Parasitic wasp‐associated polydnavirus and venom affect arthropod community assembly on wild cabbage via the caterpillar host

Plant responses to insect feeding mediate interactions with other community members and may influence community assembly on the induced plant. Since plants perceive herbivory through damage patterns and salivary components, changes in these herbivore traits, such as caused by parasitism, may alter herbivore‐induced plant‐mediated interactions.

Upon parasitism, many parasitoids inject a combination of polydnavirus and venom into their host. In doing so, the host's behaviour and physiology are modified. The injection of polydnavirus has been identified as a prominent driver of plant‐mediated interactions initiated by parasitised herbivores. However, most evidence for such plant‐mediated effects comes from greenhouse or laboratory studies.

We investigated the ecological consequences of parasitism and its associated injection components on assembly of arthropod communities in an open‐field setting. We exposed wild Brassica oleracea plants to herbivory by unparasitised Pieris brassicae caterpillars, caterpillars parasitised by Cotesia glomerata , caterpillars injected with polydnavirus and/or venom, and we used uninduced plants as control. We monitored the naturally occurring arthropod community on these plants throughout the growing season, focusing on both overall community composition and the abundance of individual species.

Arthropod community composition was marginally affected by components of parasitism and subtle effects were found for interactions with specific herbivore species. Plants damaged by P. brassicae injected with both polydnavirus and venom were colonised more often by Mamestra brassicae caterpillars, Myzus persicae aphids and aphid parasitoids compared to uninduced plants or plants induced by unparasitised P. brassicae caterpillars.

Our results show that parasitoid polydnavirus and venom, either alone or in combination, affect colonisation by members of the arthropod community in a natural field setting. These modulating effects deserve attention in future studies that investigate the dynamics of plant‐arthropod communities.