leaves containing non-host herbivores and by spending considerable
amounts of time on such leaves. Since most plants in the field are likely to
be infested by more than one herbivore species, Voset al. (2001) modelled
the implications of such time wasting at the community level. They
showed that an increase in herbivore diversity on a plant initially
promoted the persistence of parasitoid communities. However, when
herbivore diversity was very high, parasitoids became extinct, due to
insufficient parasitism rates. Hence, the informational value of plants
determines the host-encounter rate of parasitoids and this can have far-
reaching consequences for parasitoid–host interactions and community
structure.Visual cues
Although we may stress the importance of odour cues during foraging, a
parasitoid’s sensory world is not restricted to the chemical dimension.
Individual substrates may also differ in respect of other sensory inform-
ation. Plants or plant parts, for instance, may be recognized by visual cues
(colour, shape or patterns of feeding damage), as well as other physical
characteristics (surface structure, vibration conduction). The parasitoid’s
ability to use detailed visual information during foraging sets it apart from
other parasites. In a parasitoid’s natural habitat, there are several contexts
in which visual cues could provide useful information. Parasitoids may
employ colour signals to locate floral nectar, a source of food (see below),
or to find herbivores feeding on conspicuously coloured plant structures.
In those cases in which the coloration of plant structures is consistently
and specifically altered due to herbivory (e.g. galls or leaf-mines), the
coloration of infested plant structures could become a particularly reliable
stimulus to host-seeking parasitoids. Irrespective of coloration, the shape
of herbivore damage can be an apparent and reliable indicator of herbi-
vore presence. Recent work has demonstrated that parasitoids use various
visual parameters during the search for food and host sites, and learn to
discriminate between sites on the basis of these visual characteristics
(Wäckers and Lewis, 1999). In visual learning experiments,Microplitis
croceipesperformed equally well in shape discrimination as in colour
learning. Parasitoids were conditioned in a flight chamber with plants
containing two types of paper targets, differing either in colour
(yellow/grey) or in shape (triangle/square). As only one visual alternative
carried a Spodoptera larva, parasitoids could associate the visual
information with an oviposition reward. After parasitizing six hosts, the
parasitoid’s response to targets was tested in a separate run in which
no caterpillars were present. Parasitoids that had been conditioned
to discriminate colours made 73% of their landings on the previously
rewarded colour alternative. In the case of shape discrimination, this
figure was 79%. This high relative rate of shape learning is remarkable as
it is in strong contrast to the learning capacity of honey-bees. The latter46 L.E.M. Vetet al.