Two tansy-feeding aphids, Macrosiphoniella tanacetaria (MA) and Metopeurum fuscoviride (ME), were studied at a small spatial scale in and around Jena (< 80 km 2 ) using polymorphic microsatellite markers. Both species were found in approximately 60% of sites formerly known to harbour the aphids, although, generally when they did occur, they occurred singly (MA~50%; ME~60%) and rarely together on the same plant at the same time (approximately 10%) and then usually only in the early part of the growing season. This difference may be a result of quasi-apparent competition effects elicited by ants farming ME aphids, and preferentially actively eliminating or disturbing MA aphids. In terms of population genetics, both aphids showed extreme genetic heterogeneity within a metapopulation structure, with ME more than MA (i.e. higher FST values, approximately 0.4 versus 0.15, respectively), and limited levels of interpopulation gene flow. Subpopulations often deviated from Hardy-Weinberg equilibrium and showed linkage disequilibria, as expected in animals with extended parthenogenetic reproduction, and had positive FIS values for most large samples, suggesting inbreeding, and possibly philopatry, certainly in ME. Hierarchical analysis (allele range and number per locus, analysis of molecular variance and FST) strongly suggested that the plant rather than site governs the level of genetic variation. Bayesian clustering analysis revealed that both species had heterogeneous historical genetic patterning, with K (number of subgroups) in the range 3-7. Evidence is also provided from isolation-by-distance and private allele analyses indicating that, in MA, the presence of winged autumn males, absent in ME where males are wingless, influences comparative population genetic structuring, such that ME subpopulations are comparatively more inbred and genetically differentiated than MA subpopulations. Lastly, additional spatial arrangement (ALLELES-IN-SPACE) analysis showed that, in both species, certain subpopulations were genetically isolated from the remainder, probably as a result of geographical barriers, including intervening buildings and woods. As such, the biology of these tansy aphids living in semi-natural habitats is very different from many pest aphid species examined within agro-ecosystems and infesting ephemeral crops. This is because the former appear to be much more reluctant to fly and hence show contrastingly much higher levels of interpopulation divergence, even at small spatial scales as investigated in the present study. Indeed, the number of genotypic clusters found for tansy aphids using Bayesian approaches is similar to that globally for the major pest, the peach-potato aphid, Myzus persicae.

In order to reduce parasite-induced mortality, hosts may be involved in mutualistic interactions in which the partner contributes to resistance against the parasite. The pea aphid, Acyrthosiphon pisum Harris (Hemiptera: Aphididae), harbours secondary bacterial endosymbionts, some of which have been reported to confer resistance against aphid parasitoids. Although this resistance often results in death of the developing parasitoid larvae, some parasitoid individuals succeed in developing into adults. Whether these individuals suffer from fitness reduction compared to parasitoids developing in pea aphid clones without symbionts has not been tested so far. Using 30 pea aphid clones that differed in their endosymbiont complement, we studied the effects of these endosymbionts on aphid resistance against the parasitoid Aphidius ervi Haliday (Hymenoptera: Braconidae: Aphidiinae), host-parasitoid physiological interactions, and fitness of emerging adult parasitoids. The number of symbiont species in an aphid clone was positively correlated with a number of resistance measurements but there were also clear symbiont-specific effects on the host-parasitoid interaction. As in previous studies, pea aphid clones infected with Hamiltonella defensa Moran et al. showed resistance against the parasitoid. In addition, pea aphid clones infected with Regiella insecticola Moran et al. and co-infections of H. defensa-Spiroplasma, R. insecticola-Spiroplasma, and R. insecticola-H. defensa showed reduced levels of parasitism and mummification. Parasitoids emerging from symbiont-infected aphid clones often had a longer developmental time and reduced mass. The number of teratocytes was generally lower when parasitoids oviposited in aphid clones with a symbiont complement. Interestingly, unparasitized aphids infected with Serratia symbiotica Moran et al. and R. insecticola had a higher fecundity than unparasitized aphids of uninfected pea aphid clones. We conclude that in addition to conferring resistance, pea aphid symbionts also negatively affect parasitoids that successfully hatch from aphid mummies. Because of the link between aphid resistance and the number of teratocytes, the mechanism underlying resistance by symbiont infection may involve interference with teratocyte development

Within insect communities, the population ecology of organisms representing higher trophic levels, for example, hymenopterous parasitoids, may be influenced by the structure of their insect hosts. Using microsatellite markers and ecological data, we investigated the population structure of the specialist braconid wasp parasitoid, Lysiphlebus hirticornis Mackauer attacking Metopeurum fuscoviride, a specialist aphid feeding on tansy, Tanacetum vulgare. Previous studies revealed that M. fuscoviride has a classic metapopulation structure with high subpopulation turnover. In this study, up to 100% of ramets within a host plant genet colonized by aphids were colonized by the parasitoid, yet plants with aphids but no parasitoids were also observed. Genetic differentiation measured by F ST , actual differentiation (D) and relative differentiation (G ST ) indicated highly structured parasitoid population demes, with restricted gene flow among and between parasitoid subpopulations at the various sites. Interestingly, both field data and population assignment analysis showed that the parasitoid is highly philopatric. Thus, despite the frequent local extinctions of the aphid host, the parasitoid continuously exploits its aphid host and contributes to the demise of local aphid subpopulations, rather than spreading its genes over many aphid populations. F ST values for the haplodiploid parasitoid were similar to those found in an independent study of the diploid aphid host, M. fuscoviride, hence supporting the view that an insect herbivore's population structure directly influences the ecology and genetics of the higher trophic level, in this case the wasp parasitoid.