The system is designed to mimic mechanical impedance and to identify genotypes with contrasting tolerance to compaction—fast enough to support data-driven breeding decisions.
Deep, penetrative root systems can help crops withstand drought by accessing deeper water and nutrients and may also boost carbon inputs to subsoils. However, breeding for deep rooting is constrained by a measurement gap: soil compaction blocks roots, and “root system penetrability”—the ability of the whole root system to pierce or navigate hard layers—has been difficult to phenotype rapidly, accurately, and at scale. While mechanization has raised farm productivity, it has also intensified anthropogenic subsoil compaction, reducing porosity and increasing mechanical resistance, which restricts rooting depth and contributes to yield losses. Traditional assays are often labor-intensive, invasive, or focused on primary roots, and X-ray CT remains costly and low-throughput.
A study (DOI: 10.1016/j.plaphe.2025.100143) published in Plant Phenomics on 21 November 2025 by Wolfgang Busch’s team, Salk Institute for Biological Studies, establishes RootXplorer as a robust, high-throughput phenotyping platform that enables accurate, whole-root-system assessment of soil penetration, providing a critical foundation for breeding crop varieties resilient to soil compaction, drought, and climate stress.
Using a standardized Phytagel-based cylinder system (CS), the researchers mimicked progressive soil compaction and quantified whole-root-system penetrability across dicots (Arabidopsis, soybean) and monocots (sorghum, rice). They varied Phytagel concentrations in species-appropriate media (MS or Hoagland), scanned plants with the RADICYL multi-view imager, measured bottom-layer mechanical impedance via penetrometer resistance (PR), and manually scored penetration using the root count ratio (RCR) and root area ratio (RAR). To capture secondary stresses associated with compaction, they tested hypoxia using the Arabidopsis pADH::GUS reporter and quantified water availability via media water potential (Ψ). They also validated CS realism with soil columns of different bulk densities (BD), measuring superficial porosity (SP) and PR, and built RootXplorer—an open-source, deep learning (U-Net++) computer-vision pipeline—to automate root segmentation and extraction of RCR/RAR. Results showed PR increased strongly with Phytagel concentration (r = 0.958–0.989), while penetration declined sharply: manual RCR and RAR were strongly negatively correlated with Phytagel concentration (RCR r = −0.954 to −0.991; RAR r = −0.954 to −0.982), dropping from high penetration at low gel (RCR 0.686–0.975; RAR 0.565–3.438) to near-zero at high gel (RCR 0.003–0.046; RAR 0.000–0.101), with Arabidopsis responding even to mild gradients. Low gel (0.4–0.8%) induced hypoxia (high GUS activity), whereas ≥1.0% reduced GUS signal, and increasing gel lowered Ψ (down to −0.518 MPa at 3.0%), indicating rising osmotic stress. Soil compaction similarly reduced SP (r ≈ −0.99) and increased PR, and root penetration fell with BD (r ≈ −0.98), confirming the CS replicates mechanical impedance. RootXplorer achieved high segmentation accuracy (IoU 0.978–0.983) and reproduced negative gel–penetration trends, with automated vs. manual agreement extremely high (RCR r = 0.987–0.997; RAR r = 0.978–0.999), enabling large-scale screening that uncovered substantial natural variation and conserved sorghum tolerance rankings under real soil compaction and X-ray CT.
RootXplorer provides a practical, high-throughput solution for assessing whole-root-system penetrability, enabling hundreds of plants to be screened within hours under standardized mechanical stress conditions. This capability can significantly accelerate the identification of soil compaction–tolerant genotypes for breeding pipelines, advance research into root architectural plasticity and root-type-specific mechanical strategies, and support the selection of deeper-rooting traits associated with improved drought avoidance, enhanced nutrient capture, and potentially increased carbon inputs to subsoils.
###
References
DOI
10.1016/j.plaphe.2025.100143
Original Source URl
https://doi.org/10.1016/j.plaphe.2025.100143
Funding information
This research was supported by gifts to the Salk Institute's Harnessing Plants Initiative (HPI) from the Bezos Earth Fund, the Hess Corporation, and the TED Audacious Project to W.B., and National Science Foundation (award no. 2243690) and the Governor's University Research Initiative program (05–2018) from the State of Texas grants to L.H.E.
About Plant Phenomics
Plant Phenomics is dedicated to publishing novel research that will advance all aspects of plant phenotyping from the cell to the plant population levels using innovative combinations of sensor systems and data analytics. Plant Phenomics aims also to connect phenomics to other science domains, such as genomics, genetics, physiology, molecular biology, bioinformatics, statistics, mathematics, and computer sciences. Plant Phenomics should thus contribute to advance plant sciences and agriculture/forestry/horticulture by addressing key scientific challenges in the area of plant phenomics.