To develop a predictive model of carbon stock and their dynamics for the Amazonian forest • Collaborative research effort between Brazil’s National Institute for Amazonian Research (INPA) and the Museum of Natural History of the Smithsonian. • Use Landsat TM images of central Amazonia to classify landscape • Combination of field studies and computer modeling to estimate biomass of above ground biomass • Land-use history after forest clearing has deterministic effect on forest regeneration trajectory • Herbivores have strong negative impact on regeneration of degraded lands • Logging strategies in second growth forests enhance survival of commercially important trees • Provides baseline data for developing carbon trading with Annex 1 countries as suggested in Kyoto protocol