The spatio-temporal dynamics of plant invasions

点击翻译按钮获取中文摘要

A large number of ecologists have studied plant invasions, but very few have concentrated their efforts on the initial stage of dispersal and subsequent spread of invaders. Consequently, it is still extremely difficult to predict the spread of invaders. The objectives of my research program are to study the respective contribution of natural (rivers) and anthropogenic (roads) structures as dispersal corridors for exotic plants, and the long-term impacts on plant diversity of an invasive weed that uses rivers and roads to spread. I will focus my research efforts on the invasive plants of southern Québec because this region is particularly vulnerable to invasions, it has been abundantly sampled for herbarium specimens, and I have already developed a model to reconstruct the spread of plants in southern Québec. I will reconstruct, with the help of empirical data, the spread of a large set of exotic species to characterize general invasion patterns along a major invasion pathway (Richelieu River). I will also develop a model facilitating the reconstruction of the dynamics of plant invasions in a landscape strongly affected by human activities, and particularly by the road system. I hypothesize that a high number of exotic species that were previously restricted to natural corridors (rivers) became invaders with the improvement of the road network, and during or following the extension of the highway network. To test this hypothesis, I will mainly use herbarium specimens as information sources. Finally, I will describe and understand the long-term dynamics of a purple loosestrife invasion using paleoecological tools. I hypothesize that a purple loosestrife invasion consists of a first step, in which the plants form very dense stands, and of a second step, in which colonies decline and allow the re-establishment of native plants. The spatio-temporal models that will be created represent a major and unique effort to understand the spread of invasive plants along corridors, especially in a world heavily impacted by human activities and road systems. These retrospective models are extremely important in the modelling process because they are essential for a better understanding of the dynamics of invasions and for testing predictive models.