Synchrotron-based 3D X-ray computed tomography reveals root system architecture: Plastic responses to phosphorus placement

Synchrotron-based 3D X-ray computed tomography reveals root system architecture: Plastic responses to phosphorus placement

We used synchrotron-based X-ray computed tomography (SRXCT) to visualize root distribution in soil cores. X-ray CT is emerging as a leading technique to study plant roots, but SRXCT offers potential advantages compared with conventional X-ray sources, including producing X-rays of higher intensity t...

Full description

Saved in:
Bibliographic Details
Main Authors: Frederik J.T. van der Bom, Philipp D. Lösel, Andrew M. Kingston, Zhe H. Weng, Nelly S. Raymond, Anton Maksimenko, Daniel Hausermann, Michael J. Bell, Casey L. Doolette, Enzo Lombi, Peter M. Kopittke
Format: Article
Language:English
Published: Elsevier 2025-05-01
Series:Geoderma
Subjects:
X-ray computed tomography
Root system architecture
Root distribution
Soil phosphorus
Synchrotron
Image segmentation
Online Access:http://www.sciencedirect.com/science/article/pii/S0016706125001375
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We used synchrotron-based X-ray computed tomography (SRXCT) to visualize root distribution in soil cores. X-ray CT is emerging as a leading technique to study plant roots, but SRXCT offers potential advantages compared with conventional X-ray sources, including producing X-rays of higher intensity that are collimated, monochromatic and tuneable; delivering high-resolution data whilst avoiding issues such as beam-hardening and source divergence. We demonstrate the suitability of SRXCT for observing the root system of wheat plants growing in two soils (Calcisol and Ultisol) in response to placement of different phosphorus fertilisers. To optimize scanning quality, we tested the use of an inverse ‘mask’ in front of the soil cores to achieve a more uniform attenuation along the sample, thereby avoiding saturation of the detector along the thinnest parts of the soil cores. Secondly, we developed a deep learning approach for segmentation and quantification of root length and diameter. Our results demonstrate the use of SRXCT as a tool for studying root system architecture in soil at high spatial resolution. The SRXCT method marks a new stride towards advancing our understanding of root structures in unprecedented detail, opening further avenues for exploring plant-soil interactions.
ISSN:1872-6259

Similar Items