Abstract
There is a plethora of research showing that herbivorous arthropods can overcome constitutive and induced direct plant-defences such as secondary metabolites. However, plants can also use secondary metabolites in indirect defences. Plants may produce volatile compounds in response to herbivory (caused by mechanical damage and herbivore-related elicitors) and these herbivore-induced plant-volatiles (HIPV) elicit responses in predatory arthropods thus establishing an external 'immune system' for the plant. It is widely accepted that these induced indirect defences occur throughout the plant kingdom. This raises the question how herbivorous arthropods cope with indirect defense mechanisms, such as those mediated by HIPV. Herbivores have four options to develop resistance against indirect defenses: (1) avoiding HIPV-producing plants or plant parts, (2) decreasing production of elicitors, (3) blocking recognition by the plant through elicitor-modification and (4) interfering with the HIPV-biochemistry of the plant. We propose to study these four scenario's in a tri-trophic system consisting of tomato plants (Lycopersicon esculentum), two-spotted spider mites (Tetranychus urticae) and predatory mites (Phytoseiules persimilis). This tri-trophic system is accessible to experimentation because HIPV-production is well studied, molecular assays for tomato defense responses have been developed and olfactory responses of predators are well analysed. Moreover, the spider mites have been shown to harbour genetic variation for their responses to secondary plant metabolites. Also, it has been shown that some genetically different spider mite strains evoke a distinctly different induced indirect defense-response in tomato. We aim to select several spider mite strains and lines that differ in the degree to which they induce HIPV-production in tomato. Next, we aim to discriminate between the four types of resistance against indirect defenses by a combination of behavioural analysis, assessment of gene transcription in tomato plants and quantification of HIPV-output. In case of plant (part) avoidance we expect spider mites to select plant(parts) that will release less HIPV after herbivory. The tomato defense genes we want to test are selected from EST libraries and spotted onto micro-array slides. In case of a decreased elicitor secretion we expect transcription of a selection of defense genes to decrease, wheras in vase of elicitor modification we expect absence of gene-activation. Finally, if herbivore secretions interfere with the HIPV-biochemistry we expect gene-transcription in planta but a modified HIPV-release. Because the plant's indirect defense response depends on recognizing (chemical products released by) its attacker and the attacker may prevent recognition or interefere with the plant defense response, we hypothesize that (by analogy with several plant-pathogen systems) plants and herbivores are involved in an arms-race leading to gene-for-gene co-evolution. We intend to test one of the assumptions underlying this hypothesis: the existence of genetic variation with a simple mode of inheritance for induction of HIPV by herbivorous mites in a given genotype of tomato plants. This is done by artificial selection of isofemale lines of herbivorous mites that differ in the degree and/or the way they modify HIPV release and are maintained by mating virgin spider-mite females with their partheno-genetically produced, haploid sons. The number of genes involved in these traits will be estimated from the response to selection, but we will also attempt to test simple models on mode of inheritance against that on tomato gene-transcription induced by parental mite-lines, their hybrids and offspring from backcrosses of hybrid females with males from parental lines.
Netherlands