I hypothesized that in hyper-fragmented landscapes tree assemblages become largely impoverished and homogeneous at landscape level by a non-random extirpation of ecological groups of trees. Basically, seed, seedling and adult tree assemblages will be described in twenty forest fragments ranging 3-3,500 ha based on (30x30m = 0.09 ha plots).

To assess immigration rates I will follow seed rain and seedling recruitment in a monthly basis during at least one year and compare the list of adult trees with that of seeds and seedlings collected in the centre of the 0.09ha plots. Seed and seedling species not represented as adult trees will be considered as immigrant. By these means I will be able to know what group of species can maintain movement of seeds and therefore seedling recruitment that permit the maintenance of their populations and colonization abilities and what groups of tree species do not.

The next step is to generate guidelines for management of such fragmented landscapes based on tree species enrichment focused on species with reduced populations and propagule movement in forest fragments.