Potencial agrícola e usos dos solos na microrregião de Caxias - Maranhão - Brasil

• Ano de defesa: 2022
• Autor(a) principal: JANSEN, Ludimila Nathasha da Silva
• Orientador(a): FARIAS FILHO, Marcelino Silva
• Banca de defesa: FARIAS FILHO, Marcelino Silva , CAMPOS, Milton Cesar Costa , RODRIGUES, Taíssa Caroline Silva
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Maranhão
• Programa de Pós-Graduação: PROGRAMA DE PÓS-GRADUAÇÃO EM GEOGRAFIA (PPGGEO)
• Departamento: DEPARTAMENTO DE GEOCIENCIAS/CCH
• País: Brasil
• Palavras-chave em Português: potencial agrícola; uso do solo; Google Earth Engine; solos de Caxias; Cerrado.
• Palavras-chave em Inglês: agricultural potential; land use; Google Earth Engine; soils of Caxias - MA.
• Área do conhecimento CNPq: Conservação do Solo; Ciências Agrárias
• Link de acesso: https://tedebc.ufma.br/jspui/handle/tede/tede/4544

The adequate use of the soil enables the development of more stable and harmonious productive systems and avoids or minimizes the chances of major negative environmental impacts. In order to achieve the best use of the soil it is necessary to know its attributes, potentials and limitations in order to define the appropriate activities for each particularity. The microregion of Caxias, located in the east of the state of Maranhão, has soils with significant variability, with different fertility and use capacity and has areas with significant forestry and agricultural activities. In this sense, this research analyzed the agricultural potential of the micro-region and its uses, based on the soil data described in the Economic Zoning of Maranhão - ZEE, totaling 18 profiles in the study area, aiming to indicate the most appropriate types of management for each of the environmental conditions present there. For this, we used the SAAT methodology, based on management levels for different uses, which is defined by limiting factors: oxygen deficiency, fertility deficiency, water deficiency, susceptibility and impediments to mechanization, with the addition of parameters (Nutrient availability, Aluminum toxicity, Phosphorus fixation) proposed by Pereira and Lombardi Neto (2004) and for the mapping of land use and land cover the Random Forest machine learning model was used as an aid, on the Google Earth Engine platform and later the processing, crossing and analysis of the phenomena that happened in the study area with the use of QGIS software. As results, it was found that 47% of the soils in the micro-region of Caxias are more suitable for natural pasture at management level A, with the possibility of good aptitude for plantations at management level C, followed by good aptitude for regular pasture. Based on the results achieved from the Google Engine land use and cover classification, 42% of the soils in the study area are zoned for pasture, with growing forestry. The areas of conflicts showed that on average 40% of the activities developed in the micro- region are suitable for agricultural aptitude, 37% of the areas are under-utilized and 22% present over-utilization of the soils regarding agricultural aptitude. The Random Forest proved to be effective for the activity of mapping the use and cover, achieving a good precision (96%) with a satisfactory classification, according to the results of the Kappa Index. The dynamics between use and cover and agricultural suitability of the soils present in the study area indicates the requirement and attention to the management for the conservation of these systems.