Raul Sanchez Vicens

Spatial Patterns of Farmland Abandonment in Rio de Janeiro State

While land use change in Brazil is characterized by strong cropland expansion, the federal state of Rio de Janeiro, located in Southeast Brazil, shows an opposite trend: a significant amount of cropland has been abandoned. The temporal and spatial distribution of farmland abandonment in Rio de Janeiro state was analyzed at the municipal level from 1991 to 2013. Developments differ strongly between the different regions. The Norte Fluminense Region, which holds the highest share of cropping area in the state, showed the highest cropland abandonment. This decrease was mainly due to a reduction of the areas used for the cultivation of banana, maize, and sugarcane. In the Baixadas and Metropolitan regions, the reduction of areas for orange plantations played an important role, while the reduction of areas for coffee production was important for mountainous areas. Another transition seems to be the substitution of sugarcane cultivation with pineapple cultivation. Areas for coconut crop production increased mainly in the Quissama municipality. Areas for manioc crop production remained stable throughout the analysis. The patterns identified in this paper will provide an important background for the policymakers in implementing spatially explicit plans for the agricultural sector.

Study of the Floodflow Dynamics in the Pantanal of Cáceres/MT.

The Brazilian Constitution lists some biomes and ecosystems considered as national patrimony, among which the Pantanal Matogrossense is inserted, legally establishing conditions that ensure the preservation of the environment, including the use of its natural resources (Article 225, § 4). In this context, the research proposes to analyze the dynamics of the floodflow in the Pantanal of Cáceres/MT and aims to contribute a methodology to obtain results that can support the planning, management, and monitoring of natural resources, as well as providing important information for agriculture, geology, hydrology, and ecological models. In this research, images from the MODIS sensor from the year 2015, with a resolution of 250 m, were used for the application of the methodology. The software ArcGis and Matlab were used for the image processing. The normalized difference water index obtained through 500 m resolution MODIS images was used as a priori analysis of the regions with water and without water, after a threshold was defined for all of the images and the processing was done to obtain the matrix with days with water and without water in the studied area. The results obtained provide a classification for the floodflow, the quality parameters derived using study area show that the proposed method performed better with classes < 30 (96%) and 180 270 (90%) of 4% and 10% of false positives, respectively.