The project aimed at contributing to understand the ecological and genetic mechanisms underlying the possibilities of establishment of exotic insects in the genus Megastigmus (Hymenoptera: Torymidae), specialized in the exploitation of forest seeds, and at defining suitable methods of monitoring and management of the invasive insect populations. For that, we intended to compare two distinct models host seed/ Megastigmus, crossing different first date of invasion and taxonomic structure of the forests: (i) true firs (Abies)/Megastigmus where the resource, indigenous, is exploited by a native species, M. suspectus, in competition with at least 5 invasive species of North American origin, and (ii) true cedars (Cedrus)/Megastigmus where the resource, itself exotic, is exploited by 2 invasive species without native candidates. In a large seed sampling, involving 169 Abies stands in France and 55 in other European countries (Germany, Belgium, Bulgaria, Denmark, Spain, Greece, Italy, Norway, Poland, Romania, Switzerland, Tchéquie) and from the Middle East (Lebanon, Turkey), 90% of the seed lots presented chalcid damage, with a rate of attack higher than 50% in ca. 1/3 of them. On the whole, 5 invasive species coming from North America were observed (M. rafni, M. pinus, M. milleri, M. specularis, and M. lasiocarpae) beside the native species M. suspectus. At present, exotic chalcids largely predominate in Western Europe, displacing the native seed chalcids. M. rafni appears to be the most important species especially in natural fir stands with a much larger range than in the 1990s. In a similar way on Cedar, the exotic species M. schimitscheki, largely expanded since its introduction in the early 1990s and it supplants from now the other species introduced since a longer date (years 1950), M. pinsapinis, who was the only species observed in the years 1980. Several biological features likely to be implied in the apparent competitive superiority of both M. rafni and M. schimitscheki were examined in comparison to these of native species and dominated exotic species: length of female ovipositor, capability of developing in unpollenated seeds, parthenogenesis, phenology of adult emergence, prolonged diapause. These elements appear relatively contradictory between them to explain the statute dominating of M. rafni, at least. For instance the species has the longest ovipositor, allowing it to oviposit in all cone species independently of cone size, and is capable of developing until adult in unpollenated seeds. However, it is the latest species to emerge and it is not a thelytokous species conversely to the native M. suspectus. Bag tests were carried out using simultaneously or not the different species to look for resource sharing but the results will be available only during summer 2006 following adult emergence. On cedar, the same bagging experiments showed a division of the niche between M. schimitscheki and M. pinsapinis. Both prefer the apical part of the cone in non-competition situation but when the two species are placed in the same bag, their attacks added on the apical part rather than distribute on the cone. When insects from North America were compared with these of Europe using mtDNA sequencing (cyt b), it appeared with no doubt that the individuals present in Europe belong well to North-American species, in particular M. milleri who was regarded as absent from Europe until the end of 1990s. This study in addition showed that the species M. pinus, presents in many European sites, has two haplotypes very different, suggesting two cryptic species. A phylogeographic study of M. rafni using both mitochondrial and nuclear markers EF-1alpha elongation gene, ITS2) showed the existence of two distinct clades separated by strong genetic distances (3.5%), suggesting different routes of invasion. A set of 6 microsatellite markers was used to analyse the genetic structure of the populations of M. rafni in the areas of origin and invasion. The results suggested an absence of structure according to the fir host species, but a clear differentiation of the Danish populations, which have a very low variability, suggesting a probable bootleneck. On Cedrus, 5 loci microsatelittes were used to study the genetic structuring of the populations of Megastigmus schimitscheki. Populations of the native zone (Cyprus, Turkey) were highly polymorphic and spatially structured, and largely differentiated from those of the introduction zone of southeastern France, fewly polymorphic. We suggested a strong bottleneck connected to a recent and limited introduction, even limited to only one origin. Only two of these populations show a stronger differentiation however, and are both located on the current face of expansion in direction of the Alps of the South. The invasive history of the exotic species on Cedar and fir trees thus appears largely different. The additional impact of the invasive exotic species on the potential of natural regeneration of the European fir plantations is quantitatively consequent. Whereas in the years 1980, seed damage by M. suspectus very seldom exceeded 15%, percentages of attack higher than 50% are not rare any more (17.6% of the cases) due to the addition of the exotic species. On the other hand, the qualitative impact on the genetic pool of the fir plantations seemed much more limited. In particular, self-fertilized fir trees were equally infested as cross-fertilized trees. On a more general level, the taxonomic closeness of the native trees with respect to original host of the exotic insect seemed to constitute a relevant criterion to estimate the probability of establishment of exotic insects associated to forest trees. We also developed simple and cheap methods for monitoring insect infestation in seed lots at importation, based on differential flotation in solvents related to seed density. A booklet will be published gathering these data. We expect it could help in defining regulations at importation of forest seeds whereas such regulations do not exit in th European Union yet, and thus favour invasion of exotic seed insects