Cooperation - An Introduction to Behavioural Ecology - page 337

Player B

Table 12.1 The
Prisoner's dilemma
game (Axelrod &
Hamilton, 1981);
the pay-off to
player A is shown
with illustrative
numerical values

Cooperate Defect

R 3

Reward for mutual
cooperation

= S 0

Sucker's pay-off

=

Cooperate

Player A

T = 5
Temptation to defect

P = 1
Punishment for
mutual defection

Defect

In other words, cooperation is not an evolutionarily stable strategy (ESS) because in
a population of cooperators a mutant that defected would spread. Defect, however, is
an ESS; in a population of 'all defect' a mutant cooperator does not gain an advantage.
Any population with a mixture of heritable strategies will, therefore, evolve to 'all
defect'. More generally the conditions for this conclusion to hold in the matrix in
Table 12.1 are:

( ST + )

TRPS >>> > and R ,

2

which define the Prisoner's dilemma game. The problem is that while an individual can
benefit from mutual cooperation, it can do even better by exploiting the cooperative
efforts of others.
It is important to realize that the Prisoner's dilemma game is just an illustration of the
problem of cooperation, and not a solution. If we want to find a solution to the problem
of cooperation, we need to find situations where the lifetime pay-offs for cooperation or
defection are not as given in Table 12.1, and hence where the Prisoner's dilemma does
not hold.

Solving the problem of cooperation

How, then, can we account for the evolution of cooperative behaviour? As discussed in
Chapter 11, one possible solution to this problem is that cooperation can be favoured by
kin selection when it is directed towards relatives, and therefore provides indirect fitness
benefits. However, as discussed above, we also need to be able to explain cooperation
between non-relatives - in this case cooperation must provide some direct fitness benefit
to the cooperator. The distinction here between direct and indirect fitness is whether a
gene maximizes its transmission to the next generation by increasing either the fitness
of the individual it is in (direct fitness) or another individual with a copy of the same gene
(indirect fitness). Within these two broad categories, the explanations for cooperation
can be divided further in a number of ways (Fig. 12.2; Sachs et al ., 2004, Lehmann &
Keller, 2006; West et al ., 2007c).