Ocorrência e distribuição espacial de açaizal no contexto do uso da terra em áreas nativas na Amazônia oriental.

Abstract

The Amazon estuary is formed by an extensive hydrographic network and islands covered by floodplain forests, which provide the açaí palm tree with ideal growth conditions. The mapping and determination of the distribution of the natural açai palm (Euterpe oleracea Mart.) massifs in the floodplains of the Brazilian Amazon has always been a challenge both in assessing the expansion potential and in measuring the production of the crop in the region. Thus, the aim of this study is to map and analyze the most representative açaí-producing floodplain system on the islands of Igarapé-Miri (Jarimbu, Mamangal, Itaboca, Mutirão and Buçu), using remote sensing and geoprocessing techniques, in order to characterize the dimension and spatial distribution, as well as evaluate changes in the landscape and cultivars, correlating the occurrences with the physicochemical properties of lowland soils. Thousands of riverine families live on these islands, who practice the extraction of açaí, developing their own forms of management. The set of studies developed here involves the mapping of land use and land cover with the identification of açai grove patterns, as well as their spatial distribution in the midst of native forests. For this purpose, the unsupervised classification method, ISODATA, was used, with efficient identification and zoning of seven classes of land use and vegetation cover: Hydrography, Exposed Soil, Urban Area, Alluvial, Lowlands, Arboreal and Agriculture. Massifs were identified in the Alluvial and Lowland classes. The islands with the highest percentage of the Alluvial class are Mamangal, Buçu and Jarimbu with 70.3%, 69.2% and 63.7%, respectively. The soils mapped in all islands showed silty clay loam and silty loam texture, in addition to high base saturation (above 50%), high organic matter content and significant presence of potassium and phosphorus. Through the spatiotemporal analysis it was possible to observe the variations of some classes over seven years. The dynamics of use and coverage was defined by images from the RapidEye (5m) and Planet (3m) satellites, taken in 2013 and 2019, respectively. Advances in the exposed soil class of more than 50% were observed for the Jarimbu, Mutirão and Itaboca islands, the latter responsible for more than 50% of expansion in the urban area class. In the five islands, the alluvial class, responsible for the areas with açais, emerged in the last seven years with an increase in these areas at the expense of the reduction in the arboreal class. For data validation, the confusion matrix was used, with mean accuracy of the islands evaluated as very good, mean global accuracy 77.74% and mean Kappa index 0.73, indicating strong agreement with the reference data and classification. The potential of remote sensing to assess variations over time produced accurate maps of changes in floodplain forests, proving to be effective in helping to monitor and manage biodiversity.