Representative elementary volume in agricultural soils: Effects of image resolution and tillage practices on pore structure

Determining the representative elementary volume (REV) is crucial for accurately characterizing soil micromorphology using X-ray microtomography (µCT). This study examined the effects of image resolution (1.64, 5.25, and 35 µm) and soil management practices (conventional tillage – CT and minimum tillage – MT) on REV determination for various properties, including porosity, pore number, anisotropy, surface area, fractal dimension, connectivity of pores, and tortuosity. REV was defined by the stabilization of the coefficient of variation across subvolumes. The REV values ranged from 2.0 × 10⁻ 2 to 1.2 × 10⁻¹ mm³ at 1.64 µm, 2.3 × 10 0 –1.9 × 10 ¹ mm³ at 5.25 µm, and 5.4 × 10 3 –2.7 × 10⁴ mm³ at 35 µm, showing strong variation depending on the property and resolution used. Generally, CT yielded smaller REVs for most properties, indicating that tillage practices resulted in structural homogenization, while MT preserved greater pore complexity, especially at coarser scales. Fractal dimension, anisotropy, and tortuosity were particularly sensitive to resolution (1.64 µm), occasionally producing unrealistic values ( FD < 2, DA > 1, τ < 1) due to algorithmic limitations. For all measured properties, the REV was achieved at volumes of 1.2 × 10⁻¹ mm³ (1.64 μm) and 2.7 × 10⁴ mm³ (35 μm) for soil under CT, although it could not be defined for the 5.25 μm resolution. In the case of MT, the REV was achieved in the following subvolumes: 1.2 × 10⁻¹ mm³ (1.64 μm), 1.9 × 10 ¹ mm³ (5.25 μm), and 2.7 × 10⁴ mm³ (35 μm). Overall, REV is dependent on the specific property measured, the resolution used, and the management practices applied. Additionally, pore properties reflect a trade-off between accuracy at finer scales and representativeness at coarser scales.