Análise por sensoriamento remoto de áreas de açaizais em florestas de várzea no município de Mazagão (Amapá)

Abstract

The Amazon estuary is formed by a maze of islands and adjacent regions, where there are Lowland forests, areas of special importance due to the high values of productivity and soil fertility. Among the most important wood products not explored these areas highlight the açaí (Euterpe oleracea Mart.), One Arecaceae, which produces edible fruit, which is extracted from the pulp of the fruit and the palm heart. Recent years have seen a major expansion of açai as a result of açaí market growth in the state of Amapá. Such expansion can be explained both by the planting of new areas such as the management of native palm heart areas, resulting in increased density of acai individuals and reduce individuals of other forest species. The mapping through remote sensing images, can facilitate the further evaluation of the areas of wetlands, taking into account the productive potential of the same. This study aimed to test a methodology to identify areas of thickening of açai, using medium and high spatial resolution images. The study area was the river mouth Mazagão a lowland area located in the southeast of the city of El Jadida, the state of Amapá, in the Amazon River estuary. For the study used two optical images of middle and high spatial resolution, Landsat 8 / OLI and RapidEye, respectively, considering the lower cloud cover as possible. For each image was the Vegetation Index (NDVI) for characterization and identification of vegetation pixel values in 10 points in lowland forest area with high density of açaizal (PAA) and 10 points in areas without thickening of açaizal ( PFN). A point on the Landsat image corresponded to 36 points in the RapidEye image. Data processing was performed in ENVI 5.0, ArcGIS 10.1 and R 3.1.1 for analysis of descriptive statistics generated between the pixel values. For Landsat 8 / OLI average NDVI values were 0.48 ± 0.03 for PAA and 0.51 ± 0.02 for PFN (p <0.05, n = 10). For RapidEye image values were 0.57 ± 0.08 for PAA and 0.61 ± 0.06 for PFN (p <0.0001, n = 360). With a proven methodology was possible to identify areas of dense palm heart areas both in medium resolution images (Landsat 8 / OLI) as the high-resolution images (RapidEye)