A Standardized X-Ray µCT Workflow for Segmenting Roots, Pores, and Water-Associated Phases in a Homogeneous Plant-Growth Substrate

Quantifying roots, air-filled pores, and water-associated phases simultaneously in intact plant–substrate systems remains a methodological challenge in X-ray micro-computed tomography (µCT), because the gray-value ranges of these phases partly overlap and automated binary thresholding is often insufficient. This Communication presents a standardized and parameter-documented µCT image analysis workflow for segmenting and quantifying four phases (roots, air-filled pores, solid substrate matrix, and water-associated phase) in intact pot samples containing a homogeneous plant-growth substrate. The workflow combines 3D median filtering (3 × 3 × 3 kernel), average-based interactive thresholding, sequential spatial morphological operations, and watershed-based object separation for root-size distribution analysis. It was applied to 15 intact pot samples from a greenhouse experiment with perennial ryegrass (Lolium perenne) grown under three phosphorus fertilization treatments used as a demonstration system. The use of a standardized substrate enabled the derivation of global gray-value thresholds (Tpore ≤ 90; Tsolid ≥ 130) that were applied consistently across all datasets. A sensitivity analysis showed that these global thresholds overestimated root volume by approximately 10% relative to depth-dependent thresholds. Subsequent morphological dilation and subtraction operations separated the root mask from matrix-adjacent water-associated voxels. The sum of segmented phase volume fractions reached 99.758 ± 0.173% of the total ROI volume, confirming internal consistency of the workflow. Root-specific surface area differed among demonstration groups (1.99–2.23 mm−1), illustrating the workflow’s sensitivity to structural differences between samples. This approach can be transferred to controlled pot experiments with homogeneous substrates, providing a transparent framework for µCT-based analysis of intact root–substrate systems. Its practical interest lies in providing a non-destructive, high-resolution, and well-documented 3D analysis framework for root growth, supporting further research on plant–substrate interactions and treatment effects.