Summary1. Quantitative food webs were constructed describing the community of leafmining moths in the genus Phyllonorycter (Gracillariidae) and their parasitoids at a site in southern England. 2. Twelve species of Phyllonorycter were reared from four species of tree. The moths were attacked by 27 species of hymenopterous parasitoids. 3. Four webs were constructed for successive moth generations over 2 years. A summary web was built and a comparison made between webs expressed in terms of number of mines and number of females recruiting to the next generation. 4. All hosts and most parasitoids were recorded in the ®rst year, but new host± parasitoid associations were discovered throughout the sampling period. The diversity of mines and parasitoids was similar across generations. 5. The structure of the parasitoid community was strongly in¯uenced by the host plant. 6. The extent of potential indirect interactions between hosts was assessed by constructing quantitative parasitoid overlap diagrams. The measure of in¯uence used was the importance of one species as a source of parasitoids attacking the other. The greatest potential for apparent competition was between species of host attacking the same species of plant.

Summary. It is well known that feeding by Pieris brassicae caterpillars on cabbage leaves triggers the release of volatiles that attract natural antagonists such as the parasitoid Cotesia glomerata. The temporal dynamics in the emissions of parasitoid attracting volatiles has never been elucidated in this system. In a time course experiment, caterpillar infested leaves attracted the parasitoid within one hour after infestation. At such an early stage of infestation, as much as fifty percent of the parasitoids flew towards the infested plant in a wind tunnel bioassay, while only five percent flew towards the non-infested control plant. Three hours after infestation and later, the response to the volatiles from the infested plant reached its maximum and then continued at a constantly high level for the remaining 14 hours of the experiment. Chemical analyses of volatiles collected from infested leaves at short time intervals during the first 24 hours identified a total of ten compounds, comprising green leaf volatiles, terpenoids, and a nitrile. Significant increase of emission within the first 5 hours following initial herbivory was detected for (Z)-3-hexen-1-ol, (Z)-3-hexen-1-yl acetate, cineole and benzylcyanide. Subsequently, a coupled bioassay-chemical analysis procedure was developed allowing for testing and analyzing the same sample for future identification of the bioactive compounds. This was achieved by using stir bar sorptive extraction for the analysis of solvent extracts of caterpillar-damaged leaves.

Summary 1.Hyssopus pallidus Askew (Hymenoptera, Eulophidae) is a gregarious ectoparasitoid of the two tortricid moths species Cydia molesta Busck and C. pomonella L. (Lepidoptera, Tortricidae). It paralyses and parasitizes different larval instars of both species inside the apple fruit, which leads to the death of the caterpillar. 2. We assessed the influence of host species characteristics and host food on the performance of the parasitoid female in terms of clutch size decisions and fitness of the F 1 generation. 3. A comparison of clutch size revealed that female parasitoids deposited similar numbers of eggs on the comparatively smaller C. molesta hosts as on the larger C. pomonella hosts. The number of parasitoid offspring produced per weight unit of host larva was significantly higher in C. molesta than in C. pomonella , which is contrary to the general prediction that smaller hosts yield less parasitoid offspring. However, the sex ratio was not influenced by host species that differed considerably in size. 4. Despite the fact that less host resources were available per parasitoid larva feeding on C. molesta caterpillars, the mean weight of emerging female wasps was higher in the parasitoids reared on C. molesta . Furthermore, longevity of these female wasps was neither influenced by host species nor by the food their host had consumed. In addition we did not find a positive relationship between adult female weight and longevity. 5. Parasitoid females proved to be able to assess accurately the nutritional quality of an encountered host and adjust clutch size accordingly. These findings indicate that host size is not equal to host quality. Thus host size is not the only parameter to explain the nutritional quality of a given host and to predict fitness gain in the subsequent generation.

While most risk assessments contrast a transgenic resistant to its isogenic line, an additional comparison between the transgenic line and a classically bred cultivar with the same resistance gene would be highly desirable. Our approach was to compare headspace volatiles of transgenic scab resistant apple plants with two representative cultivars (the isogenic 'Gala' and the scab resistance gene-containing 'Florina'). As modifications in volatile profiles have been shown to alter plant relationships with non-target insects, we analysed headspace volatiles from apple plants subjected to different infection types by gas chromatography-mass spectrometry. Marked differences were found between healthy and leafminer (Phyllonorycter blancardella) infested genotypes, where emissions between the transgenic scab resistant line and the two cultivars differed quantitatively in four terpenes and an aromatic compound. However, these modified odour emissions were in the range of variability of the emissions recorded for the two standard cultivars that proved to be crucial references.