IOBC/wprs Bull. 29(8): 15-23. 1 Investigating the ecology of inducible indirect defence by manipulating plant phenotype and genotype Marcel Dicke, Maaike Bruinsma, Tibor Bukovinszky, Rieta Gols, Peter W. de Jong, Joop J.A. van Loon, Tjeerd A.L. Snoeren, Si-Jun Zheng Laboratory of Entomology, Wageningen University, P.O. Box 8031, NL-6700 EH Wageningen, The Netherlands, www.dpw.wau.nl/ento/english Abstract A challenge for ecologists has been to understand how individual traits of organisms affect species interactions and community dynamics. Recent breakthroughs provide ecologists with delicate manipulative tools in which mechanistic knowledge of well-characterized genotypes and phenotypic plasticity can be exploited to study the effect of individual plant traits on interactions in ecosystems. Food webs are overlaid with infochemical webs that mediate direct and indirect interactions. It is increasingly clear that indirect interactions can have important effects on community dynamics. Infochemicals are interesting in this respect because they cannot be directly used in bodybuilding, yet the responses they elicit have important consequences for fitness, and thus for interactions in a community. Infochemicals from plants influence interactions with members of different trophic levels, such as carnivores and herbivores. The infochemical-phenotype of plants is plastic: infochemical emission is an active and specific process that is induced by herbivory. The infochemicals attract carnivores that affect the herbivore population. Additionally, the infochemicals also affect herbivore behaviour and characteristics of neighbouring competitor plants. Careful manipulation of the phenotypically plastic emission by plants provides unique opportunities to investigate the effect of the infochemicals on food-web interactions. This novel approach creates an essential link between molecular, chemical, behavioural and community ecology. Introduction A major challenge for biology in the 21 st century is to integrate research approaches that address different levels of biological organisation: i.e. from subcellular processes all the way to community processes. A pressing issue in ecology is to understand how direct and indirect interactions among individual organisms influence food webs and community dynamics. This can now be addressed through a novel integration of approaches: from ecogenomics, through behavioural ecology to community ecology. Chemical cues are a major source of information for very different organisms ranging from microorganisms to mammals (e.g. Dicke and Grostal 2001, Kats and Dill 1998, Roitberg and Isman 1992, Tollrian and Harvell 1999). Chemical information affects various behaviours that underlie population dynamics and food web interactions, including the selection of food, the selection of mates, competition and the avoidance of predators (e.g. Dicke and Vet 1999, Hilker et al. 2002, Kats and Dill 1998, Roitberg and Isman 1992, Sabelis et al. 1999, Turlings and Benrey 1998). Therefore, chemical information is an important factor influencing species interactions and most likely also community processes (Vet 1998). However, the study of chemical information conveyance has been mostly restricted to studies at the level of individual organisms and the identification of the chemicals that convey the information. The influence of chemical information on food web processes has received little attention (Hunter 2002, Van der Meijden and Klinkhamer 2000, Vet 1998), in contrast to effects of direct trophic interactions (Morin 1999). Yet, circumstantial evidence indicates that chemical information from phenotypically plastic plants can have important influences on food web dynamics through indirect effects that combine bottom-up and top-down effects (Dicke and Vet 1999, Sabelis et al. 1999). Empirical support should come from manipulative experiments (manipulations of plant infochemical phenotype) that compare food web processes in the presence and absence of plant infochemicals. The ability to manipulate the infochemical phenotype of plants in specific and well- known ways is indispensable for this approach. Such experiments have recently come within reach. A