Many different techniques are used today with different objectives to model species geographic distribution. Some of these techniques are based on environmental analysis where algorithms search for environmental conditions similar to those where the species are known to occur, resulting in potential areas where the environmental conditions would be favorable for the growth of these species. The present study is based on predictive modeling of species geographic distribution using genetic algorithm as a tool for plant conservation in modeling the "cerrado" biome of the State of São Paulo. The methodology used and the results obtained were considered satisfactory for the generation of geographic distribution models for plant species, based on abiotic data for the regions studied. The accuracy of the model in predicting the occurrence of species in the "cerrado" is greater if only points that occur within the "cerrado" and not in transition areas are used. In order to minimize problems due to the lack of convergence of the genetic algorithm GARP ("Genetic Algorithm for Rule Set Production"), 100 models were generated for each specie. The use of modeling tools can help understand the distribution patterns of a biome or ecosystem in a regional analysis.
biodiversity modeling; genetic algorithms; plant species