# Source paper archive

- Source title: Meta(xylem)morphoses: Root vascular patterning is dynamic and convergent in temperate grasses
- Author(s): Jones, D. H., Schoen, M., Damodaran, D., Trautewig, C., Baca Cabrera, J. C., Ware, A., Melzer, M., Schierenbeck, M., Zhao, D., Gimenez, R., Lobet, G., Pandey, B. K., Couvreur, V., Bennett, M. J., Rodriguez, D., Schneider, H.
- Source landing URL: https://www.biorxiv.org/content/10.64898/2026.11.19.767855v1
- PDF URL: https://www.biorxiv.org/content/10.64898/2026.11.19.767855v1.full.pdf
- PDF SHA-256: 88f6189138b7f41490bc21b1f99c7876ca977f19b513eb9afaa80c1075e45b67
- Converter: pymupdf4llm 0.0.27
- Limits: Automated PDF text extraction; reading order, equations, tables and formatting may be incomplete or incorrect. Images are omitted, and no OCR was performed. Markup and links are neutralized for safe downloading. Consult the original PDF for the authoritative version.

---

&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


**Meta(xylem)morphoses: Root vascular patterning is dynamic and convergent in**


**temperate grasses**


Dylan H Jones* &#91;†1&#93;, Maria Schön &#91;†1&#93;, Darshan Damodaran &#91;1&#93;, Corinna Trautewig &#91;1&#93;, Juan


Baca Cabrera &#91;3&#93;, Alexander Ware &#91;4&#93;, Michael Melzer &#91;1&#93;, Matias Schierenbeck &#91;1&#93;, Dongxue


Zhao &#91;5&#93;, Raul Gimenez &#91;5&#93;, Guilliaume Lobet &#91;6&#93;, Bipin Pandey &#91;4&#93;, Valentin Couvreur &#91;6&#93;,


Malcolm Bennett &#91;4&#93;, Daniel Rodriguez &#91;5&#93;, Hannah M Schneider* &#91;1,2&#93;


1 Department of Physiology and Cell Biology, Leibniz Institute of Plant Genetics &amp; Crop


Plant Research (IPK) OT Gatersleben, Corrensstr 3, Seeland 06466, Germany


2 Division of Root Sciences, Department of Crop Science, Georg-August-University


Goettingen, Goettingen 37075, Germany


3 Forschungszentrum Jülich GmbH, Institute of Bio- and Geoscience: Agrosphere


(IBG-3), Jülich, Germany


4 Plant and Crop Sciences, School of Biosciences, University of Nottingham,


Loughborough LE12 5RD, UK


5 Centre for Crop Science, The Queensland Alliance for Agriculture and Food


Innovation (QAAFI), The University of Queensland, Brisbane, QLD, Australia


6 Earth and Life Institute, Université catholique de Louvain, 1348, Belgium


  - Author for Correspondence


  - These authors contributed equally


Short title: Metaxylem remodelling in grasses


1


1


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


**Abstract**


In cereal roots the metaxylem vessels are the main channels of axial water transport.


Metaxylem size and number are considered constant traits characterising different root


classes. Here we show that metaxylem number is progressively remodelled along root


axes, producing mature roots with vascular anatomy differing by position. At initiation,


adventitious wheat root tips produce multiple files of metaxylem, and then


progressively fewer, until a single central metaxylem file exists, resembling the


vascular anatomy of the seminal roots. This pattern was observed in wheat, barley,


rye, oat, and _Brachypodium_, but not in maize or sorghum. We investigated the


developmental regulation of this phenomenon by creating a cell-resolved spatial


transcriptomic atlas of wheat adventitious root tips in these different metaxylem


production states and integrated this with bulk RNA-seq analysis of whole root tips.


We developed a pipeline integrating these complementary transcriptomic datasets to


reveal differentially expressed genes that included candidates controlling meristem


maintenance, vascular patterning, and metaxylem wall lignification. Hydraulic


modelling revealed changes in vessel number can alter the distribution of hydraulic


pressure levels in the root, while maintaining or increasing the capacity of roots to


transport water. We propose that this developmental transition preconfigures different


regions of the root for their changing hydraulic environments, providing a mechanism


to balance hydraulic robustness and efficiency over the root life cycle.


1


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


**Introduction**


Water transport is fundamental to plant growth, survival, and productivity. In cereals,


the transport efficiency of water and mineral nutrients from the soil to the plant aerial


tissue is determined not only by the size and architecture of the root system, but also


by the cellular organization of its vascular tissues (Jones et al., 2025). Despite this


central role in plant physiology, root vascular systems remain substantially less studied


than their aerial counterparts, particularly in mature field-grown plants where


recovering intact root systems is technically challenging (Ryan et al. 2016).


Consequently, how vascular organisation develops and is maintained across the


lifespan of cereal roots remains poorly understood.


The principal conduits for axial water transport in cereal roots are the metaxylem


vessels, which are positioned within the stele, the central vasculature cylinder. Once


differentiated, metaxylem vessels undergo programmed cell death to leave continuous


files of hollow, lignified cells that form vessels responsible for water transport from the


roots to the aerial tissue. Lignification enhances the cell files physical strength,


allowing them to withstand the negative pressures generated by transpirational pull


(Zimmermann et al., 2002). Axial conductance (Kx) is the sum of the conductances of


the individual vessels which as this scales to the fourth power of vessel radius &#91;as per


the Hagen-Poiuseuille law (Tyree and Zimmermann, 2002)&#93;, small perturbations in


vessel size can have large consequences for water transport capacity. Therefore, a


single wide vessel has a far greater transport capacity (Sanderson et al., 1988; Steudle


and Peterson, 1998) than multiple narrower vessels with the same total cross

sectional area. Alternatively narrower vessels can present an advantage under water


limitation, slowing water use to delay soil water depletion and reducing the risk of


cavitation (Harrison Day et al., 2023). Variation in metaxylem size and number have


accordingly been linked to water use, drought tolerance, and yield across various


cereals under controlled (Kadam et al., 2015; Klein et al., 2020; Strock et al., 2021;


Clément et al., 2022; Affortit et al., 2024) and field conditions.


Unlike dicotyledonous roots which develop additional vascular files through secondary


growth, cereal roots lack a vascular cambium and build their metaxylem files entirely


through primary growth. Metaxylem files are patterned close to the root apical


meristem, after which their number is developmentally stable, while many root


2


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


morphological features such as lateral root patterning, secondary cell wall


modification, and cortical senescence (Van Norman et al., 2013; Jones et al., 2025;


McCahill et al., 2025; Dhar et al., 2026) change with tissue age. This fixed post

developmental pattern of the metaxylem can therefore be read along the root axis as


a stable diary of the developmental decisions made at the root tip during growth. While


many studies rely on anatomical sampling at a single position to infer genetic


regulation or physiological impact (Koehler et al., 2025), this static approach misses


the broader temporal dynamics of tissue differentiation. Uncovering these dynamic


pattern variations over the course of root growth reveals how plants actively reprogram


cell fates long after early embryonic patterning is complete. Consequently, the


functional consequences of pattern variations occurring over the course of root growth


remain largely unconsidered.


Grass root systems are a particularly informative system in which to address this


question as they comprise several different classes of axial roots that arise in


sequence. Seminal roots emerge from the embryo and support early seedling


establishment (Hochholdinger et al., 2018), followed by adventitious roots which


emerge in acropetal sequence from stem nodes (Hochholdinger et al., 2004; Liu et al.,


2020; Schneider et al., 2021), and eventually make up the majority of the root system


in older plants. These classes can differ significantly in anatomy, growth rate, lifespan


and function (Liu et al., 2020; Koehler et al., 2025), whilst further differences can be


made between adventitious roots originating from different stem nodes. However,


these observations of difference are typically made at a single comparable sampling


position. Maintaining vascular continuity during growth requires developmental


coordination and necessitates conductive elements remain connected despite


continuous cell division, differentiation, and elongation occurring at the root apex.


Considerable progress has been made in understanding the molecular regulation of


vascular patterning and continuity in _Arabidopsis_, where auxin transport, canalisation,


and programmed cell death coordinate the formation of connected vascular networks


(De Rybel et al., 2016; Berleth et al., 2000; Scarpella, 2017). In comparison, far less


is known about how metaxylem continuity is established and maintained in grass roots,


and still less in the mature adventitious roots which dominate the later root network.


3


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


Here we surveyed root anatomy in intact mature root systems across multiple grass


and cereal species, sampling different root classes and positions along the axis. We


describe a previously unrecognised developmental transition during which


adventitious roots progressively re-pattern their metaxylem during growth. This


coordinated reduction in metaxylem files drives an anatomical convergence on a


vascular anatomical phenotype common to adventitious and seminal roots. This


transition phenomenon was conserved across a range of temperate grasses surveyed


(wheat, barley, rye, oat, and brome) but was absent in panicoid grasses (maize and


sorghum). To investigate its developmental basis, we utilised cell-resolved spatial


transcriptomics to create an atlas of adventitious root tip development at different ages


and combined this with bulk RNA sequencing to detect root class and age interactions.


To understand the bearing of this vascular repatterning on root physiology we utilised


hydraulic modelling to understand functional dynamics at the local and whole organ


levels. Together these analyses reveal an unexpected phase of vascular remodelling


in adventitious roots of grasses and provide new insight into how vascular continuity


is maintained during root development.


**Materials and Methods**


_**Plant material and growth conditions.**_ Plants were grown in three experimental


systems: hydroponic culture, soil-filled mesocosms and field trials. The species,


genotypes and system used for each experiment are summarised in Table S1.


_Hydroponic culture._ Seeds were surface-disinfected with 10% NaClO for 5 min, rinsed


three times in distilled water and germinated on filter paper moistened with tap water


in a controlled-environment chamber at 22 °C, 60% relative humidity and a 16 h light/8


h dark cycle with a photosynthetic photon flux density of 400 µmol photons m⁻² s⁻¹


(ISTA, 2018). After seven days, fifteen uniform seedlings per genotype were


transplanted into 75.5 L hydroponic tanks (Rubbermaid Brute; 38.3 × 70.7 × 44.1 cm)


containing a continuously aerated nutrient solution that was replaced every two weeks


(Schierenbeck et al., 2026). Plants were maintained at 17 °C day and 15 °C night


under a 16 h photoperiod (900 µmol photons m⁻² s⁻¹). For the bulk RNA-seq and


spatial transcriptomic experiments and the developmental time series, plantings were


staggered across four dates at two-week intervals, so that young and old adventitious


and seminal roots were available at a single harvest.


4


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


_Soil mesocosms_ . Seeds were pre-germinated on paper rolls for five days, after which


two seedlings were transplanted per column and later thinned to one after


establishment. Plants were grown in 30 L PVC mesocosms (110 cm length, 23 cm


internal diameter) lined with a plastic sleeve and perforated at the base for drainage,


at 25/16 °C under a 16/8 h photoperiod and a photosynthetic photon flux density of


830 µmol photons m⁻² s⁻¹ for 48 days. The substrate (bulk density 0.86 g cm⁻³)


consisted of 50% medium sand (particle size 0.20–0.63 mm), 20% agricultural


Chernozem A-horizon sieved soil of sandy clay loam texture (Altermann et al. 2005


;USDA, 2017) from the experimental fields of IPK Gatersleben (Germany), 15% perlite


(Perligrain Classic, 0.1–6 mm) and 15% vermiculite (Floragard, 1–2 mm), amended


with a controlled-release fertiliser (Osmocote 5 3-4M, 16N–8P–12K–2.2% MgO plus


micronutrients) at 5 g L⁻¹. Nutrient availability was further supplemented with nutrient


solution as described by Schierenbeck et al. (2026): maize received 825 mL per plant


and temperate cereals 525 mL per plant, each split into five applications. Plants were


watered twice weekly according to crop demand to maintain non-limiting water


conditions, giving a cumulative irrigation volume of 4.17 L per plant for maize and 3.67


L per plant for temperate cereals. Before harvest, irrigation was withheld for


approximately four days to allow drainage; columns were then dismantled, root


systems washed free of substrate, and stored in 50% ethanol until imaging.


_Field trials._ The sorghum trial was conducted at Warwick, Queensland, Australia


(28°12′46″S, 152°06′14″E) during the 2024/25 summer season, comprising 250 plots


(5 × 3 m) of 53 hybrids at 0.75 m row spacing and two plant population densities (6


and 12 plants m⁻²), on a deep black cracking clay (Vertosol; Isbell, 2021), sown on 20


December 2024. The wheat trial was conducted at Gatton Research Station,


Queensland (27°32′44″S, 152°19′54″E) during the 2024 winter season, as a factorial


combination of two plant populations (60 and 120 plants m⁻²) and 36 cultivars (13


commercial varieties and 21 breeding lines) in a split-plot design with four replicate


blocks, on black and grey cracking clays (Isbell, 2021; Powell, 1982), sown on 15 May


2024 into soil near field capacity. Root sampling for both trials is described below.


_Root sampling and anatomical phenotyping._ Root systems from mesocosm-grown


plants were separated into classes by origin; seminal and nodal adventitious roots


were distinguished in all species, and in maize and sorghum the embryonic primary


5


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


root was distinguished from the seminal roots. Each root was divided into 10 cm


segments (5 cm for wheat), and a 2 cm subsegment was taken from the midpoint of


each segment. These sections were imaged, annotated and measured with the Rapid


Anatomics Tool (RAT; Jones et al., 2025) following region-of-interest definition in FĲI


(ImageJ 1.54; Schindelin et al., 2012), yielding metaxylem file number, individual and


total metaxylem area, stele area and complexity (the metaxylem perimeter-to-area


ratio normalised by that of the stele); these measurements underlie Figure 1D and


Figure 2. Field root systems were sampled by soil coring (sorghum, 4.4 cm auger to 1


m, two on-row cores per plot; wheat, 5 cm auger to 0.9 m, one core adjacent to a


representative plant per plot); cores were cut into 10 cm depths, washed and, for


wheat, a 2 cm subsegment of the largest-diameter root taken from alternate depths


(0–10, 20–30, 40–50, 60–70, 80–90 cm). The three thickest roots per depth were


retained in 60–70% ethanol and pooled by depth (wheat, n = 12 cores; sorghum, n =


6 cores) before imaging as above using the RAT.


_Histology and light microscopy._ For the representative sections in Figure 1a and b,


fine hand sections of seminal and second-whorl nodal roots (5 cm segments, midpoint


± 1 cm) were fixed in 70% ethanol, 10% acetic acid and 20% water. All staining was


performed in the dark with gentle shaking. To visualise suberin and lignin, sections


were stained with Basic Fuchsin (Thermo Scientific; 0.2% w/v in 70% ethanol) for 10


min and washed twice in 100% ethanol, then with Fluorol Yellow 088 (ChemCruz;


0.01% w/v in methanol) for 5 min and washed in 50% ethanol, and counterstained with


Calcofluor White (Sigma-Aldrich) for 30 min before rinsing thoroughly in MilliQ water.


Sections were imaged the same day on a Keyence BZ-X1000 with a 10× objective,


acquiring Basic Fuchsin in the TRITC channel (excitation 545 nm, emission 570–640


nm), Fluorol Yellow in the GFP channel (470 nm, 500–550 nm) and Calcofluor White


in the DAPI channel (360 nm, 435–485 nm). For the high-resolution sections in Figure


1C, four-to-five-millimetre segments from seminal and adventitious roots were fixed by


microwave-assisted aldehyde and osmium tetroxide fixation (2% glutaraldehyde and


2% paraformaldehyde in 0.05 M cacodylate buffer, pH 7.3, followed by 1% osmium


tetroxide), dehydrated through a graded acetone series and embedded in Spurr's resin


using a PELCO BioWave 34700 microwave processor (Ted Pella, Redding, CA, USA;


full schedule in Table S2); sectioning and light microscopy followed Kavka et al.


6


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


(2024), images were saved in 8-bit, inverted, and a Cyan LUT was applied and images


were cropped to the stele.


_**Spatial transcriptomics:**_ _sample preparation._ Young (AY; 5–10 cm, 32 days after


germination) and old (AO; 30–40 cm, 46 days) node-2 and node-3 adventitious root


tips were collected from hydroponically grown wheat, trimmed to 3 cm and then to 8


mm, rinsed and embedded in optimal cutting temperature compound (OCT). To


maximise the number of tips per section, 16 (AY) and 12 (AO) root tips were arrayed


in a custom three-dimensionally printed insert (0.2 mm walls) within a 7 mm Tissue

Tek cryomould in 1 mm OCT layers of three to four tips; layers were frozen on dry ice


and assembled into a single compound block.


_Capture, library preparation and sequencing_ . Stereo-seq capture, library preparation


and sequencing were performed as a commercial service by BGI Genomics


(Shenzhen, China) following the standard STOmics Stereo-seq protocol, as was the


bulk RNA-seq library preparation and sequencing (below). Compound blocks were


sectioned transversely from the face furthest from the tips in 10 µm increments,


selecting a section immediately proximal to the meristematic region of the root tip,


where the root cap was visible and metaxylem were present but not yet fully


differentiated and still nucleated, and captured on Stereo-seq chips (500 nm bead-to

bead distance).


_Spatial data processing._ Reads were processed with the SAW pipeline (Gong et al.,


2024): coordinate and molecular identifiers were taken from read 1 (nucleotides 1–25


and 26–35 respectively) and the cDNA from read 2, coordinate identifiers were


mapped to the chip's spatial coordinates allowing one mismatch, and reads with an N


or low-quality base in the molecular identifier or a poly-A body were discarded.


Retained reads were aligned to the IWGSC RefSeq v2.1 wheat genome (Zhu et al.,


2021) with the STAR-based aligner within SAW (Dobin et al., 2013); uniquely mapped


reads were annotated to genes, molecular identifiers sharing a coordinate and gene


were collapsed, and a spatial gene-expression matrix was generated for each chip.


_Cell segmentation and annotation._ Four sections were selected per chip. For each, a


maximum-projection image of the nuclear (DAPI) and cell-wall (Fluorescent Brightener


28) stains was co-registered with the Stereo-seq expression matrix, and individual


7


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


cells were annotated manually in FĲI by drawing region-of-interest polygons on the


cell-wall image; polygons were converted to coordinate arrays and captured


transcripts assigned to the enclosing cell. Cells were assigned to one of 15 tissue


classes, yielding 6,685 annotated cells across eight sections.


_Clustering and marker analysis._ In Scanpy (v1.10.3; Wolf et al., 2018), per-cell counts


were normalised to counts per 10,000 and log-transformed, 3,000 highly variable


genes were selected (seurat_v3) and principal component analysis was performed on


the scaled data (50 components). Batch effects between chips and sections were


corrected with Harmony (harmonypy v0.0.10; Korsunsky et al., 2019), a


neighbourhood graph was constructed (50 neighbours over 50 principal components),


cells were clustered with the Leiden algorithm (leidenalg v0.10.2; Traag et al., 2019)


at resolution 1.4, and visualised by UMAP (umap-learn v0.5.6; McInnes et al., 2018).


Tissue-enriched markers were identified using scikit-learn (Pedregosa et al., 2011) by


a one-versus-rest Wilcoxon rank-sum test of each tissue against all others, with the


area under the receiver-operating curve derived from the U statistic, retaining genes


with AUC ≥ 0.70 and a Benjamini–Hochberg-adjusted p &lt; 0.05. The response of each


gene to root-tip age was tested within tissue by a two-sided Mann–Whitney U test


between young and old cells and summarised as the log 2 fold change (young versus


old); tissue-relative expression was min–max scaled across the 15 tissues for display.


_Bulk RNA-seq._ For each replicate, a pool of ten root tips from different plants was


collected, with each of the four sample groups (seminal young, seminal old,


adventitious young and adventitious old) represented by four biological replicates.


Total RNA was extracted with TRIzol (Invitrogen), genomic DNA removed with the


TURBO DNA-free Kit (Invitrogen) and RNA quantified by Qubit (Invitrogen); library


preparation and sequencing were performed by BGI Genomics (above). Clean reads


were aligned by BGI to the IWGSC RefSeq v1.1 wheat genome (IWGSC, 2018) with


HISAT2 (v2.2.1; Kim et al., 2019) and gene-level counts summarised with


featureCounts (Rsubread v2.18; Liao et al., 2014); gene identifiers were subsequently


converted to RefSeq v2.1 (Zhu et al., 2021), so that the bulk and spatial datasets


shared a common annotation. Differential expression was assessed with DESeq2


(v1.46; Love et al., 2014); genes were called differentially expressed at a Benjamini–


Hochberg-adjusted p &lt; 0.05 with no fold-change threshold, association with


8


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


metaxylem file number instead being defined by concordant direction of change across


the two target contrasts. The target contrasts were adventitious young versus


adventitious old (AY/AO) and adventitious young versus seminal young (AY/SY), with


adventitious old versus seminal old (AO/SO) and seminal young versus seminal old


(SY/SO) retained as exclusion contrasts.


_Candidate gene filtering._ Genes associated with metaxylem convergence were


selected by successive filtering (Figure 3a, b). Of the 82,074 genes detected in


common, those differentially expressed and concordant in direction across the AY/AO


and AY/SY contrasts, and not driven by the AO/SO or SY/SO exclusion contrasts,


were retained (5,093 genes). This set was filtered against the spatial data: genes more


highly expressed in metaxylem than in a pseudobulk of all other tissues were retained


(1,255 genes); of these, genes concordant in direction between the bulk contrast and


the young-adventitious metaxylem pseudobulk were retained (739 genes); and finally


genes with their highest per-cell mean expression in metaxylem were retained (448


genes). In parallel, the same spatial filtering applied to genes also significantly


differentially expressed in the per-section spatial pseudobulk yielded a stringent set of


five candidates.


_Functional enrichment._ Over-representation of MapMan bins (Mercator4 v6;


Schwacke et al., 2019) and Gene Ontology terms was tested by Fisher's exact test


with Benjamini–Hochberg correction, with fold enrichment calculated relative to the


expected proportion. To limit the influence of very lowly expressed genes, enrichment


used candidate sets against a background restricted to a mean expression of at least


one count per million; terms were called enriched at an adjusted p &lt; 0.05.


_Root hydraulic modelling._ Root prototypes were defined by varying the longitudinal


distribution of metaxylem number and cross-sectional area along a 70 cm wheat crown


root of 28 segments (2.5 cm each). Individual vessel cross-sectional area took one of


three values, 1,200, 4,000 or 8,000 µm² (the 12.5th percentile, median and 87.5th


percentile of the measured distribution; narrow N, medium M and wide W), and


metaxylem file number one of 7, 5, 3 or 1. Prototypes ran through all combinations of


initial and final file number and of initial and final area class, at one of three transition


rates (one file lost per 2.5, 5 or 10 cm) and beginning either at the root base or after


35 cm, giving 405 prototypes. Root radial conductivity (k_r, m MPa⁻¹ s⁻¹) was modelled


9


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


with GRANAR–MECHA (Heymans et al., 2020). A reference crown-root anatomy was


reconstructed in GRANAR from the median measured parameters (root cross

sectional area 440,000 µm², stele area 64,000 µm², six cortical layers) and coupled


with MECHA (Couvreur et al., 2018) using the default subcellular parameters (Table


S3). A single longitudinal k_r profile was applied to all prototypes, a high value for the


apical 2.5 cm and a low value for the remaining mature root, corresponding to


contrasting apoplastic-barrier scenarios (Table S3) and falling within the range


reported for wheat (Baca Cabrera et al., 2026; Bramley et al., 2009). Segment axial


conductance (k_x, m⁴ MPa⁻¹ s⁻¹) was calculated for each segment from metaxylem


number and vessel area by the Hagen–Poiseuille equation (Tyree &amp; Zimmermann,


2002), adjusted for axial hydraulic limitation (Barry et al., 2025). Whole-root


conductance was simulated with MARSHAL (Meunier et al., 2020) on a 30-day-old,


70 cm crown root and its laterals generated in CPlantBox (Giraud et al., 2023) and


parameterised from the hydroponic measurements (Table S4), with crown- and lateral

root parameters taken from the prototype parameterisation and previous studies


respectively. Outputs were whole-root-system conductance (K_rs), segment xylem


pressure (MPa) and axial water flow (J_xl).


_Statistics._ Anatomical data (Figures 1 and 2) were analysed in GraphPad Prism v10.4:


differences between root classes at each axial position by two-way ANOVA with


Tukey's multiple-comparison correction; the decline in metaxylem file number and


complexity along the axis by one-phase exponential decay fits; and relationships


between metaxylem area, file number and stele area by Pearson correlation.


Candidate gene expression (Figure 3) was compared in GraphPad Prism v10.4 by


one-way ANOVA with Tukey's multiple comparisons across the four sample groups,


and by a two-sided Mann–Whitney U test between young and old adventitious cells


within each tissue.


**Results**


**Metaxylem anatomy converges between adventitious and seminal roots in**


**wheat**


10


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


Figure 1. **Metaxylem anatomy converges between adventitious and seminal**


**roots in wheat.** (a, b) Root class origins and sampling positions, with the transverse


section imaged at each. Roots were divided into 5 cm segments and one section taken


from each midpoint (± 1 cm); sections i–x span ~2.5 to ~47.5 cm from the stem in (a)


seminal (pink) and (b) second-node adventitious (blue) roots. Stained for


callose/cellulose (cyan), lignin (magenta) and suberin (yellow). &#91;Scale bar, 0.5 mm&#93; (c)


11


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


Stele cropped sections in roots of different lengths, each sampled within 1 cm of the


base and 1 mm of the tip: (c.i) 8 cm and (c.ii) 30 cm seminal roots; (c.iii) 8 cm and


(c.iv) 35 cm adventitious roots, the latter also sectioned 17–18 cm from the base. (D)


Metaxylem file number, total metaxylem area, complexity and stele area versus


distance from the stem, for adventitious (blue) and seminal (pink) roots; complexity


ratio is the metaxylem perimeter-to-area ratio normalised by that of the stele.


Asterisks, significant difference between classes at a position (two-way ANOVA,


Tukey’s multiple comparison correction). Error bars, SD; n = 16.


Cross sections of young adventitious roots (grown in soil mesocosms) reveals that


their vascular anatomy is distinct from comparable positions in seminal roots,


particularly by the significant difference in metaxylem file number (Figure 1c.i, c.ii, d)


(Table S5). This is apparent both in the older basal sections of roots (Figure 1a.i-iv,


b.i-iv) and immediately adjacent to the root tip at comparable distances from the base


of the root (Figure 1c.i, c.iii). As cereal roots do not undergo secondary growth, the


vascular anatomy observed along the mature root reflects the developmental state of


the apical meristem at the time each segment was produced. This enables us to


reconstruct developmental changes in metaxylem patterning from mature root


anatomy. Sampling at approximately 2.5 cm from the root base, adventitious roots had


on average 5 metaxylem files in the stele, whereas seminal roots had only a single


centrally positioned file. As adventitious roots grew longer, the number of metaxylem


files decreased rapidly, converging on the seminal root phenotype within 30 cm of the


root base, while seminal roots maintained their single file throughout development.


Metaxylem file number differed significantly between root classes when comparing the


segments taken in the first 30 cm, and within the adventitious roots the presence of


significant difference in file number in segments and any distal segment ceased


between segments after 20 cm. These observations reveal a previously undescribed


developmental transition in which anatomically distinct root classes converge towards


a common ‘unipolar’ vascular architecture.


This convergence in metaxylem number was not accompanied by a significant change


in the total cross-sectional area of metaxylem files or in stele along the root axis in


either seminal or adventitious roots. At all positions these values differed significantly


between root classes. There was a significant and strong correlation between total


12


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


metaxylem cross sectional area and stele area in both seminal (r=0.862) and


adventitious roots (r=0.956), whereas the number of metaxylem files weakly correlated


with stele area (r=0.377). This indicates that while total metaxylem cross sectional


area and stele area may be developmentally coupled, the number of files of metaxylem


may be a consequence of separate developmental factors and not a process resulting


from allometry. Adventitious roots were significantly more complex than seminal roots


in the four most basal segments (Figure 1d), and both classes reduced significantly in


their mean complexity (normalised perimeter-to-area ratio) along the root axis.


Sections within the first 20 cm of adventitious root and the 15cm of seminal root


segments showed significant difference when compared to distal sections, then from


25 cm onwards no differences were present between or within root classes.


The conjugation of these datasets, sampling at maturity and near the meristem, across


different root classes and ages reveal a distinct developmental phase in the root tip in


which metaxylem pattern changes with developmental time.


**Root vascular remodelling exists in temperate cereal crops**


13


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


Figure 2. **Adventitious root metaxylem file number converges on that of seminal**


**roots with distance from the base across multiple grasses.** Cladogram of the


seven grasses surveyed, each paired with metaxylem file number versus distance


from the stem for seminal (pink), second-node adventitious (blue) and, where


sampled, soil-core (green) roots. Mesocosm-grown plants, two pooled genotypes per


species, n = 6; field plants sampled from soil cores, the three thickest roots per core


from two cores per plot across four plots and three genotypes, pooled by depth (wheat


n = 12, sorghum n = 4). Asterisks, significant difference between adventitious and


seminal roots at a position (two-way ANOVA, Tukey correction). Cladogram adapted


from Grass Phylogeny Working Group III (2024)


To determine whether developmental vascular remodelling is unique to wheat or


represents a broader feature of cereal root development we quantified metaxylem


pattern across many major cereals including oat, rye, barley, several genotypes of


wheat across different environments, brome, sorghum, and maize. Remarkably, the


same progressive reduction in metaxylem file number along the adventitious root axis


was observed in every temperate species. This was reproducible across hydroponic


culture, soil mesocosms, and field grown wheat; indicating that vascular remodelling


is not an artefact of growth conditions, but reflects an intrinsic and conserved


developmental program which leads to convergence between seminal and


adventitious root anatomy (Figure 2). This was confirmed by two-way ANOVA, through


discovery of significant effects of position along axis, root class, and segment by class


interactions, in wheat, barley, rye, oat, and brome.


We extended our studies to also include field grown plants with roots extracted from


soil cores as discontinuous sample positions. Consistent with our controlled

environment experiments, no root with more than one metaxylem file was found in


samples collected from below 30 cm in any sampled wheat genotypes in our field trials.


In contrast, sorghum maintained multiple metaxylem files in all segments across the


whole sampled soil profile, reinforcing our conclusion that adventitious root metaxylem


pattern remodelling differs fundamentally between the pooideae and panicoid grasses


surveyed, regardless of growing conditions.


14


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


**A cell-resolved spatial atlas for wheat root tips resolves tissue identities**


Figure 3. **Cell-annotated spatial RNA-seq shows high concordance between**


**manually and statistically defined cell groups expressing known wheat marker**


**genes.** (a) Sampling and section preparation: (a.i) Adventitious Young (AY; 32 days


after germination, 5–10 cm, n = 16) and Adventitious Old (AO; 46 days, 35–40 cm, n


= 12) second-node tips were excised, trimmed to 5 mm, (a.ii) OCT-embedded and


(a.iii, a.iv) sectioned transversely from the base to just above the quiescent centre,


15


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


where metaxylem were present but undifferentiated and still nucleated; (a.v) captured


on Stereo-seq chips (one per age). (b) UMAP of 6,685 annotated cells coloured by


(b.i) tissue identity, (b.ii) Leiden cluster (resolution 1.4) and (b.iii) sample of origin. (c)


Representative annotated AY and AO sections, coloured by tissue and by Leiden


cluster identity. (d) Correspondence between clusters and the 15 tissues, as the


fraction of overlapping cells. (e) Colour scales for (d) and (f). (f) Marker gene


expression across the annotated tissues; rows are tissues (Metaxylem to Epidermis)


and columns are marker genes grouped by tissue. Colour shows relative expression,


the mean expression of each gene per tissue min–max scaled from 0 to 1 across the


15 tissues. Asterisks denote significant enrichment of a gene in that tissue, tested by


a one-versus-rest Wilcoxon rank-sum test with the area under the receiver-operating


curve derived from the U statistic, retaining genes with AUC ≥ 0.70 and a Benjamini–


Hochberg-adjusted p &lt; 0.05 (* p &lt; 0.05, ** p &lt; 0.01, *** p &lt; 0.001). Markers were


derived by ranking each tissue the genes that best distinguished it from all others and


retaining those above the AUC and adjusted-p thresholds. Wheat genome version 2.1


identifier background denotes the source of the marker: blue, identified from the


anatomical annotations in this study; light yellow, wheat root atlas (Ke et al., 2025);


darker yellow, wheat atlas marker validated across species; light green, atlas marker


overlapping an annotation marker from this study; dark green, cross-species atlas


marker overlapping an annotation marker, purple boxes show overlap between cell


type and expected marker gene.


Having established that vascular remodelling is a conserved developmental


programme in temperate cereals, we next sought to identify the molecular changes


accompanying this transition. We generated whole-transcriptome spatial


transcriptomic profiles (untargeted polyA-captured on the chip's oligo-dT array) of


young and old adventitious root tips using Stereo-seq (Figure 3a), and assigned each


encircled capture point (arrayed in a 500 nm grid) to an individual cell of one of fifteen


tissue classes with an exclusion zone two capture points thick between cells. This was


achieved using the co-registered cell wall image, and provided 6,685 annotated cells,


with 4 root tips sections for both the young and old root tips (Figure 3c). Following


unguided harmonisation of these cells, this cell pool showed no strong chip or root


16


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


piece specific separation (Figure 3b.iii). Unsupervised clustering of the 6,685


annotated cells (Leiden resolution 1.4) produced clusters that resolved the major


tissue types and closely mirrored the manually assigned identities (Figure 3b.ii).


The clustering highlighted similarities between certain cell types. The clustering


strongly validated the assignment of the hypodermis annotation as separate to the rest


of the cortex, showing very few hypodermal cells outside cluster 6. The separately


annotated cortical cell file layers, (Cortex 1 to 5) predominantly grouped together in


cluster 0 (Figure 3d), however there was overlap between the innermost cortical cell


file layers (1 and 2) and the endodermis in cluster 4 (Figure 3d). Xylem and metaxylem


clustered together in cluster 9, with all metaxylem falling into this cluster, though a


small proportion of xylem also appeared in other adjacent clusters. Remaining


vascular tissues clustered less cleanly, though both phloem and phloem pole cells


peaked in fraction of assigned cells uniquely in their own clusters (7 and 10


respectively). Stele cells neighbouring metaxylem files were assigned their own


identity in the manual annotations, however these distributed across clusters 2 and 5


along with the stele parenchyma (Figure 3d).


**Tissue marker genes confirm identity assignment and reveal root tip age**


**differences**


Having shown the assigned cell types are transcriptionally distinct, we proceeded to


validate these identities using established markers and sought novel markers for our


additional cell types. All manually annotated tissue types could be confirmed through


peak expression of marker genes from literature in the expected tissue. Several


markers from literature also showed high (and in some cases peak) expression in other


tissues, affirming the importance of multiple references; however, this challenge was


limited to select pericycle and endodermal genes as expected from the cluster


overlaps, which had no bearing on downstream analysis. Marker genes for metaxylem


were highly specific to the appropriate cells, as were xylem markers, being clearly


absent from the other group despite the similarities at clustering. We also confirmed


that these marker genes were stable between age groups, showing no significant fold


change in the target tissue between young and old root tips.


17


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


From our data set we were able to derive candidates for cell type specific markers.


Several of these overlapped with marker genes identified in the single cell and spatial


wheat root atlas produced by Ke et al. (2025) (Figure 3e, Table S6) (atlas markers in


yellow, cross species bright yellow, and green where overlapping with markers


identified in this study). Several peroxidases appeared as metaxylem markers in our


dataset, which were highly metaxylem specific. In combination, the clustering and


marker gene expression analysis strongly support our cell resolved spatial


annotations; these also allow identification of redundant annotations or tissue regions


containing previously underexplored heterogeneity.


18


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


Figure 4 **Integrated filtering of the bulk and spatial transcriptomes nominates**


**candidate regulators of metaxylem convergence.** (a) Filtering pipeline: root tips of


the four groups (seminal young, SY; seminal old, SO; adventitious young, AY;


adventitious old, AO) were profiled by bulk RNA-seq and Stereo-seq and filtered


through six steps (1 to 6, corresponding to the rows of b). (b) Genes retained at each


step, for the primary bulk-anchored set and the parallel spatial-pseudobulk-concordant


set; up- and down-regulated in blue and pink. Step 1, genes detected in both platforms


(82,074); 2, differentially expressed in the bulk (DESeq2, Benjamini-Hochberg

adjusted p &lt; 0.05); 3, concordant across the two target contrasts (AY/AO, AY/SY)


excluding SY/SO, AO/SO; 4, more highly expressed in pooled metaxylem than in a


pseudobulk of all other tissues; 5, concordant between the bulk and the spatial


metaxylem pseudobulks; 6, per-cell expression peaks in metaxylem. (c) MapMan


enrichment of the bulk and metaxylem (MX) young-up and old-up sets; left, gene count


per bin; right, mean fold change; crosses, bins with no genes in the set. (d to g) Four


candidates, per-tissue CPM in the spatial data (AY left with white border, AO right) and


per-group CPM from the bulk; (d) a class III peroxidase (TraesCS7A03G1097700), (e)


a BEL1-like homeodomain factor (TraesCS3D03G0793500), (f) an LHW-like bHLH


factor (TraesCS5D03G0335500) and (g) an AGAMOUS-type MADS-box factor


(TraesCS5A03G0321700). Bars, mean ± SD (n = 4). Significance: (d–g, left) two-sided


Mann-Whitney U test comparing young and old adventitious cells within each tissue;


(d–g, right) one-way ANOVA with Tukey's multiple comparisons, shared letters


denoting groups that do not differ. * p &lt; 0.05, ** p &lt; 0.01, **** p &lt; 0.0001


Across the two transcriptomics experiments, expression of 82,074 genes were


quantified. From the bulk RNA-seq we identified genes differentially expressed


(DESeq2, Benjamini-Hochberg-adjusted p&lt;0.05) between roots contrasting in the


number of metaxylem that were being produced at the tip (Table S7). This was done


using the contrast groups Adventitious Young versus Adventitious Old (AY/AO), and


Adventitious Young versus Seminal Young (AY/SY), with Adventitious Old versus


Seminal Old (AO/SO) and Seminal Young versus Seminal Old (SY/SO) as exclusion


filters to remove genes that relate primarily to differences in age, phenology, or root


class rather than metaxylem file number. Taking just the gene set where the direction


19


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


of differential expression was concordant in both of the target contrast groups gave


5,093 genes we associated with metaxylem number. This set was then filtered using


the spatial data and restricted to genes that were more highly expressed in the pooled


metaxylem than in a pseudobulk pool of all other tissues which brought our set down


to 1,255 genes of interest. We then filtered this list again, looking for genes that were


concordant in direction of change from the many to one bulk pool, and between the


young and old spatial pseudobulks (yielding 739 genes). This list was then filtered for


genes that showed their peak expression in metaxylem, giving 448 genes of interest.


A more stringent filter was also applied at the start, keeping only genes that were also


significantly differentially expressed in the spatial pseudo bulk looking for genes that


were differentially expressed between chips AY and AO chips. After filtering for


concordance with the bulk RNA-seq and applying the other described filters this


yielded five genes of interest, one upregulated in young adventitious root tips relative


to all other sample groups (Figure 4d), and four downregulated, one given as an


example in Figure 4f. Functional enrichment varied with filtering stage and root tip age


(Figure 4b), with transcriptional co-regulation enriched in the young metaxylem

specific set and cell-cycle organisation and microtubular network terms in both old root


tip sets.


The most strongly metaxylem specific genes in the spatially cell filtered set of 488


genes were peroxidases (Class III, which catalyse lignification of vessel walls),


comprising the whole top six (Figure 4d), several of which were identified as marker


genes in the earlier analysis (Figure 3e). Though their overall and per cell expression


changed with root tip age, they were more highly expressed in metaxylem than any


other cell type in both conditions. Several bHLH transcription factors appeared in the


448 set list, including an orthologue of Lonesome-Highway (LHW), although closer


inspection revealed age effects in the seminal roots and a less clear spatial expression


pattern, it is notable this candidate with a well characterised role in determining xylem


pole number in _Arabidopsis_ and rice was surfaced. In the five gene subset with


significant changes in expression in the spatial pseudobulk were two genes with


functional annotations relating them to meristem maintenance (Figure 4e,f). These


were a BEL1-Like homeodomain protein (a positive regulator of meristematic cell


division), and an Agamous MADS-box Transcription factor, a repressor of


meristematic cell division, more highly expressed in the poly-metaxylem file tip state


20


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


and mono-metaxylem file tip state respectively. Genes homologous to or bearing


functional annotations of Forked or Forked-like family members appeared in the


unfiltered bulk list, but had a similar expression pattern in seminal as well as


adventitious roots and were expressed more strongly in xylem and stele parenchyma


than metaxylem.


This conservative filtering strategy prioritizes specificity over breadth, potentially


excluding regulators with broader spatial expression or those acting upstream of root


class and meristem age. Despite these stringent criteria, the remaining candidate


dataset captured significant transcriptional shifts across successive stages of


metaxylem determinacy, including meristematic maintenance, cell division, cell


elongation, and cell wall modification. Rather than pointing to a single master


regulator, these results demonstrate a multi-layered regulatory program governing the


metaxylem transition, even within the narrow developmental window captured by the


spatial data. Consequently, while these cell-type filters may not capture the entire


regulatory network, they successfully isolate the core parallel processes driving this


anatomical shift, providing a robust foundation for future functional validation.


Figure 5. **Metaxylem file number and vessel size shape whole-root conductance**


**and xylem pressure** . Hydraulic properties of modelled wheat root prototypes of


21


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


defined metaxylem configuration, simulated with GRANAR, MECHA and MARSHAL.


Each configuration is named by metaxylem file number and cross-sectional area class


(N, narrow; M, medium; W, wide); Prototypes share one of three basal anatomies


(1_Medium, 5_Medium and 7_Narrow) against which the others are compared. Each


prototype radiates from the shoot junction and is coloured along its length by the


metaxylem hydraulic pressure potential (MPa; colour bar), with tip stele anatomy


metaxylem shown in cross section at each end. The table below each panel gives the


total metaxylem cross-sectional area, whole-root conductance (Krs), tip conductance


(Kx) and the metaxylem pressure potential at the tip; each is expressed as a


percentage of the reference anatomically unchanging configuration, with the reference


itself (highlighted) given as absolute values


**Metaxylem number and vessel size shape whole-root conductance and xylem**


**pressure**


To assess the functional consequences of metaxylem file remodelling, we simulated


the hydraulic properties of roots with different patterns of metaxylem change under


different basal and distal anatomies, using GRANAR (Heymans et al., 2020) and


MECHA (Couvreur et al., 2018) to estimate realistic radial uptake values and


MARSHAL (Meunier et al., 2019) to integrate positional values. Using this integrated


modelling framework allowed varying of only metaxylem patterns along the axis and


no other traits in order to target the specific functional impacts of metaxylem file


patterning exclusively. Root prototypes were digitally constructed, contrasting in basal


and distal metaxylem file number, initial and final mean metaxylem cross-sectional


area, transition rate (distance in cm for a mx file to be lost) and root length. Here we


focus on the contrast between anatomically uniform reference prototypes and several


root morphologies whose longitudinal transition patterns match the observed


phenotypes. At the level of a single cross section, axial conductance (Kx) is


proportional to the sum of the fourth power of the vessel radii, and can therefore be


dominated by the width of the individual files in the segment, so a change to fewer,


larger vessels has a predictably multiplying effect on segment conductance. The


behaviour of the root as a whole is less intuitive: the axial conductance of each


segment governs all the water that must pass through it toward the shoot, so a change


in the basal region, through which the entire root hydraulic movement travels, weighs


22


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


far more heavily than the same change near the tip, and capturing this requires


integrating conductance along the whole root. This is also due to the additive effects


of axial limitations, meaning for a specific axial conductance, additional root length will


not affect whole root conductance (Meunier et al. 2017, Baca Cabrera et al. 2024).


This integration showed that the whole-root cost of a progressive loss of metaxylem


files depends on how that loss occurs. Relative to the matched unchanging prototype,


scenarios in which the metaxylem file pattern transitioned to a wider single central file,


in the manner seen in wheat adventitious roots, typically increased whole-root


conductance (Krs). This increase occurs even when seven narrow files converging to


a single medium area file (7N:1M), reduces the final metaxylem area to less than half


of the starting area. In a further scenario, where five medium vessels with a total area


of 20,000 µm² transitioned to a single vessel of 8,000 µm², whole-root conductance


was maintained at 98% of the unchanging prototype, despite the axial conductance of


the post-transition segment being only 80% of that of the starting condition.


Conversely, in the seminal root prototypes, which start and end with a single vessel


and change only in its cross-sectional area, reducing that area by 70% or increasing


it by 100% relative to the 4,000 µm² unchanging scenario, results in an eleven-fold


decrease and a four-fold increase in Kx respectively, but only a 65% decrease and a


31% increase in Krs (Figure 5, Table S8).


These divergent hydraulic responses demonstrate that whole-root conductance can


either buffer or magnify the localized impacts of vascular remodelling depending on


the specific geometry of the transition. This indicates that anatomical continuity along


the longitudinal axis is an independent and highly impactful property determining root


hydraulic function, decoupled from simple cross-sectional area metrics.


**Discussion**


**A heteroblastic phase change in root apical meristems of temperate cereals**


The metaxylem anatomy of cereal adventitious roots is not fixed at root genesis, but


in the developing region is progressively remodelled during root growth until a final


production state of a single central file is reached. As this pattern is largely constitutive


in seminal roots, vascular patterning is convergent in the axial root classes. Viewed


along the root axis in a progressively distal view, the metaxylem undergo a confluence,


23


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


though when viewed in the direction of xylem sap flow and from the apical meristem


towards older tissue this reads as a furcation (Alemán-Sancheschúlz et al., 2026). At


maturity, it is the case that the basal, older root segments typically contain more files


of metaxylem than distal segments, but as we show through our time series studies,


this is better interpreted as the distinct vascular anatomy as produced by a ‘young’


root tip rather than a consequence of tissue age. For these reasons we consider the


metaxylem pattern along mature cereal adventitious roots a feature that is more readily


interpreted in distance from the point of organogenesis at the root junction as opposed


to relative to the root tip.


Given how fundamental metaxylem function it is striking that this phenomenon is not


more widely documented, however this speaks to the technical challenges of


recovering intact lengths of adventitious roots. Several studies have however


presented data where this transition is visible (Wu et al., 2011; Ouyang et al., 2020;


Baca Cabrera et al., 2026); mostly focusing on the changes in lignification and


secondary barrier formation along the root axis. Insights into species not included in


our surveys can be gleaned from these studies, and interestingly present examples of


contrast to the pattern we observe in the cool season grasses. Rate of vascular change


as a trait appears to vary within wheat, and shows variation within a panel of winter


wheat cultivars released over 120 years of breeding (Baca Cabrera et al., 2026;


Behrend et al., 2025). This difference between cultivars was heritable and showed a


gradient in appearance that aligned with cultivar release year, with more modern


cultivars showing a greater extent of vascular remodelling within the same length of


root sampled.


What defines this as an anatomical developmental transition rather than a gradual loss


of meristematic proliferative potency is that after this transition occurred, in no cases


did the root cease to grow at the end of this phase of change, or did there cease to be


a central metaxylem file in the root, as can be seen in lateral roots (Shishkova et al.,


2008). In all of these species in which the distinct remodelling occurred, regardless of


the starting condition, or the length of root produced before this transition was


complete, the endpoint was the same and irrevocable, and had no clear bearing on


final root length.


**Meristem progression drives the meta-xylem-morphoses**


24


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


The reduction in metaxylem file number reflects a change in what the root apical


meristem produces over developmental time, rather than any alteration to a segment


once it has matured. Consistent with this, when we imaged root tips across a range of


root lengths, the youngest, shortest tips contained the most metaxylem files close to


the tip, and this number fell as roots lengthened and tips aged. We can therefore divide


adventitious root development into two phases: an initial transition phase following


emergence, in which multiple files are present and the vasculature is actively


remodelled, and a subsequent stable phase, in which a single central file is produced


and maintained.


How the number of metaxylem files produced at the meristem, and the change in that


number, is regulated remains a key question that will require further exploration. A


next step will be to clarify the cellular mechanism of this remodelling: whether multiple


metaxylem initials relocate to become neighbouring fuse, and or are eliminated,


whether a single initial divides and branches (Miki et al., 2026; Saito et al., 2019)


before differentiation in the early phase and ceases to branch later, or whether a


combination of these occurs, with multiple initials both changing their division pattern


and being eliminated. Given the large spatial distance over which this transition occurs


relative to the size of vascular cells pre-elongation, resolving this will require high

resolution three-dimensional anatomical surveys, or high frequency time-point


sampling with lineage tracing of cell files in the root tip. What is clear is that this process


must be tightly regulated to preserve hydraulic continuity and root function (Sawchuk


et al., 2013; Etchells et al., 2016; Turco et al., 2019). We suggest that this behaves as


a developmental phase transition of the root meristem, comparable to the phase


transitions long recognised in shoot meristems, such as the switch from vegetative to


reproductive (Huĳser and Schmid, 2011) growth, but expressed through vascular


anatomy rather than organ identity. Such transitions are well characterised in shoots,


yet remain little explored in roots and even less so in grasses (Brady et al., 2006, 2007;


Yang et al., 2024; Rodriguez-Alonso et al., 2018). Developmental regulators can


similarly be conserved or recruited to control distinct developmental outcomes across


grasses (Klein et al., 2022; Xiao et al., 2022; Gallagher et al., 2023).


To target these changes, we profiled the transcriptomes of root meristems producing


contrasting vascular anatomies over the course of development, using bulk RNA-seq


25


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


to define a candidate pool and spatial transcriptomics to filter it to genes with focused


metaxylem expression and concordant change. This conservative, two-technique


strategy isolates a smaller set of higher-confidence targets than either method alone,


to prioritise for functional study in transformed or mutagenised populations. We


devised a strategy that draws on the complementary strengths of the different


transcriptomic platforms. Spatial transcriptomics provides excellent spatial resolution


and allows direct annotation of cell types. It is, however, resource-intensive and low in


capture rate (De Veylder et al. 2026), and because metaxylem cells are scarce,


especially in the mono-file state, our spatial data alone struggles to provide the


replication needed to serve as our main discovery platform. Bulk RNA-seq is


complementary, robust and well-replicated but not cell-type specific, and so carries


the differential-expression discoveries, while the spatial data validates and filters the


candidates. This combines the stringency of the spatial with the robustness of the bulk,


an approach that should generalise to other low-abundance cell types. Filtering


against the spatial data often increased the mean fold change of the enriched


functional groups (Figure 3B), indicating that it not only shortens the candidate list but


enriches for the more strongly varying genes of interest.


The candidates fall into two groups, the first relating directly to the vessels. An LHW

like bHLH transcription factor was present among the metaxylem-associated genes,


of interest given the established role of LHW in setting xylem pole number in


_Arabidopsis_ (De Rybel et al., 2013; Ohashi-Ito et al., 2024), and its disruption of


metaxylem file reduction in rice (Reeger et al., 2021). Several peroxidases were


among the most strongly metaxylem-specific genes, most plausibly reporting


metaxylem differentiation rather than driving the change in file number due to their role


in catalysing the oxidative polymerisation of mono-lignols during cell wall lignification


(Marjamaa et al., 2009). Their per-cell expression also fell from young to old tips, which


is itself informative: the signal is strongest when metaxylem files are most numerous,


so bulk data alone might have ascribed it simply to a greater number of metaxylem


cells, whereas the cell-resolved data show a genuine change in expression per cell.


The second group relates more to the meristem. Among the genes differentially


expressed between young and old adventitious tips were a BEL1-like homeodomain


factor and an AGAMOUS-type MADS-box factor, both with a predicted role in


meristem maintenance and determinacy (Kumar et al., 2007; Cao et al., 2024).


26


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


AGAMOUS is known for specifying a shift in identity at the floral meristem (Bartlett,


2017; Abraham-Juárez et al., 2020; Tapia-López et al., 2008; Montiel et al., 2020), and


its increased expression in older tips is consistent with a meristem progressing


towards a more determinate, single-file output. Together these analyses do not


immediately resolve the regulatory network underlying the transition (Brady et al.,


2011), but they isolate candidate processes and, alongside the validated cell-resolved


atlas of the wheat root tip, establish wheat as a tractable system in which to dissect


developmental phase transitions in roots.


**Evolutionary divergence of the transition across grasses**


The transition is not universal among grasses. It occurred in every genotype of the


temperate species we surveyed, wheat, barley, rye, oat and _Brachypodium_, but was


absent in the PACMAD cereals maize and sorghum, the adventitious roots of which


retained multiple metaxylem files along their entire length and remained anatomically


distinct from their seminal roots. Basal metaxylem file number and area can differ with


species, genotype and stem node of origin in both temperate and tropical cereals, and


have been the focus of many studies linking root anatomy to plant physiology and


agronomic (Strock et al., 2021; Klein et al., 2024; Affortit et al., 2024) performance.


Basal anatomy has been shown to be an important trait, but our results indicate that


in many species sampling at this position is insufficient.


Why the transition is present in the temperate grasses surveyed but not the tropical


ones is intriguing, and plausibly relates to adaptation to their respective ecological


niches; the contrast may in part reflect the different water-use and conservation


strategies associated with C3 and C4 (Gowik and Westhoff, 2011) photosynthesis.


The BOP clade offers further contrasts: though changes in metaxylem number do


occur, neither rice nor intermediate wheatgrass show the rate or extent of remodelling


(Ouyang et al., 2020; Clément et al., 2022) seen in the other members we surveyed,


which may reflect the distinct demands of a semi-aquatic habit and a perennial lifecycle


respectively. The trait may therefore segregate along several axes, including clade,


photosynthetic type, lifecycle and adaptation to water availability (Kadam et al., 2015)


or other edaphic gradients such as soil density, nutrient availability and temperature


(Zobel, 1992; Peralta Ogorek et al., 2025). Alongside a fuller characterisation of the


27


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


developmental landscape of this remodelling, surveying more diverse grasses


adapted to contrasting environments should help resolve its regulation and function.


**Functional consequences**


The remodelling also carries consequences for the hydraulic architecture of the root.


Modelling of the measured anatomy indicated that converging on a single, enlarged


central vessel is an efficient way to maintain whole-root axial conductance (Hacke et


al., 2001; Tyree and Zimmermann, 2002) as the root remodels, because the outcome


depends more on the configuration through which metaxylem area changes than on


the amount of area lost; concentrating area into fewer, larger vessels preserves or


increases conductance, whereas an equivalent loss of area spread across an


unchanged number of files does not. Because axial conductance integrates along the


whole root, holding one configuration along the entire axis would compound its costs


(Meunier et al., 2017; Baca Cabrera et al., 2024), whereas the graded transition


matches the supply capacity of each region to the demand placed on it: multiple files


where the cumulative flow from the whole root passes basally, converging to a single


efficient vessel distally where flow is lower.


Having two distinct vascular states may balance several competing demands over the


life of the root, in particular a trade-off between robustness and efficiency. A single


wide conduit is more vulnerable to failure through embolism (Harrison Day et al., 2023)


than several narrower ones in which the failure of any one removes less of the total


conductance; retaining files basally, where the consequences of failure are greatest,


while converging distally as tension rises places robustness and efficiency where each


is most useful. The risk of failure at greater depths may also be reduced, being


upstream of the additive longitudinal pressure gradient produced by friction (Bouda et


al., 2018) of water transport. Thus, rather than responding plastically to the local soil


hydraulic environment, developmental remodelling of metaxylem patterning may


preconfigure different regions of the root for the hydraulic conditions they are likely to


experience later in the plant life cycle. Older basal tissues retained in the upper soil


profile maintain multiple conduits, whereas younger tissues extending into deeper soil


are produced with a hydraulically distinct vascular configuration. The reduction in


vessel area may also (Richards and Passioura, 1989) conserve water, favouring an


architecturally distributed root network over a high-conductance one built from fewer


28


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


roots, and could suit growth through vertically heterogeneous soil, where water,


mechanical impedance, temperature and nutrients all change with depth, as too does


the intensity of environmental fluctuations (Baumberger et al., 2024; Philipp et al.,


2025). There may additionally be a construction cost: for a given area, many small


vessels present far more wall surface than one large vessel, since perimeter scales


less steeply than area, so consolidating metaxylem may reduce the lignified wall built


per unit of conductive area and be more economical in carbon. We treat these


ecological and economic interpretations as provisional, since our functional evidence


is confined to the modelling.


This work uncovers a previously undescribed developmental transition in cereal roots


and provides a validated cell-resolved atlas of different states of the adventitious root


tip. This incident of change in vascular patterning is developmentally novel and, on the


evidence of the modelling, physiologically consequential, both in understanding the


consequence of vascular change in adventitious roots, and in understanding the


functional differences of adventitious and seminal roots. Consideration of this


phenomenon will be essential in future of how vascular pattern is built and maintained


across the life of cereal roots and the consequences thereof.


**Author Contributions:** D.H.J.: conceptualization, data curation, formal analysis,
investigation, methodology, visualization, writing-original draft; M.Sö.: data curation,
formal analysis, investigation, methodology, visualization, writing-review &amp; editing;
D.D.: investigation, resources, methodology; C.T.: investigation, resources,
methodology; J.B.C.: investigation (hydraulic modelling), formal analysis, data
curation, methodology, writing-review &amp; editing; A.W.: conceptualization, data
curation, methodology, writing-review &amp; editing; M.M.: investigation; M.Sc.:
investigation, resources, methodology, writing-review &amp; editing; D.Z.: investigation,
writing-review &amp; editing; R.G.: investigation, writing-review &amp; editing; G.L.: writingreview &amp; editing; B.P.: writing-review &amp; editing; V.C.: writing-review &amp; editing; M.B.:
writing-review &amp; editing; D.R.: project administration, funding acquisition, writingreview &amp; editing; H.M.S.: conceptualization, funding acquisition, supervision, project
administration, methodology, writing-review &amp; editing. All authors reviewed and
approved the final manuscript.


**Supplementary Material**


Table S1 List of genotypes used in this study


Table S2 Mean, standard deviation and n for each species, root class and axial
position, for all measured vascular traits (Figures 1 and 2).


29


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


Table S3 Subcellular hydraulic parameters and apoplastic barrier configuration used

for MECHA simulations.


Table S4 Root architectural and hydraulic parameters used for whole-root hydraulic

simulations.


Table S5 Whole-root-system architecture of hydroponically grown wheat, by node
and axial position (mean, SD, n).


Table S6 Cell-type marker genes used to validate the spatial annotation, with the
tissue each marks, its source and discrimination statistics (Figure 4).


Table S7 Bulk RNA-seq: raw read counts per sample, DESeq2 contrasts, and the
candidate-filtering stage each gene reaches (Figure 3).


Table S8 Hydraulic simulation outputs for all 405 root prototypes (Figure 5).


**Funding**


This study was supported by the Grains Research and Development Corporation
(GRDC) through grant "Root structure and function traits: Overcoming the root
phenotyping bottleneck in cereals (UOQ2312-009RTX) (D.H.J), the Leibniz
Association (Resilient Roots to H.M.S.), and the European Union (ERC, 101162856,
FATE) (H.M.S.). JCBC is funded by Deutsche Forschungsgemeinschaft (DFG,
German Research Foundation) (SFB 1502/1–2022, Projektnummer: 450058266. DZ
is the recipient of an Australian Research Council Early Career Industry Fellowship
(IE230100179, “Drought tolerance in sorghum: the roots of the solution”) and funded
by Australian Government. VC is a Research Associate of the Belgian Fonds de la
Recherche Scientifque (FNRS), co-funded by the European Union (ERC grant i
101043083). The views and opinions expressed are, however, those of the author(s)
only and do not necessarily refect those of the European Union or the European l
Research Council. Neither the European Union nor the granting authority can be held
responsible.


**Conflicts of interest**


The authors declare no conflicts of interest.


**Data availability statement**


Data available in article supplementary material and upon request.


30


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


**References**


**Abraham-Juárez MJ, Schrager-Lavelle A, Man J, Whipple C, Handakumbura P,**


**Babbitt C, and Bartlett M** . Evolutionary Variation in MADS Box Dimerization Affects


Floral Development and Protein Abundance in Maize. Plant Cell. 2020:32(11):3408–


3424. https://doi.org/10.1105/tpc.20.00300


**Affortit P, Faye A, Jones DH, Benson E, Sine B, Burridge J, Ndoye MS, Barry L,**


**Moukouanga D, Barnard S, et al.** Root metaxylem area influences drought tolerance


and transpiration in pearl millet in a soil texture dependent manner. BioRxiv


2024:2024.11.09.622826. https://doi.org/10.1101/2024.11.09.622826


**Alemán**  - **Sancheschúlz, G., Anfodillo, T., Pérez**  - **Maussán, A. I., Ayala**  - **Hernández,**


**M. M., &amp; Olson, M. E.** Da Vinci's mischief: xylem conduits in the stems of woody plants


do not furcate. New Phyt. 2026: _250_ (5), 3001-3013. https://doi.org/10.1111/nph.71097


**Baca Cabrera JC, Jones DH, Vanderborght J, Behrend D, Schneider HM, and**


**Lobet G** . Root anatomical gradients and cultivar differences underlie variation in root


hydraulic properties in German winter wheat. J Exp Bot. 2026a:erag177.


https://doi.org/10.1093/jxb/erag177


**Baca Cabrera JC, Vanderborght J, Couvreur V, Behrend D, Gaiser T, Nguyen TH,**


**and Lobet G** . Root hydraulic properties: An exploration of their variability across


scales. Plant Direct. 2024:8(4):e582. https://doi.org/10.1002/pld3.582


**Barry L, Cabrera JCB, Lucas M, Lobet G, Boursiac Y, and Grondin A** . Role of


xylem in root hydraulics: Functionality and implications for drought adaptation.


Quantitative Plant Biology. 2025a:6:e42. https://doi.org/10.1017/qpb.2025.10026


**Bartlett ME** . Changing MADS-Box Transcription Factor Protein–Protein Interactions


as a Mechanism for Generating Floral Morphological Diversity. Integr Comp Biol.


2017:57(6):1312–1321. https://doi.org/10.1093/icb/icx067


**Baumberger M, Haas B, Sivakumar S, Ludwig M, Meyer N, and Meyer H** . High

resolution soil temperature and soil moisture patterns in space, depth and time: An


interpretable machine learning modelling approach. Geoderma. 2024:451:117049.


https://doi.org/10.1016/j.geoderma.2024.117049


31


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


**Behrend D, Nguyen T, Hüging H, Cabrera JB, Lobet G, Seidel S, Srivastava A,**


**Bazzo CG, Kramer N, Nachtweide P, et al.** Biomass partitioning and canopy


architecture of six German winter wheat cultivars released between 1895 and 2002.


2025.


**Berleth T, Mattsson J, and Hardtke CS** . Vascular continuity and auxin signals.


Trends in Plant Science. 2000:5(9):387–393. https://doi.org/10.1016/S1360

1385(00)01725-8


**Bouda M, Brodersen C, and Saiers J** . Whole root system water conductance


responds to both axial and radial traits and network topology over natural range of trait


variation. Journal of Theoretical Biology. 2018:456:49–61.


https://doi.org/10.1016/j.jtbi.2018.07.033


**Brady SM, Long TA, and Benfey PN** . Unraveling the Dynamic Transcriptome. Plant


Cell. 2006: **18** (9):2101–2111. https://doi.org/10.1105/tpc.105.037572


**Brady SM, Orlando DA, Lee J-Y, Wang JY, Koch J, Dinneny JR, Mace D, Ohler U,**


**and Benfey PN** . A High-Resolution Root Spatiotemporal Map Reveals Dominant


Expression Patterns. Science. 2007:318(5851):801–806.


https://doi.org/10.1126/science.1146265


**Brady SM, Zhang L, Megraw M, Martinez NJ, Jiang E, Yi CS, Liu W, Zeng A,**


**Taylor-Teeples M, Kim D, et al.** A stele-enriched gene regulatory network in the


Arabidopsis root. Mol Syst Biol.2011:7(1):MSB2010114.


https://doi.org/10.1038/msb.2010.114


**Cao C, Guo S, Deng P, Yang S, Xu J, Hu T, Hu Z, Chen D, Zhang H, Navea IP, et**


**al.** The BEL1-like homeodomain protein OsBLH4 regulates rice plant height, grain


number, and heading date by repressing the expression of OsGA2ox1. The Plant


Journal. 2024:119(3):1369–1385. https://doi.org/10.1111/tpj.16857


**Clément C, Schneider HM, Dresbøll DB, Lynch JP, and Thorup-Kristensen K** .


Root and xylem anatomy varies with root length, root order, soil depth and environment


in intermediate wheatgrass (Kernza®) and alfalfa. Ann Bot. 2022a:130(3):367–382.


https://doi.org/10.1093/aob/mcac058


32


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


**Couvreur V, Faget M, Lobet G, Javaux M, Chaumont F, and Draye X** . Going with


the Flow: Multiscale Insights into the Composite Nature of Water Transport in Roots.


Plant Physiol. 2018a:178(4):1689–1703. https://doi.org/10.1104/pp.18.01006


**De Rybel B, Mähönen AP, Helariutta Y, and Weĳers D** . Plant vascular


development: from early specification to differentiation. Nat Rev Mol Cell Biol.


2016a:17(1):30–40. https://doi.org/10.1038/nrm.2015.6


**De Rybel B, Möller B, Yoshida S, Grabowicz I, Barbier de Reuille P, Boeren S,**


**Smith RS, Borst JW, and Weĳers D** . A bHLH Complex Controls Embryonic Vascular


Tissue Establishment and Indeterminate Growth in Arabidopsis. Developmental Cell.


2013:24(4):426–437. https://doi.org/10.1016/j.devcel.2012.12.013


**Dhar S, Park J, and Lee J-Y** . In-and-out signals for the root vascular tissue


development. Current Opinion in Plant Biology. 2026a:91:102875.


https://doi.org/10.1016/j.pbi.2026.102875


**Etchells JP, Smit ME, Gaudinier A, Williams CJ, and Brady SM** . A brief history of


the TDIF-PXY signalling module: balancing meristem identity and differentiation during


vascular development. New Phytologist. 2016:209(2):474–484.


https://doi.org/10.1111/nph.13642


**Gallagher JP, Man J, Chiaramida A, Rozza IK, Patterson EL, Powell MM,**


**Schrager-Lavelle A, Multani DS, Meeley RB, and Bartlett ME** . GRASSY TILLERS1


(GT1) and SIX-ROWED SPIKE1 (VRS1) homologs share conserved roles in growth


repression. Proceedings of the National Academy of Sciences.


2023:120(51):e2311961120. https://doi.org/10.1073/pnas.2311961120


**Giraud M, Gall SL, Harings M, Javaux M, Leitner D, Meunier F, Rothfuss Y, van**


**Dusschoten D, Vanderborght J, Vereecken H, et al.** CPlantBox: a fully coupled


modelling platform for the water and carbon fluxes in the soil–plant–atmosphere


continuum. in silico Plants. 2023:5(2):diad009.


https://doi.org/10.1093/insilicoplants/diad009


**Gowik U and Westhoff P** . The Path from C3 to C4 Photosynthesis. Plant Physiol.


2011a:155(1):56–63. https://doi.org/10.1104/pp.110.165308


33


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


**Hacke UG and Sperry JS** . Functional and ecological xylem anatomy. Perspectives


in Plant Ecology, Evolution and Systematics. 2001a:4(2):97–115.


https://doi.org/10.1078/1433-8319-00017


**Harrison Day BL, Johnson KM, Tonet V, Bourbia I, Blackman C, and Brodribb TJ**


. The root of the problem: diverse vulnerability to xylem cavitation found within the root


system of wheat plants. New Phytologist. 2023:239(4):1239–1252.


https://doi.org/10.1111/nph.19017


**Heymans A, Couvreur V, LaRue T, Paez-Garcia A, and Lobet G** . GRANAR, a


Computational Tool to Better Understand the Functional Importance of Monocotyledon


Root Anatomy. Plant Physiol. 2020a:182(2):707–720.


https://doi.org/10.1104/pp.19.00617


**HOCHHOLDINGER F, WOLL K, SAUER M, and DEMBINSKY D** . Genetic Dissection


                                       of Root Formation in Maize (Zea mays) Reveals Root type Specific Developmental


Programmes. Ann Bot. 2004:93(4):359–368. https://doi.org/10.1093/aob/mch056


**Hochholdinger F, Yu P, and Marcon C** . Genetic Control of Root System


Development in Maize. Trends in Plant Science. 2018:23(1):79–88.


https://doi.org/10.1016/j.tplants.2017.10.004


**Huĳser P and Schmid M** . The control of developmental phase transitions in plants.


Development. 2011:138(19):4117–4129. https://doi.org/10.1242/dev.063511


**Iii GPWG** . A nuclear phylogenomic tree of grasses (Poaceae) recovers current


classification despite gene tree incongruence. New Phytologist. 2025:245(2):818–834.


https://doi.org/10.1111/nph.20263


**Klein H, Gallagher J, Demesa-Arevalo E, Abraham-Juárez MJ, Heeney M, Feil R,**


**Lunn JE, Xiao Y, Chuck G, Whipple C, et al.** Recruitment of an ancient branching


program to suppress carpel development in maize flowers. Proceedings of the


National Academy of Sciences.2022:119(2):e2115871119.


https://doi.org/10.1073/pnas.2115871119


34


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


**Jones DH, Baca Cabrera JC, Behrend D, Wells DM, Swift JF, Atkinson JA, Schön**


**M, Lobet G, Hanlon MT, and Schneider HM** . The Rapid Anatomics Tool (RAT): A


low-cost root anatomical phenotyping platform reveals changes in root anatomy along


the root axis. Plant Phenomics. 2026:8(1):100150.


https://doi.org/10.1016/j.plaphe.2025.100150


**Jones DH, Kajala K, Kawa D, Lopez-Valdivia I, Kreszies T, and Schneider HM** .


The root cortex of the Poaceae: a diverse, dynamic, and dispensable tissue. Plant


Soil. 2025:514(2):1627–1662. https://doi.org/10.1007/s11104-025-07498-0


**Kadam NN, Yin X, Bindraban PS, Struik PC, and Jagadish KSV** . Does


Morphological and Anatomical Plasticity during the Vegetative Stage Make Wheat


More Tolerant of Water Deficit Stress Than Rice? Plant Physiol. 2015a:167(4):1389–


1401. https://doi.org/10.1104/pp.114.253328


**Ke Y, Pujol V, Staut J, Pollaris L, Seurinck R, Eekhout T, Grones C, Saura-**


**Sanchez M, Van Bel M, Vuylsteke M, et al.** A single-cell and spatial wheat root atlas


with cross-species annotations delineates conserved tissue-specific marker genes


and regulators. Cell Reports. 2025:44(2):115240.


https://doi.org/10.1016/j.celrep.2025.115240


**Klein SP, Kaeppler SM, Brown KM, and Lynch JP** . Integrating GWAS with a gene


co-expression network better prioritizes candidate genes associated with root


metaxylem phenes in maize. The Plant Genome. 2024a:17(3):e20489.


https://doi.org/10.1002/tpg2.20489


**Klein SP, Schneider HM, Perkins AC, Brown KM, and Lynch JP** . Multiple


Integrated Root Phenotypes Are Associated with Improved Drought Tolerance. Plant


Physiol. 2020:183(3):1011–1025. https://doi.org/10.1104/pp.20.00211


**Koehler T, Kim Y, Tung S-Y, Heymans A, Tyborski N, Steiner F, Wild AJ, Pausch**


**J, Ahmed MA, and Schneider HM** . Node of origin matters: comparative analysis of


soil water limitation effects on nodal root anatomy in maize (Zea mays). Ann Bot.


2025:136(5–6):1031–1046. https://doi.org/10.1093/aob/mcaf075


35


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


**Korsunsky I, Millard N, Fan J, Slowikowski K, Zhang F, Wei K, Baglaenko Y,**


**Brenner M, Loh P, and Raychaudhuri S** . Fast, sensitive and accurate integration of


single-cell data with Harmony. Nat Methods. 2019:16(12):1289–1296.


https://doi.org/10.1038/s41592-019-0619-0


**Kumar R, Kushalappa K, Godt D, Pidkowich MS, Pastorelli S, Hepworth SR, and**


**Haughn GW** . The Arabidopsis BEL1-LIKE HOMEODOMAIN Proteins SAW1 and


SAW2 Act Redundantly to Regulate KNOX Expression Spatially in Leaf Margins. Plant


Cell. 2007:19(9):2719–2735. https://doi.org/10.1105/tpc.106.048769


**Liu Z, Giehl RFH, Hartmann A, Hajirezaei MR, Carpentier S, and von Wirén N** .


Seminal and Nodal Roots of Barley Differ in Anatomy, Proteome and Nitrate Uptake


Capacity. Plant Cell Physiol. 2020:61(7):1297–1308.


https://doi.org/10.1093/pcp/pcaa059


**Love MI, Huber W, and Anders S** . Moderated estimation of fold change and


dispersion for RNA-seq data with DESeq2. Genome Biol. 2014:15(12):550.


https://doi.org/10.1186/s13059-014-0550-8


**Marjamaa K, Kukkola EM, and Fagerstedt KV** . The role of xylem class III


peroxidases in lignification. J Exp Bot. 2009:60(2):367–376.


https://doi.org/10.1093/jxb/ern278


**McCahill IW, Abushal LT, Khahani B, Probert CF, Flockhart EL, Gregory GA, Li**


**EZ, Zhang Y, Baumgart LA, O’Malley RC, et al.** Shoring up the base: The


development and regulation of cortical sclerenchyma in grass nodal roots. Plant


Physiol. 2025:199(1):kiaf215. https://doi.org/10.1093/plphys/kiaf215


**Meunier F, Couvreur V, Draye X, Vanderborght J, and Javaux M** . Towards


quantitative root hydraulic phenotyping: novel mathematical functions to calculate


plant-scale hydraulic parameters from root system functional and structural traits. J


Math Biol. 2017:75(5):1133–1170. https://doi.org/10.1007/s00285-017-1111-z


**Meunier F, Heymans A, Draye X, Couvreur V, Javaux M, and Lobet G** . MARSHAL,


a novel tool for virtual phenotyping of maize root system hydraulic architectures. in


silico Plants. 2020a:2(1):diz012. https://doi.org/10.1093/insilicoplants/diz012


36


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


**Miki Y, Saito S, Niki T, and Gladish DK** . Origin and Development of Phloem and


Xylem as Revealed in 3D Models of the Vascular Cylinder in Primary Roots of Oryza


sativa L. cv. Hitomebore. Plants. 2026a:15(4):607.


https://doi.org/10.3390/plants15040607


**Montiel G, Gaudet M, Laurans F, Rozenberg P, Simon M, Gantet P, Jay-Allemand**


**C, and Breton C** . Overexpression of MADS-box Gene AGAMOUS-LIKE 12 Activates


Root Development in Juglans sp. and Arabidopsis thaliana. Plants. 2020:9(4):444.


https://doi.org/10.3390/plants9040444


**Ohashi-Ito K, Iwamoto K, and Fukuda H** . LONESOME HIGHWAY-TARGET OF


MONOPTEROS5 transcription factor complex promotes a predifferentiation state for


xylem vessel differentiation in the root apical meristem by inducing the expression of


VASCULAR-RELATED NAC-DOMAIN genes. New Phytologist. 2024a:242(3):1146–


1155. https://doi.org/10.1111/nph.19670


**Ouyang W, Yin X, Yang J, and Struik PC** . Comparisons with wheat reveal root


anatomical and histochemical constraints of rice under water-deficit stress. Plant Soil.


2020:452(1):547–568. https://doi.org/10.1007/s11104-020-04581-6


**Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, Blondel**


**M, Müller A, Nothman J, Louppe G, et al.** Scikit-learn: Machine Learning in Python.


2018. https://doi.org/10.48550/arXiv.1201.0490


**Peralta Ogorek LL, Gao Y, Farrar E, and Pandey BK** . Soil compaction sensing


mechanisms and root responses. Trends in Plant Science. 2025:30(5):565–575.


https://doi.org/10.1016/j.tplants.2024.10.014


**Philipp L, Sünnemann M, Schädler M, Blagodatskaya E, Tarkka M, Eisenhauer**


**N, and Reitz T** . Soil depth shapes the microbial response to land use and climate


change in agroecosystems. Applied Soil Ecology. 2025:209:106025.


https://doi.org/10.1016/j.apsoil.2025.106025


**Reeger JE, Wheatley M, Yang Y, and Brown KM** . Targeted mutation of transcription


factor genes alters metaxylem vessel size and number in rice roots. Plant Direct.


2021:5(6):e00328. https://doi.org/10.1002/pld3.328


37


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


**Richards R and Passioura J** . A breeding program to reduce the diameter of the


major xylem vessel in the seminal roots of wheat and its effect on grain yield in rain

fed environments. Australian Journal of Agricultural Research. 1989a:40(5):943–950.


https://doi.org/10.1071/AR9890943


**Rodriguez-Alonso G, Matvienko M, López-Valle ML, Lázaro-Mixteco PE,**


**Napsucialy-Mendivil S, Dubrovsky JG, and Shishkova S** . Transcriptomics insights


into the genetic regulation of root apical meristem exhaustion and determinate primary


root growth in Pachycereus pringlei (Cactaceae). Sci Rep. 2018:8(1):8529.


https://doi.org/10.1038/s41598-018-26897-1


**Ryan, PR, Delhaize, E, Watt, M and Richardson, AE** . Plant roots: understanding


structure and function in an ocean of complexity. Ann Bot. 2016:118(4):555-559.


https://doi.org/10.1093/aob/mcw192


**Saito S, Niki T, and Gladish DK** . Comparison of Promeristem Structure and


Ontogeny of Procambium in Primary Roots of Zea mays ssp. Mexicana and Z. mays


‘Honey Bantam’ with Emphasis on Metaxylem Vessel Histogenesis. Plants.


2019:8(6):162. https://doi.org/10.3390/plants8060162


**Sanderson J, Whitbread FC, and Clarkson DT** . Persistent xylem cross-walls reduce


the axial hydraulic conductivity in the apical 20 cm of barley seminal root axes:


implications for the driving force for water movement. Plant, Cell &amp; Environment.


1988:11(4):247–256. https://doi.org/10.1111/j.1365-3040.1988.tb01143.x


**Sawchuk MG and Scarpella E** . Polarity, Continuity, and Alignment in Plant Vascular


Strands. J Integr Plant Biol. 2013:55(9):824–834. https://doi.org/10.1111/jipb.12086


**Scarpella E** . The logic of plant vascular patterning. Polarity, continuity and plasticity


in the formation of the veins and of their networks. Curr Opin Genet Dev. 2017:45:34–


43. https://doi.org/10.1016/j.gde.2017.02.009


**Schierenbeck M, Tawale AB, Lopez-Valdivia I, Jones DH, Wolf A, Linow P,**


**Trautewig C, and Schneider HM** . A thermal time framework drives coordinated


below- and above-ground development in temperate cereal crops.


2026:2026.04.24.720680. https://doi.org/10.64898/2026.04.24.720680


38


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


**Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T,**


**Preibisch S, Rueden C, Saalfeld S, Schmid B, et al.** Fĳi: an open-source platform


for biological-image analysis. Nat Methods. 2012:9(7):676–682.


https://doi.org/10.1038/nmeth.2019


**Schneider HM, Yang JT, Brown KM, and Lynch JP** . Nodal root diameter and node


number in maize (Zea mays L.) interact to influence plant growth under nitrogen stress.


Plant Direct. 2021:5(3):e00310. https://doi.org/10.1002/pld3.310


**Shishkova S, Rost TL, and Dubrovsky JG** . Determinate Root Growth and Meristem


Maintenance in Angiosperms. Ann Bot. 2008:101(3):319–340.


https://doi.org/10.1093/aob/mcm251


**Steudle E and Peterson CA** . How does water get through roots? J Exp Bot.


1998:49(322):775–788. https://doi.org/10.1093/jxb/49.322.775


**Strock CF, Burridge JD, Niemiec MD, Brown KM, and Lynch JP** . Root metaxylem


and architecture phenotypes integrate to regulate water use under drought stress.


Plant, Cell &amp; Environment. 2021:44(1):49–67. https://doi.org/10.1111/pce.13875


**Tapia-López R, García-Ponce B, Dubrovsky JG, Garay-Arroyo A, Pérez-Ruíz RV,**


**Kim S-H, Acevedo F, Pelaz S, and Alvarez-Buylla ER** . An AGAMOUS-Related


MADS-Box Gene, XAL1 (AGL12), Regulates Root Meristem Cell Proliferation and


Flowering Transition in Arabidopsis. Plant Physiol. 2008:146(3):1182–1192.


https://doi.org/10.1104/pp.107.108647


**Turco GM, Rodriguez-Medina J, Siebert S, Han D, Valderrama-Gómez MÁ,**


**Vahldick H, Shulse CN, Cole BJ, Juliano CE, Dickel DE, et al.** Molecular


Mechanisms Driving Switch Behavior in Xylem Cell Differentiation. Cell Reports.


2019:28(2):342-351.e4. https://doi.org/10.1016/j.celrep.2019.06.041


**Tyree MT and Zimmermann MH** . Hydraulic Architecture of Whole Plants and Plant


Performance. . In. Xylem Structure and the Ascent of Sap, MT Tyree and MH


Zimmermann, eds. (Springer Berlin Heidelberg: Berlin, Heidelberg), pp. 175–214.


https://doi.org/10.1007/978-3-662-04931-0_6


39


&#91;bioRxiv preprint doi: https://doi.org/10.64898/2026.11.19.767855; this version posted November 20, 2026. The copyright holder for this&#93;(https://doi.org/10.64898/2026.11.19.767855)

preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in
&#91;perpetuity. It is made available under a CC-BY-NC 4.0 International license.&#93;(http://creativecommons.org/licenses/by-nc/4.0/)


**Van Norman JM, Xuan W, Beeckman T, and Benfey PN** . To branch or not to


branch: the role of pre-patterning in lateral root formation. Development.


2013:140(21):4301–4310. https://doi.org/10.1242/dev.090548


**Wolf FA, Angerer P, and Theis FJ** . SCANPY: large-scale single-cell gene


expression data analysis. Genome Biol. 2018:19(1):15.


https://doi.org/10.1186/s13059-017-1382-0


**Wu H, Jaeger M, Wang M, Li B, and Zhang BG** . Three-dimensional distribution of


vessels, passage cells and lateral roots along the root axis of winter wheat (Triticum


aestivum). Annals of Botany. 2011:107(5):843–853.


https://doi.org/10.1093/aob/mcr005


**Xiao Y, Guo J, Dong Z, Richardson A, Patterson E, Mangrum S, Bybee S,**


**Bertolini E, Bartlett M, Chuck G, et al.** Boundary domain genes were recruited to


suppress bract growth and promote branching in maize. Science Advances.


2022:8(24):eabm6835. https://doi.org/10.1126/sciadv.abm6835


**Yang B, Sun Y, Minne M, Ge Y, Yue Q, Goossens V, Mor E, Callebaut B,**


**Bevernaege K, Winne JM, et al.** SPL13 controls a root apical meristem phase change


by triggering oriented cell divisions. Science. 2024:386(6723):eado4298.


https://doi.org/10.1126/science.ado4298


**Zimmermann U, Schneider H, Wegner LH, Wagner H-J, Szimtenings M, Haase A,**


**and Bentrup F-W** . What are the driving forces for water lifting in the xylem conduit?


Physiologia Plantarum. 2002:114(3):327–335. https://doi.org/10.1034/j.1399

3054.2002.1140301.x


**Zobel RW** . Soil Environment Constraints to Root Growth. . In. Limitations to Plant


Root Growth, JL Hatfield and BA Stewart, eds. (Springer: New York, NY), pp. 27–51.


https://doi.org/10.1007/978-1-4612-2894-3_2


40


