| 29 Apr 2026
Root-Type-Specific Water Uptake in Maize across Soil Texture and Moisture Gradients
Abstract. Understanding how distinct root types contribute to water uptake under variable soil conditions is crucial for improving crop water use efficiency. We quantified root water uptake in six-week-old maize grown in sandy or loamy soil under well-watered and water-limited conditions. Time-series neutron radiography combined with deuterated water labelling enabled estimation of effective radial water uptake of crown and seminal roots using a diffusion–convection model.
Plant development responded strongly to the soil environment, with the greatest reductions in transpiration, shoot height, and root length occurring under sandy-dry conditions. Root water uptake patterns varied with soil texture and water availability. Crown roots exhibited pronounced hydraulic plasticity, with radial fluxes approximately threefold higher in sandy than in loamy soil. In contrast, seminal roots showed no impact of soil conditions on root water uptake. Overall, crown roots showed two times higher radial fluxes than seminal roots.
These findings demonstrate how soil texture and moisture interact with root-type-specific hydraulic function, providing insights for breeding and modelling drought-resilient maize.
General comments
The manuscript describes an interesting study that tries to analyze the effect of soil texture and soil water content (SWC) on the uptake
and transport of D2O by crown and seminal roots of maize plants. The data analysis depends on a partly established model for uptake of D2O using neutron radiography. The observed variations/changes in D2O transport among root types on changes of soil water content are interpreted as
changes in radial and axial fluxes. The authors conclude that fluxes are mainly influenced by soil texture especially for crown roots, more so
than soil water content (SWC) but do so while assuming a fixed internal anatomical organization of the root cross-section, aside from differences in overall root diameter, even though the total flux (measured via weight loss) is nearly 30% less for loamy soil and more than 60% for sandy soil. Root lengths were
also affected but less so than transpiration. A rough estimate shows that the changes in transpiration and root lengths do not completely match up with
changes in axial and radial fluxes, about 20 to 25%. This is not discussed. Neither are the changes in shoot sizes discussed which may severely impact
radial and axial fluxes which I think should be addressed more fully. This may partly be caused by ignoring potential changes in root anatomy caused by
soil texture and/or low SWC within the model (see my point 1 in the specific comments). In addition, a more extensive discussion on the changes of local
SWC due to the D2O injeection is required as this disturbs the root water uptake pattern, the measurement is a perturbation, especially in sandy soil that
is drier and therefore experiences a larger local SWC change.
Specific comments
1.I am perfectly fine with the idea that the model is a significantly simplified version of reality. However, this simplification
very likely comes with a major issue. Here, internal root anatomy is assumed fixed, e.g. a constant 5% conductive xylem fraction, a constant
not derived from microscopy data. But seminal roots and crown or nodal roots typically do not have the same relative conductive
area and this area is not constant over time (the changing fraction of early metaxylem and late metaxylem over time). More importantly,
xylem organisation with the roots changes on soil texture and especially soil water content. See for a recent article with references:
https://doi.org/10.1002/csc2.70084 by Yutong Jiang and JK Whalen. Root anatomy can change substantially (e.g. changes in stele diameter,
changes in xylem bundle numbers, xylem sizes) with conditions but when they do the model used by the authors
does not allow for this as it is presented here. This implies that anatomical changes upon variations in soil texture or SWC are interpreted
in terms of changes in axial and/or radial fluxes. At the moment it is completely unclear how this affects the modelling results of this manuscript.
So, when I pose a counter-hypothesis stating that anatomical variations/changes between CR and SR upon differences in soil texture and SWC
cause the observed changes in D2O fluxes, how would the authors provide proof that this alternative hypothesis is less likely to be correct.
What is needed therefore is at minimum an analysis where the effect of changes of Ax, the xylem stele diameter, especially in relation to overall root
diameter are modelled, basically a sensitivity analysis, such that the importance of anatomical variations can be at least somewhat quantified. Better,
would be to perform microscopy on roots that have been grown under the investigated experimental conditions and quantify these anatomical changes
and insert these into the model.
2.The authors should explain why they chose this specific form of Eq.7 which is unclear to me. Also explain what to do when a turns negative.
How is this handled or is this an impossibiolity and az > -1 always imposed? (otherwise the solution is imaginary if b is a non-integer?)
Also, doesn't a have the dimension of 1/cm or 1/mm or 1/m, the inverse of z (not mentioned)? In case b is odd, the flux can be negative is az<-1
which I also find difficult to interpret. Maybe the authors can expand a bit on this.
3. The authors mention that the lower boundary condition for tracer concentration entering the root segments is assumed, not meaured. This may influence the fitted concentrations and so the fluxes. This is a point that requires some discussion.
4.Another issue is that it is not clear whether the authors looked at seminal and crown roots that include the lateral roots or not. Judging from the
reported average root diameters this might not be the case. This point needs to be clarified. Also, did the authors verify that the measured root lengths
per NT are the same or similar to standard measurements looking at excavated roots whose lengths have been quantified? For a older than 45 days maize plant (or is it 6 weeks?) root lengths of less than 10m appear quite low to me, I would have expected a root length of more than 30m or so. Lateral root diameters are significantly less than primary axis roots, so how is this treated in the model?
5.It is not clear from the experimental section how many plants were used per condition. In the Fig.3 caption it is indicated N=6 so I suppose 24 plants
were measured in total, correct? This needs to be mentioned clearly in the methods section. How many segments of CR and SR were measured per plant and how where these segments treated in the statistical analysis? What about the variation among individual segments among individual plants for both CR and SR as compared to variation between plants? Also, what are the estimated error margins for the modelled parameters, based on the assumption that the model assumption themselves are correct? It is not clear to me how the in-between plant variations compare with the in-between segments and within-model variations. So, how are the segments treated, as independent observations or not as this may lead to pseudoreplication?
6.It is reported that the plants were older than 45 days (!! in the methods section, the abstract says six-weeks-old which is somewhat inconsistent),
but as you measured 24 plants, I assume, and each measurement took at least 3h or so, what was the difference in age of the plants? I ask this as the metaxylem of maize plants develops over time (late metaxylem becomes more important over time).
Technical comments:
Equation 8 is a rearrangement of Equation 3 and is thereby somewhat redundant.