One of the consequences of habitat fragmentation is the increa- sing isolation of the resulting habitat patches. The latter pre- vents plant species from migrating between fragments by seed dispersal. Particularly forest plant species, with their often limited dispersal capabilities, are sensitive to this process. A better knowledge of the colonization patterns of forest plant spp, of their dispersal distances and the factors effecting dis- persal is essential to save these species from extingtion in our dynamic landscapes. Particularly longdistance dispersal remains understudied. This study aims at predicting colonisation patterns of forest plant species in a fragmented landscape using quantita- tive migration models, like the ones of Neubert & Caswell (2000) and Higgins & Cain (2002). These models will be calibrated and validated with data from local demographic and regional dispersal processes. Forest policy relevance for forest expansion.