to exchange grooming for aid in
coalitions,鈥� says Lee Dugatkin from the University of Louisville, 鈥渨hile no such
exchange exists among related individuals.鈥�

So it鈥檚 not really surprising that some researchers doubt whether game theory
and tit-for-tat can ever describe nature. 鈥淭hese are fragile explanations
because they rely on a number of prerequisites which are not often met in the
animal world,鈥� says Robert May from Oxford University. 鈥淭he animals are required
to remember specific encounters, individuals and outcomes of each game and they
must anticipate future encounters with the same individual.鈥� He points out that
the theory also gives the same value to games in the future, without taking into
account the fact that they may not happen.

Theoretical playground

Gilbert Roberts from the University of Newcastle also has reservations. 鈥淭he
Prisoner鈥檚 Dilemma has become a theoretician鈥檚 playground. It becomes more and
more diverged from the behaviours it is trying to explain,鈥� he says. But with
colleague Thomas Sherrat, he has come up with a finding that he hopes will start
to unite the theory with studies in animal behaviour. 鈥淚 was inspired by the
observation that most examples of cooperation are not all-or-nothing
behaviours,鈥� says Roberts. 鈥淏irds that preen one another can vary the amount of
time they spend doing so.鈥� He wondered how animals decide how much of an
investment to make in another individual.

Roberts and Sherrat used the Prisoner鈥檚 Dilemma to see which strategies of
variable investment could possibly evolve and lead to stable populations of
cooperators. Some players were 鈥渟hort changers鈥�, consistently cooperating less
than their opponent had in the previous round. Others matched their opponent鈥檚
investment exactly, or refused to cooperate at all. The strategy that emerged as
successful was called 鈥渞aise-the-stakes鈥�, a form of tit-for-tat, where players
make a small investment in each other at first, but with each round of the game,
become slightly more generous than their opponent was in the previous round. The
big advantage of this finding is that it can be tested in nature. 鈥淭his opens
the way to simple experiments to test the theories of how reciprocal altruism
might work,鈥� says Laurent Keller from the University of Lausanne in Switzerland,
鈥渂ecause it is easy to measure the amount of time that animals spend
驳谤辞辞尘颈苍驳.鈥�

Christmas card panic

Could this also explain why the giving of gifts at Christmas is apt to get
out of hand? 鈥淩eciprocation is much more important than kin selection in
humans,鈥� says Hamilton. 鈥淎s animals, we have devoted a huge amount of resources
in the brain to recognising faces. That鈥檚 why we are such great tit-for-tatters.
And gift giving is deeply symbolic in all human cultures. How else do you
explain the depth of panic you feel when you receive a Christmas card from
someone who you have inadvertently omitted from your list?鈥�

There could be more to it than that, however. 鈥淗umans will help someone, even
though they are extremely unlikely to meet them again,鈥� says Martin Nowak from
the Institute for Advanced Study, Princeton. Working with colleague Karl
Sigmund, Nowak has come up with a model called indirect reciprocity that may
explain what looks on the surface like indiscriminate generosity. This time, the
Prisoner鈥檚 Dilemma is played out against a social background. Instead of pairs
of players reacting to one another in isolation from the rest of the cyber
population, other players witness acts of cooperation and betrayal. Each player
is assigned an 鈥渋mage score鈥� which is apparent to all the other players, so an
individual can judge whether or not to cooperate with another on the basis of
how altruistic they have seen that opponent to be.

The beauty of this model is that cooperation evolves even where individuals
may never meet each other again. And there is evidence that such a scoring
system can occur in the natural world. Arabian babblers, for example, are small
birds of unassuming appearance found in the Arabian peninsula. Within their
groups, generosity is associated with social status and the highest-ranking
babblers are the ones that do the most grooming and guarding. Consequently, the
birds fall over one another to do a good turn for a neighbour.

Humans may not be quite so generous, but we do have a tendency to be
indiscriminately nice and trusting. 鈥淲e are much nicer to one another than
appears to benefit our selfish genes,鈥� says Dawkins.

So, it鈥檚 presents for the family to bolster your genetic ties, gifts for
friends because you know they will be buying them for you, and something for the
office nerd to show everyone that you鈥檙e a wonderful colleague. And you thought
Christmas philanthropy was a selfless pursuit? Bah! Humbug!

THERE is only one example of animals following a tit-for-tat strategy where
altruism, or the lack of it, in one individual is reciprocated directly by
another. It involves a behaviour known as predator inspection, which some fish
display. Here one or two fish move slowly and carefully away from the school,
towards a potential predator. The scouts deliberately risk being eaten, but do
the group a favour by assessing whether an intruder is dangerous.

It looks as though the fish may base their decision about whether or not to
accompany another individual on such inspections on previous experience. To test
this idea Lee Dugatkin from the University of Louisville positioned a mirror
along the length of a tank containing guppies. This meant that a single fish
inspecting potential predators is led to believe it is accompanied by a
companion. The mirror is either parallel to the sides of the tank, so the fish
image swims towards the predator with the scout fish (it cooperates); or the
mirror is placed at an angle, so the other fish appears to swim away (it backs
down).

鈥淭he fish show all the characteristics of tit-for-tat,鈥� says Dugatkin. 鈥淭hey
start off behaving cooperatively, but they retaliate against a fish who appears
to swim away.鈥� In other words, they do not make the inspection with their
reflected image next time. 鈥淎nd they are forgiving, only remembering one move
back in a sequence of trials,鈥� says Dugatkin. 鈥淪o if the companion fish appears
to swim away time after time, and then the mirror is moved and it cooperates
just once, the real fish cooperates in response on the next inspection.鈥�