Microtomografia de raios X na análise do volume elementar representativo e de mudanças na rede de poros em função do manejo e ciclos de umedecimento e secamento

Abstract

This thesis work is basically based on two axes, the first involving the definition of repre- sentative elementary volumes (REV) and the second, after the definition of the REV, the characterization of the pore network of a red latosol under different management systems and the analysis of the effects of wetting and drying cycles (W-D) on its porous system. REV measurements are fundamental in soil physics to ensure that the evaluated properties do not vary due to statistical fluctuations caused by the measurement scale, reflecting the intrinsic heterogeneity of the soil microstructure. One technique that can be successfully employed in REV analysis is micro-X-ray computed tomography (μCT). However, some questions arise in REV analysis, and some of them motivated this thesis work: 1) Is REV influenced by soil management for the same soil type?; 2) Do processes such as wetting and drying cycles affect REV?; 3) Does the analyzed pore size scale influence REV?; 4) Once REV is defined, are intra-aggregate pores affected by processes such as W-D cycles, regardless of the type of management to which the soil is subjected? The choice of intra-aggregate pore analysis in this study was due to its importance in protecting organic matter, water retention, microbial activi- ty, and aggregate stability. The management practices studied were conventional tillage (CT), no-tillage (NT), and minimum tillage (MT). The soil studied was a dystrophic Red Latosol widely used in agriculture in Brazil. The W-D cycles were performed using the capillary rise process for moistening and a tension table for drying. For the REV study, two different sub- volume selection schemes named cube and parallelepiped were used, as well as the CT, MT, and NT systems. In the study that evaluated the influence of resolution, three-dimensional (3D) images with the following voxel sizes were used: 1.64 μm, 5.25 μm, and 35 μm. The systems analyzed in this study were CT and MT. In the study on changes in the pore architec- ture of small aggregates (2-4 mm) as a function of W-D cycles, samples were analyzed under NT, CT, and secondary forest (F). In general, the physical properties evaluated included im- age-based porosity (P), fractal dimension (FD), degree of anisotropy (DA), connectivity (C), number of pores (NP/V), surface area (SA/V) per volume, and tortuosity (τ). The REV was defined by the stabilization of the coefficient of variation in progressively larger subvolumes. The results show that the REV depends mainly on soil management for P and C (both selec- tion schemes). The parallelepiped method presented a lower REV due to the larger volume analyzed in relation to the cube method. The REV obtained by the cube method was more sensitive to changes in the analysis scale, showing an increasing trend with the W-D cycles applied for P and FD. The REV showed dependence on resolution, with greater variability in the properties measured for 1.64 μm. Soil management influenced the REV, with CT general- ly presenting lower REVs, indicating greater homogeneity of the pore network. Artifacts in the images or a small amount of information generated anomalous values for some physical parameters (anisotropy, tortuosity, and fractal dimension) at a resolution of 1.64 μm, high- lighting the limitations of analysis algorithms in reduced subvolumes. Based on the results obtained, REV cannot be considered static, as it is influenced by soil management and pro- cesses that occur in it. Interestingly, even for very small sample sizes, REV can still be deter- mined, and the values of the physical parameters obtained are representative within the analy- sis scale. Regarding the analysis of changes in the intra-aggregate pore network as a function of different management practices and W-D cycles, it was found that the cycles did not signif- icantly affect (p > 0.05) the following physical properties: P, NP/V, FD, τ, and C. The effect of cycles on soil permeability (k) and hydraulic conductivity (K) was also studied, based on the most voluminous pore (MVP). No differences were observed in PMVP, FAMVP, ENMVP, τMVP, hydraulic radius, k, and K between 0 and 12 W-D cycles for MVP. Comparing soil management types after 12 W-D, for example, F samples became more porous than CT and NT samples. In contrast, the NT pore system had a smaller fractal dimension and was more tortuous than that of the CT and F samples. These results show that, for highly weathered soils, such as the latosol studied, the intra-aggregate pore network proved resilient to changes with W-D cycles, regardless of the type of management adopted. The results of this study provide basic information on the definition of representative elementary volumes and the de- pendencies of these volumes as a function of anthropogenic processes The results of this study provide basic information on the definition of representative elementary volumes and the dependencies of these volumes on anthropogenic or natural processes occurring in the soil. This type of result is fundamental because it helps define the minimum volumes of inter- est that should be adopted for X-ray tomography analysis. However, these volumes must be determined for each measured property and soil type, depending on the textural class, which was not investigated in this study. The results obtained also show that pores contained in small soil aggregates are not very sensitive to changes due to processes such as wetting and drying cycles, regardless of the type of management adopted. This type of information is in- teresting given the importance of intra-aggregate pores in the processes that occur in the soil.