Papers

Summary

– This study aims: i) to characterize the clay mineralogy and study the spatial variability of the granulometric, chemical and mineralogical attributes of the soil based on geological, pedological and geomorphological information; ii) to evaluate the potential of diffuse reflectance spectroscopy (DRS) and magnetic susceptibility (MS) in the estimation of hematite (Hm) and goethite (Gt) iron oxides, granulometric and chemical attributes in the landscape compartments; iii) to use DRS and MS in the study of the quantification and spatial variability of soil CO2 emission (FCO2); and iv) to investigate the efficiency of the use of MS and iron oxides in pedotransfer functions (PTFs) for estimate the FCO2. The study area, 870 ha, is located in the municipality of Guatapará-São Paulo, and is cultivated with sugar cane managed with straw. The area is located in the geomorphological province of Plateau Western Paulista, presenting three materials of origin related to the transition between Serra Geral Basalts, Eluvial-Colluvial Deposit and Alluvial Deposit, four types of Oxisols and one Entisols, besides geomorphology characterized by concave and convex areas. A total of 372 samples were collected at a depth of 0.00-0.25 m for soil attributes. For the FCO2 study, an area of 90 ha was delimited. In order to evaluate the indirect methods DRS and MS, the data were compared with the results of chemical, granulometric and mineralogical analyzes. All data were submitted to classical and geostatistical statistical analysis to evaluate spatial dependence and variability. The variability of some soil attributes was stratified by compartments (geology, pedology and geomorphology), in order to understand the influence of each compartment on the variability of the attributes studied. The stratification of soil attributes for geological, pedological and geomorphological compartment is effective to study the variation of the attributes. DRS is an effective tool in the characterization of iron oxides and in the identification of different soil classes. MS is useful both for better detailing and re-adjustment of soil classes, iron content and textural class intervals, helping to identify management zones. All attributes studied have spatial dependence. The interpolated maps based on attribute prediction by MS and DRS show pattern of variability similar to maps based on observed data. However, for soils with high levels of iron oxides, such as ferric Oxisols, the MS was more indicated to predict the attributes studied. Among the attributes studied, iron oxides and MS were the main predictors of FCO2, proving their efficiency as pedoenvironmental of soil formation factors and processes.