but we only know how these changes are mediated in a few systems.
Recent trends in behavioural ecology suggest that functional and mecha-
nistic approaches to the study of behaviour are increasingly being merged
into a more holistic approach (Krebs and Davies, 1997). Hopefully this
new way of thinking will soon reach the area of parasite manipulation of
host behaviour.Conclusions
The study of host manipulation by parasites presents a special
complication for behavioural ecologists: the trait of interest is the product
of the interaction between the genotypes of two different organisms.
The behaviour of a parasitized host is the simultaneous expression of
host and parasite genes, and investigating the function, evolution and
proximate basis of a change in host behaviour can be very challenging.
In this chapter, I have considered only cases in which the modified
host behaviour appeared to be mainly the product of parasite genes and
therefore of benefit to the parasite. However, without proper study of
both the immediate causes of the change in host behaviour and of its net
fitness consequences for the parasite, it is premature to consider it as a
parasite adaptation. Behavioural ecologists have been accused in the past
of adopting a Panglossian view of nature, in which every feature of an
organism is always beneficial (Gould and Lewontin, 1979). Although
changes in host behaviour are often detrimental to the host, they are not
necessarily always advantageous for the parasite. And, even if they are
advantageous for the parasite, they may not be true adaptations. Given
the tools now available, the next step in this field of research should
be to integrate investigations into the functional, causal and historical
aspects of host manipulation. First, clear, quantitative predictions derived
from optimality theory or game theory need to be tested in experimental
systems or in comparative analyses to provide a deeper understanding
of when the ability to manipulate hosts evolves and how strongly it
manifests itself. Secondly, the proximal cause of behavioural changes
needs to be elucidated in more than a handful of systems. Thirdly,
information on both the magnitude and type of manipulation induced
by parasites and on the physiological mechanisms used to induce
them needs to be mapped on parasite phylogenies to provide insight
into their evolutionary history. There have been recent developments
in the broader implications of host manipulation by parasites, such
as its influence on the development of parasite communities, on the
evolution of other sympatric parasites and on the evolution of hosts them-
selves (see Poulin and Thomas, 1999; Laffertyet al., 2000). To better
understand the phenomenon itself, however, the three steps outlined
above will be necessary if we are to go beyond the mere natural-history
perspective.254 R. Poulin