Carbon geomicrobiology, saturation deficit and sequestration potential of Brazilian agricultural soils

• Ano de defesa: 2024
• Autor(a) principal: Rodríguez Albarracin, Heidy Soledad
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Atividade enzimática; Carbon sequestration; Digital soil mapping; Enzyme activity; Espectroscopia do solo; Mapeamento digital do solo; Mineralogia; Mineralogy; Sequestro de carbono; Soil spectroscopy
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/11/11140/tde-12042024-152725/

The ecosystem service of climate regulation provided by soil is due to its capacity to sequester C, and soil organic carbon (SOC) is essential for its health. The capacity of the soil to retain OC depends on the minerals and their interaction with the microbiota. Chapter 1 of this work analyzes the potential for COS sequestration in the clay fraction for soils in the Piracicaba region, in the state of São Paulo, based on an equation for the potential C saturation deficit of fine soil particles, adjusted for tropical agricultural soils. This potential was adjusted using a spatial regression model. In the surface layer, the sequestration potential is mainly explained by the relative abundance of kaolinite, hematite, goethite and gibbsite determined by Vis-NIR-SWIR spectroscopy. A direct relationship was observed with goethite and gibbsite. At a depth of 80 to 100 cm, kaolinite and hematite were responsible for the greatest variation in sequestration potential. The contribution of each mineral to COS sequestration potential was also mapped, with high contributions from goethite and gibbsite in the deep soil layers. Chapter 2 was based on the adjustment of the C sequestration potential model with microbiological and mineralogical variables. The modeling and mapping of different microbiological properties was carried out using spectral transfer functions and digital soil mapping (DSM), achieving R2 of 0.77 to 0.85. All these properties were detected using specific bands, which achieved correlations of 0.64 to 0.98 with the laboratory analyses. The autoregressive models obtained r from 0.61 to 0.7. The explanatory variables were associated with kaolinite, hematite, goethite, gibbsite and the abundance of fungi, actinomycetes, vesicular-arbuscular mycorrhizal fungi, enzymatic activity of beta- glucosidase, urease and phosphatase and particulate organic matter (POM), with the overall abundance of fungi being the most important microbiological variable. Chapter 3 was based on the development of a strategy to analyze microbiological activity at the microscale through the spectroscopic detection of 35 samples with analysis of of microbial biomass carbon (MBC) and enzymatic activity of beta-glucosidase, urease and phosphatase, and fractionation of soil organic matter (SOM) into POM and SOM associated with the mineral fraction (MAOM). In order to characterize the Vis-NIR-SWIR and Mid-IR spectra of the different fractions according to the microbiological variables, specific bands were selected for each variable. Finally, in chapter 4, a technique was developed to calculate and spatialize the activity indices of the enzymes betaglycosidase, phosphatase and urease for agricultural areas in Brazil using DSM and having as covariates a Synthetic Soil Image (SYSI), variables associated with relief, climate, biomes and mineralogical maps. The enzyme activity maps were obtained for the agricultural area of Brazil (3481362.60 km2), with a validation R2 ranging from 0.68 to 0.35. These enzyme activity indices on a 30 m scale can be considered an important contribution to monitoring and mapping the quality and health of Brazilian soils, as they are sensitive to land use and management.