Modelagem da cobertura da terra e análise da influência do reflorestamento na transformação da paisagem: Bacia do Rio Piracicaba e Região Metropolitana do Vale do Aço

Abstract

Land cover change has been causing environmental impacts on global scale. Those changes become the focus of some researches, towards understanding the dynamics of these changes and their consequences for biodiversity conservation, climatic changes and landscape changes. In this context, the LUCC models Land Use and Cover Change have wide application, once through their representations is possible that modelers build future scenarios simulations. The cities in the Piracicaba river basin and metropolitan area from Vale do Aço - RMVA, are home to some industries from steel sector and cellulose and paper in Minas Gerais state. This region is one of the major hub of economic development in the state. However, this economic development based mainly in the steel industries, paper and cellulose, has brought changes in land cover. Among these, the replacement of natural vegetation cover areas for reforestation with eucalyptus has been impacting the landscape and contributing to biodiversity losses. In order to measure the changes in land cover and landscape structure, this research employed the land change modeler model (LCM) to analyze, to model and to simulate changes in land cover from land cover data from 1985, 2010 and 2013. These results, plus the LCM simulation to 2035 were used in the landscape structure analysis for the years 1985, 2010 and 2035. The landscape structure analyze, was performed through metrics that were used to measure changes in the landscape elements, and based on this analysis assess the degree of landscape fragmentation. The results of land cover change analysis showed an increased by 11,64% on reforestation with Eucalyptus areas, whereas forest and pasture areas decreased by 9,15% and 3,55% respectively, between 1985 and 2010. Explanatory variables for potential transition sub-model calibration were tested. The higher explanatory variables related to observed changes , according to Cramer´s-V test were: 1)changed areas between 1985 and 2010, 2) distances from urban centers; 3) distance from areas with eucalypt reforestation; 4) distance to pasture area. Using a set of environmental and socioeconomic explanatory variables was possible to calibrate the potential transition sub-model based on a neural network approach with an accuracy of 94.28 %. With the calibrated land cover the simulation to 2035, the trend found showed that reforestation with eucalyptus will may keep increasing. The landscape analysis showed the matrix of the landscape is changing. While in 1985 the largest patches were forested areas, in 2010 the largest patches became eucalyptus reforestation. This change reflects on the connectivity and isolation of fragments. The simulation showed a tendency that forested areas become increasingly isolated, while the reforestation areas become connected. In conclusion it is clear the necessity of balancing economic growth and environmental conservation. We need to think the socioeconomic benefits of reforestation against their environmental impacts. If eucalyptus demand keep on grow, such as observed between 1985 and 2010, it will be necessary to establish limits or ways to improve productivity without compromising environmental quality as an alternative to keep the most stable dynamic equilibrium of the environment.