Variabilidade espacial da produtividade e estrutura arbórea de castanhais nativos na Amazônia oriental

Abstract

The brazil nut tree (Bertholletia excelsa Bonpl) is a non-timber forest resource more relevant in the south America, the species has great ecological, economic, social and cultural relevance. In southern Amapá, the species can be found in dense forests (FD) and, mainly, in open forests (FA). The forest typologies of occurrence are determinant for the growth and establishment of the brazil nut trees, because of the different availability of light and the involuntary management carried out by extractivists. The natural and human influence can modify the brazil nut trees structure, leaving a spatial signature in the landscape. Geostatistics is a tool that can help in the detection of this spatial component in the brazil nut stands, being able to express the relationship between variables and places of occurrence. Therefore, the dissertation main point was to verify the spatial dependence existence of tree and productive variables in native Amazonian brazil nut stands, comparing the responses between two forest typologies. The study was carried out in the extractive reserve of the Cajari River, located in the southwest of Amapá. Data from the 100% inventory carried out by Embrapa were used. Dendrometric and productivity variables were extracted from this base. Variography was performed for the variables and, later, the adjustment and selection of models widely used in the literature. After the models selection, the ordinary kriging was used for interpolation in thematic maps. Verifying the relationship between the variables and productivity, the relative deviation coefficient (CDR) was used. The brazil nut populations presented platycuttic distributions and tree concentration in intermediate classes of diameter (50cm≤ DAP ≤ 150 cm) in typologies. In AF there was higher concentration in the class 50cm-100cm, while in the DF it was in the 100cm-150m class. In both typologies, the variables were classified as moderate and strong spatial dependence. However, in AF there was a greater range of spatial structure variation, while in DF the range fell in half, revealing the strong presence of random factors. In kriging maps, the highest productivity follows the areas with the densest chestnut snot in AF, while the average diameter is larger in areas with lower density. In FD, the maps did not present a pattern, the largest classes seem to always occur in specific locations, on the edges of the stain. The CDR indicated that both in AF and DF, density and basal area are the most similar maps in relation to productivity. This indicates that the productivity of brazil nut trees in the eastern Amazon depends more on the density of brazil nut trees than on their individual productive capacity. The type of distribution observed is characteristic of brazil nut trees with intense seed collection and human activities. The spatial dependence found in AF indicates that the environment, with a lot of direct lighting and open areas, favored the development of the species in times of the future. In DF, extractive practice can modify the structure of the forest and positively disturb the growth of the brazil nut tree. Human disturbances caused by the opening of bites, removal of wood, involuntary dispersion of fruits, cleaning of the base of trees, may have benefited the chestnut tree, especially in a dense forest environment.