Interestingly, a form of cheating has been documented at the inter-
specific level: it appears that some non-manipulative parasite species
try to ‘hitch’ a ride with a manipulative parasite species with which
they share intermediate and definitive hosts (Thomaset al., 1997, 1998;
Laffertyet al., 2000). These hitchhikers obtain the benefits of manipu-
lation, i.e. increased transmission rate to the definitive host, but let the
other species assume the full costs. Using simple models, it is possible to
predict which conditions should favour the evolution of hitchhiking
(Thomaset al., 1998) and, even though there are two parasite species
involved, it should be possible to derive the ESS or the optimal ratio of
manipulators and non-manipulators.
Finally, the concept of kin selection and its associated notion
of inclusive fitness have shed a new light on cooperation and conflict
among individuals (Hamilton, 1964). Any organism shares alleles with its
relatives, and its inclusive fitness depends on its own success at reproduc-
tion (direct fitness) and on how much it raises the reproductive success of
its relatives by helping them in any way (indirect fitness). In the examples
discussed above where several conspecific parasites occur in the same
intermediate host, what happens if the parasite group consists mostly or
entirely of close kin? This is the case in the trematodeD. dendriticum:
because of the way in which they are acquired by their ant intermediate
host, all metacercariae inside one ant are likely to be derived from a single
egg by asexual reproduction in their first host, a snail (Wickler, 1976). The
one metacercaria that induces the manipulation of ant behaviour and
then dies thus benefits in terms of inclusive fitness: some of its genes are
transmitted via its kin. Presumably, having the full costs incurred by a
single individual rather than shared among relatives yields greater overall
benefits. BecauseD. dendriticummetacercariae are clones, it may be that
it is more advantageous for one to die than for all to suffer from a slightly
lower fitness. Systems similar to this one may be relatively common.
Because of patchiness in the distribution of eggs from different individu-
als and because of the amplification of this patchiness by asexual multi-
plication of larval stages in trematodes, it may be that manipulative
parasites sharing an intermediate host with conspecifics may in fact
be surrounded by relatives. Kin-selection theory would allow various
predictions to be made about the optimal partitioning of investments in
manipulation among co-occurring parasites once their genetic relatedness
is determined.
My aim in this section has been to illustrate how some of the powerful
conceptual tools developed by behavioural ecologists can be applied to
the study of host manipulation by parasites. Other theoretical frameworks
have been used to generate predictions regarding whether or not parasites
should manipulate host behaviour and how strong the manipulation
should be. These include population-dynamics models (Dobson, 1988)
and verbal models based on life history and ecological classification
of parasites (Kuris, 1997). The many predictions derived from these250 R. Poulin