Caracterização pedológica associada ao mapeamento digital de solos por geoestatística e espectroscopia na região do Parque Nacional de Anavilhanas (AM)

Abstract

Knowledge of soil properties generates data and can help in land use planning, conservation, production and decision-making in ecosystems and the social area. The Amazon biome is one of the most biodiverse in the world, covered by great geodiversity in a region of difficult access and little infrastructure, which leads to the need for more detailed and updated soil surveys. Therefore, the present study aimed to evaluate the potential of new geotechnologies such as spectroscopy for soil characterization and geostatistics for mapping attributes in the Anavilhanas National Park (PNA), in the state of Amazonas, northern Brazil. For this, fifteen sites covering different soils were selected according to a legacy pedological map, where 43 soil samples were collected at 20 cm depth range up to 100 cm depth. These samples were submitted to physical analysis (sand, silt, clay), chemical analysis (silicon, aluminum, iron and titanium), mineralogical and spectral analysis by near remote sensor in the region 350-2500 nm of the electromagnetic spectrum. Initially, the data were analyzed using Pearson's coefficient, through which the degree of association between soil properties and relief was observed. Subsequently, the data that showed a certain degree of significance were submitted to descriptive statistical analysis and verification of Normality. Then, a geostatistical analysis was carried out, and semivariograms were constructed for the physical-chemical attributes and relief covariates, and the best models were selected for interpolation. The attributes sand, silt, clay, Si, Al, Fe and Ti, together with the relief covariates slope, vertical curvature and shaded relief showed spatial dependence. Soil data at depth 0-20 cm and relief covariates were mapped by Ordinary Kriging using QGIS 3.16. Kriging soil maps were grouped with k-means and resulted in 05 pedological units characterized by soil type, texture, chemistry and mineralogy, as well as relief characteristics. Spectral data showed the presence of kaolinite and iron oxide hydroxides in the studied soils. Geostatistics and spectroscopy help in the characterization of soil attributes and mapping of units via geotechnologies and can subsidize other initiatives in regions of difficult access and lack of data.