Rapid Plant Invasion in Distinct Climates Involves Different Sources of Phenotypic Variation

anglais. peer reviewed

anglais. When exotic species spread over novel environments, their phenotype will depend on a combination of different processes, including phenotypic plasticity (PP), local adaptation (LA), environmental maternal effects (EME) and genetic drift (GD). Few attempts have been made to simultaneously address the importance of those processes in plant invasion. The present study uses the well-documented invasion history of Senecio inaequidens (Asteraceae) in southern France, where it was introduced at a single wool-processing site. It gradually invaded the Mediterranean coast and the Pyrenean Mountains, which have noticeably different climates. We used seeds from Pyrenean and Mediterranean populations, as well as populations from the first introduction area, to explore the phenotypic variation related to climatic variation. A reciprocal sowing experiment was performed with gardens under Mediterranean and Pyrenean climates. We analyzed climatic phenotypic variation in germination, growth, reproduction, leaf physiology and survival. Genetic structure in the studied invasion area was characterized using AFLP. We found consistent genetic differentiation in growth traits but no home-site advantage, so weak support for LA to climate. In contrast, genetic differentiation showed a relationship with colonization history. PP in response to climate was observed for most traits, and it played an important role in leaf trait variation. EME mediated by seed mass influenced all but leaf traits in a Pyrenean climate. Heavier, earlier-germinating seeds produced larger individuals that produced more flower heads throughout the growing season. However, in the Mediterranean garden, seed mass only influenced the germination rate. The results show that phenotypic variation in response to climate depends on various ecological and evolutionary processes associated with geographical zone and life history traits. Seeing the relative importance of EME and GD, we argue that a “local adaptation vs. phenotypic plasticity” approach is therefore not sufficient to fully understand what shapes phenotypic variation and genetic architecture of invasive populations.