6533b7d7fe1ef96bd1268206

RESEARCH PRODUCT

Processus écologiques et évolutifs influençant la colonisation de l'ambroisie à feuilles d'armoise (Ambrosia artemisiifolia l.) en France.

subject

description

Understanding of the mechanisms behind the success of the invasive species is essential to manage current biological invasions and to prevent the risks of the futures ones. Using a conceptual framework integrating ecological and evolutionary processes, this work aimed to analyse the factors of the common ragweed colonization in France. First of all, the study of biotic and abiotic interactions has shown the ability of common ragweed to tolerate herbivory and water stress. Common ragweed is able to buffer defoliation through an efficient compensatory growth with no consequence on the reproduction. Herbivory tolerance has been maintained in introduced populations even if herbivory pressure is low in the introduction area. Under water stress conditions, common ragweed displayed higher root biomass than the other species present in the communities it invades. In addition, differences in mean trait values suggest different niche occupation by common ragweed compared with its companion species. Common garden experiments comparing native populations vs. invasive isolated and invasive core populations have shown that adaptation ability is mainly due to phenotypic plasticity rather than trait differentiations. Quantitative genetic studies suggest a high evolutionary potential for germination traits. Morphological, phenological and physiological traits exhibit lower standard genetic variation and lower heritability and thus a lower evolutionary potential. However, variation in reaction norms suggests a high evolutionary potential for phenotypic plasticity. Herbivory and water stress tolerance are factors that potentially increase the ability to colonize à large range of habitats. Furthermore, phenotypic plasticity and evolutionary potential may also favour an increase or a shift in species ecological niche and hence may favour the distribution range expansion