| dc.description.abstract | Historical, socioeconomic, cultural and natural changes determine, over time, changes also at the landscape scale everywhere in the world. These changes are dynamic, have occurred in large numbers in recent decades and thehuman are largely responsible for these disturbances on the landscape that may have consequences both in spatial extent and duration of time. In northeastern Rio Grande do Sul (RS), southern Brazil, the landscape has undergone many changes since the century XIX and today special attention to the forest and grassland ecosystems of the region is required. The objective of this study was to quantify, describe and evaluate changes in the landscape of northeastern RS, and especially the dynamic mosaic of forest and grassland ecosystems in the region of Campos de Cima da Serra (CCS) in the Araucaria Plateau, on the use of soil over a period of 21 years. Satellite images of 1984, 1994 and 2005, obtained through Landsat 5 TM, were classified into six classes of land use. From these images landscape metrics and NDMI moisture index were extracted to characterize the dynamics of the landscape in the region during the studied period. The area was also classified according to geoeconomic regions covering the region. From the data of geoeconomic regions, years and the respective classes of land uses and related areas, a Log-linear analysis was performed. To evaluate the changes in the index NDMI over time, a G test was applied, listing the 15 levels of moisture generated in their respective areas for each year. Using a Log-linear analysis the relation between the areas represented by the moisture levels with their own levels of moisture, the years and the rainfall average of 12 months prior to satellite images analyzed was tested. A map overlaying agricultural and grassland areas according to the moisture level was done in order to visualize the direction of expansion of this class. The analysis showed that different classes of land uses increased and contractions were different in different geoeconomic regions over time (χ2 = 89120,612, df = 40, p = 0.0). Test G (χ2 = 592875,970, df = 26, p = 0.0) was significant, showing wetlands variation over time. Log-linear analysis rainfall interacted in areas corresponding to the variation of moisture levels over time (χ2 = 4,421E +10, df = 89, p = 0.0). The agricultural areas have expanded preferentially in the areas with the highest levels of humidity. The contrast between natural and anthropogenic areas increased, and agricultural areas presented great growth whereas field and forest areas reduced, becoming more fragmented. The results showed that the northeast region of the RS hads a major impact on land uses and cover over the studied period and that areas with more moisture were converted into agricultural expansion. | en |
