The software DINAMICA was initially conceived for the simulation of the Amazon Landscape dynamics, particularly the landscapes evolved in areas occupied by small colonists - farm properties smaller than 100 hectares. This model aims at incorporating the decision rules used by the colonists to choose and cultivate their plots of land, and eventually abandon them for forest regeneration.

With new propositions, the DINAMICA model involves a multiple time step stochastic simulation with dynamic spatial transition probabilities calculated within a cartographic neighborhood. The model employs special functions designed to reproduce on a map the actual size of landscape patches and it applies logistic regression to indicate the areas most favorable for each type of transition from environmental data obtained mainly from satellite imagery.

The software DINAMICA models the changes in these Amazonian landscapes, according to a simplified view, through the states and paths as shown in fig. 1. With respect to this conceptual model, it was defined four landscape elements and six transitions, which can be simplified only to three transitions, because the others transitions (regeneration to secondary forest, and secondary forest to mature forest) are considered to be deterministic, based only on the sojourn time. In this way, the model will only deal with the changes between forest to the deforested areas, deforested areas to regeneration, and new deforestation occurring from the regeneration patches.

In effect, this taxonomic simplification does not compromise the model, considering that most of the study area, where the model was initially built and tested, consists of pastures in several stages of use and abandon. Moreover, the model was not conceived to describe the current land uses in these regions, but the spatial configuration arising from the forest fragmentation and the forest regeneration of the abandoned areas. In this manner, the model deals with the effects of the deforestation and regeneration processes to the fragmentation and recovering of the original forest habitats. Using this approach, the following issues can be addressed to the model.

• What deforested areas will become secondary forests?

• What will be the spatial configuration of the landscape elements in a given future time?

• And what will be the ecological consequences of that landscape structure?

• Where will the forest fragments remain?

• How can an architecture of colonization project influence the development of a landscape, especially forest fragmentation ?

Fig. 1 Simplified model of landscape changes in the study area. The dashed lines indicate that if the regeneration process succeeds, one will find after a certain time a secondary forest with ecological characteristics similar to the original forest.