spp.) as intermediate hosts and the birds that feed on them as definitive
hosts. Both parasites induce a very similar photophilic behaviour in their
amphipod hosts, an analogous trait that arose twice independently by
convergent evolution. Another example comes from a comparison of
nematomorphs (hair-worms) and mermithid nematodes, two unrelated
taxa that both evolved similar life cycles. Juveniles of both groups develop
inside terrestrial insects until they become adults, when they must finish
their life in fresh water. Although predation by a definitive host is
not involved, the parasites still need to manipulate the behaviour of
their hosts to induce them to throw themselves in water. Indeed, both
nematomorphs and mermithid nematodes are known to trigger a marked
hydrophilia in their insect hosts, which results in the parasites emerging
into a suitable aquatic habitat (Poinar, 1991; Maeyamaet al., 1994). Again,
this suggests convergent evolution by different parasites facing similar
difficulties.
In other cases, related parasites in similar situations appear to use
different approaches to solve the same problem. For instance, two gymno-
phallid trematodes that must be transmitted by predation from bivalves to
shore birds (oyster-catchers, gulls) have chosen different manipulation of
the bivalve host to achieve a greater transmission success.Meiogymno-
phallus fossarumencysts in the mantle of its bivalve host and causes it to
reverse its position in the mud, so that the shell opening faces upward,
and to remain closer to the surface than uninfected bivalves (Bartoli,
1974). In contrast,Parvatrema affinisencysts anywhere in its bivalve
host, but somehow causes it to leave noticeable tracks on the surface of
tidal flats that indicate where it has burrowed (Swennen, 1969; but see
Mouritsen, 1997). Clearly, the problem of getting birds to feed on bivalves
has many solutions, and the two gymnophallid trematodes above have
independently solved that problem. There are other solutions as well that
have been adopted independently by echinostomatid trematodes with
identical life cycles: some of these species encyst in the foot of their
bivalve host, greatly impairing its function and leaving heavily infected
bivalves incapable of burrowing into the sediments (Lauckner, 1984;
Thomas and Poulin, 1998).
The impression one gets from the above examples is that host
manipulation by parasites has evolved independently on numerous
occasions. It is too early to estimate just how often this trait evolved;
however, based on the range of parasite taxa in which transmission via
predation occurs (Kuris, 1997) and on the wide number of taxa in which
manipulation has been documented (Moore and Gotelli, 1990), the actual
number must be much greater than 20 times. Because of convergent
evolution, manipulations of host behaviour in different systems often
appear superficially similar, as though there were only a few solutions to a
particular transmission problem. However, no doubt the true diversity of
approaches used by manipulating parasites will become apparent once
we uncover the mechanisms that different parasites may use to induce
seemingly identical alterations in their hosts’ behaviour.252 R. Poulin