The study of the biological potential of forest ecosystems become essential when the objective is rational management, aiming productivity maximization, and avoiding resource exhaustion of the forest site. Forest sector usually apply models to simulate the future of the stand, aiming make appropriated decisions regarding the management to be followed. Among other models, 3-PG was tested for several forest species across different edaphoclimatic conditions, and has been reported unanimously, as a powerful tool to help on forest stands management. The objective of this study was: calibrate, validate and simulate productivity of clonal Eucalyptus plantations using 3-PG model integrated with a geographical information system. Twenty inventory plots were selected containing genetic material C219H (Eucalyptus grandis x E. urophylla) planted on spacing 3 x 3m, even aged, with planting date at 2000, Capão Bonito region, south of São Paulo state. On each plot were evaluated edaphic, dendrometric and nutritional variables on three average trees. Incorporation of relief characteristics resulted in differences on global radiation mapping, mainly during winter, where the effects of relief were elevated, showing up to 147% comparing to flat relief. Number of frosts per month was determined by monthly mean minimum temperature, with a negative linear relation. Most of soil variables evaluated showed a log-normal frequency distribution. At geostatistical analysis, variables related to soil fertility presented higher reaches of spatial dependency at A horizon, indicating higher homogeneity on nutrient distribution at superficial soil layer. The effective CEC showed positive exponential relation with wood volume. Parallel to that curve, we proposed an equation relating the fertility index of 3-PGIS and that soil attribute. Clay content showed strong influence on water levels at field capacity and wilting point, and as consequence, on soil water availability. Stem maximum biomass, observed on forest stands, managed with high technological standards, related directly with clay+silt contents. Same soil property affected wood density. High productivity sites, as consequence, with higher clay+silt contents, resulted in trees with lower wood density. Leaf level of N, Mg, Cu and Zn increased exponentially with clay content, like DBH and foliage biomass. Otherwise, leaf level of P, K, Ca, S, B, Fe and Mn showed opposite relation, decreasing exponentially on clayey soils. Calibration, validation and application of 3-PGIS were obtained for Capão Bonito edaphoclimatic conditions, on clonal stands of the genetic material C219H. Overall, 3- PGIS model showed good performance, mainly on estimative of stem plus bark volume and stem biomass. Simulations of DBH and MAI were dispersing.
