Proximal sensing characterization of polygenetic soils variability in Brazil

• Within profile variation of soil properties were depicted by proximal sensing. • Soil physical and chemical properties varied even within a soil horizon. • Unapparent parent material discontinuities were revealed by sensors. • Both less and highly weathered soils vary more than noticed by field morphology. • pXRF and Vis-NIR provided important complementary data for soil characterization. In tropical regions, pedogenesis studies are challenging since most soils are polygenetic and studies on this approach are still lacking. Thus, complementary data is needed to understand their formation, which has been possible through proximal sensing tools. The objective of this study was to assess the efficiency of proximal sensing data to investigate the presence of lithological discontinuities and the within-profile variation of polygenetic soils formed from different parent materials and with varying weathering degrees. Soil morphology, texture, fertility, mineralogy, and reflectance analyses were conducted to characterize soil samples collected per horizon from five studied profiles. Additional samples were collected following a 15 x 15 cm grid and analyzed via portable X-ray fluorescence (pXRF) spectrometry. Parent material discontinuities were investigated through the ratios Ti/Zr, Si/Al, fine sand/coarse sand (FS/CS), and differences in the mineralogy of the sand, silt, and clay fractions. The five studied profiles were classified as: Fluvic Cambisol (CY), Sideralic Cambisol (CX), Xanthic Gibbsic Ferralsol (LA), Xanthic Ferralsol (LVA), and Rhodic Gibbsic Ferralsol (LV) per the World Reference Base (WRB/FAO) for Soil Resources. pXRF data revealed within-horizon variation of elemental contents. Chemical traits of anthropic and pedogenetic origin were successfully identified. The Ti/Zr ratio and mineralogical analysis of the sand, silt, and clay fractions were able to identify parent material discontinuities in LVA. By interpreting Vis-NIR spectra, it was possible to separate soils based on texture and mineralogy. Proximal sensor data, especially from pXRF, allowed for the detection of parent material discontinuities that were unapparent during field morphology analysis, contributing to improved details on soil genesis assessment and comprehension of previous soil formation events.