Biology Reference
In-Depth Information
(a)
(b)
100
100
80
80
60
60
40
40
20
20
0
0
Siblings
Cousins
Full-sibs
(~0.5)
Mixture of sibs
and non-sibs
(~0.0.3)
Relatedness to potential prey
Fig. 11.9
Cannibalism in Arizona tiger salamanders. (a) Larvae are more likely to develop as the cannibal
morph when reared with non-relatives. (b) Cannibal morphs were more likely to eat individuals to who they
were less closely related. From Pfennig and Collins (1993). Reprinted with permission from the Nature Publishing
Group. Photograph of a cannibal morph eating a typical morph. Photo © David Pfennig.
In a series of elegant experiments, Pfennig
et al
. (1999) then went on to test a number
of alternative explanations for this observation. They rejected alternate hypotheses that
did not rely on the indirect costs of eating relatives. For example, avoiding eating relatives
might be a disease avoidance strategy, if it is easier to acquire parasites by eating closer
relatives. This might be the case if parasites are adapted to infect specific genotypes, in
which case the parasites that have been able to infect an individual, are also likely to be
successful at infecting their relatives, who have similar genotypes. However, this hypothesis
was rejected because cannibals were not more likely to become infected with bacteria or
viruses by eating kin (in fact the opposite occurred), cannibals did not avoid diseased prey,
and cannibalism was not less common in patches with higher rates of disease.
In contrast, Pfennig
et al
. (1999) found clear support for the kin selection hypothesis,
that individuals are avoiding the indirect costs of eating relatives. He carried out a
quantitative test of Hamilton's rule by estimating the costs and benefits of eating
siblings. Hamilton's rule predicts that cannibals should be altruistic (avoid eating
siblings) if
rB
−
C
>
0, in other words if
B
/
C
>
1/r
. For full siblings,
r
=
0.5, so the eating of
full siblings should be avoided if
B
/
C
2.
Pfennig estimated the benefit (
B
) and cost (
C
) of not eating relatives, by taking
advantage of the fact that there was variation across individuals in the extent to which
they avoid eating kin. Some individuals (discriminators) show a strong tendency to avoid
eating siblings, whilst others (non-discriminators) show no tendency to avoid eating kin.
Pfennig placed either a discriminator or a non-discriminator cannibal larvae in a
container with 24 typical larvae, six of which were siblings and the remainder non-
siblings. He found a large indirect benefit (
B
) of kin discrimination, with approximately
f our siblings surviving in the presence of a discriminator cannibal and two in the presence
of a non-discriminatory cannibal, giving
B
≈ 4 − 2
>
… because there
is a large indirect
fitness benefit to
not eating
siblings
2. He also found no significant cost
of kin discrimination, with non-discriminators growing at the same rate, and reaching
maturity at the same age, as discriminators, suggesting
C
≈ 0. This suggests that
B
/
C
=
2,
in which case kin discrimination over cannibalism would be explained by kin selection.
>
