We are investigating the role of disturbance history, source-sink metapopulation dynamics, and genetic adaptation to different canopy environments in the invasion success of Alliaria petiolata in the New England understory. In these studies, we employ a combination of population-level experiments, metapopulation studies, and landscape-level historical analyses. We have initiated a statewide survey of forested locations to determine the presence and absence of garlic mustard with respect to major pathways of invasion via past and present habitat disturbances. More explicit spatial analyses will test whether past sites of open canopy could have acted as corridors for the spread of invasive populations to present locations. This broad analysis will enable us to evaluate the generality of emerging conclusions from our population-level studies for predicting the broad-scale factors controlling the distribution, pattern and abundance of exotic species. At the population-level, we have begun long term demographic modeling of A. petiolata and experiments to test for habitat-specific natural selection on physiological, phenological, and allocational traits in sites with different canopy structure (edge or understory). In long term demographic analyses, we will use matrix population models to determine the relative contributions of subpopulations in different habitats to forest invasion. In ongoing reciprocal transplant studies, we will asses whether the maternal source habitat of a propagule contributes to its germination and survival in contrasting habitats.