the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 01 Sep 2026
The more the better? A limited proportion of deep roots maximizes tree survival under drought in Mediterranean forest
Abstract. Deep roots are assumed to enhance tree survival during drought by providing access to water stored at depth. However, the respective roles of deep soil water storage (SWCDeep) and fine-root distribution (FineRootDistrib), which jointly determine access to deep water during drought, remain poorly understood. We assessed their relative importance during drought using the mechanistic hydraulic model SurEau-Ecos. The model was calibrated with leaf water potential and isotopic data collected in 2014 and 2015 for three species (Quercus ilex L., Fagus sylvatica L., and Abies alba Mill.) across two Mediterranean sites in France. Our results show trees predominantly rely on shallow water under non-limiting conditions, whereas SWCDeep mobilisation becomes essential during drought, accounting for 11–36 % of annual transpiration. Sensitivity analyses reveal a “rare-root strategy”, whereby a limited proportion of deep roots maximizes survival by enabling access to deep water while preventing rapid depletion of deep reserves. These findings highlight that fine root distribution, through its control on the rate of SWCDeep depletion, is a key driver of drought resistance, with an influence comparable to other major ecohydrological traits. Accounting for FineRootDistrib is therefore essential when assessing vulnerability to drought.
- Preprint
(2362 KB) - Metadata XML
-
Supplement
(1713 KB) - BibTeX
- EndNote
Status: open (until 17 Nov 2026)
- RC1: 'Comment on egusphere-2026-5115', Anonymous Referee #1, 22 Sep 2026 reply
Data sets
Tree xylem isotope analysis by mass and laser spectrometry S. Carriere et al. https://doi.org/10.17632/mvfvpmnxgs.3
Model code and software
Sureau - Ecos V2.0 Model J. Ruffault et al. https://forge.inrae.fr/urfm/sureau/-/tree/v2.2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 255 | 106 | 62 | 423 | 64 | 70 | 59 |
- HTML: 255
- PDF: 106
- XML: 62
- Total: 423
- Supplement: 64
- BibTeX: 70
- EndNote: 59
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
General comments
This article sets out to investigate, using a mechanistic hydraulic model, the respective roles of deep soil water storage, on the one hand, and the distribution of fine roots, on the other, in the uptake of deep water by trees during periods of drought. The article is very clear, well-researched and generally interesting. The results show that deep water uptake, whilst of marginal importance under non-limiting conditions, accounts for a substantial proportion of annual transpiration during periods of drought. One of the main conclusions of this article is that deep water uptake by trees is a very important process that must be taken into account more explicitly and precisely in hydroecological modelling. Another major result of the modelling exercise presented in this paper highlights the existence of a ‘rare root’ strategy, whereby only a few deep roots would be involved in deep water uptake during periods of drought, leading to a slow and somewhat regulated depletion of deep water resources.
As regards the terminology chosen by the authors, I found that the indiscriminate use of the terms ‘root distribution’ and ‘root density’ is somewhat confusing, particularly at the beginning of the article. Ultimately, having read the text in its entirety, I believe that what the authors have in mind corresponds more to the vertical distribution of roots (which encompasses root density at any given soil depth). I would therefore recommend clarifying this point from the outset so that the reader does not find themselves asking the same questions as I did when I first began reading the article. The discussion also leaves the reader somewhat unsatisfied in that deep water resources, which are treated as a single deep root reservoir in the model, actually constitute a complex network of distributed reservoirs of varying capacities (and almost certainly contrasted soil water potentials), as shown in Figure 1; yet the local soil–root interactions that might explain (or provide alternative explanations for) the ‘rare root’ strategy are scarcely addressed if at all.
With regard to this ‘rare root’ strategy, the following question seems to arise: does this mechanism in fact rely solely on a limited number of roots or on a limited number of active absorption sites? What about the plasticity of the rhizosphere’s hydraulic properties (as defined by Carminati and Vetterlein) or of intra-plant signalling in such a regulatory process? Overall, this is a very good article which I enjoyed reading, and I strongly recommend that it be published as soon as the few minor points raised in this report have been addressed.
Specific comments
L32 : other ecological traits such as?
L62 : root distribution is referred to in the abstract (L30) but I agree with the statement made here that fine root density is more likely the parameter of importance, or more exactly that it is fine root density at depth that most likely matters.
L93 : what is the definition of a data-model? Mechanistic as mentioned in the abstract appears more appropriate.
L116: should read “divides”
L192: space is missing following full stop.
L200: gap-filled using which technique?
L246-247: was is done at regular monthly intervals? During which part of the year? please specify the dates (this could be referred to in a table).
L255: is “understory precipitation” throughfall?
L293: this moving average value of 5 days was chosen based on an experiment in which deuterated water was injected in the trunk of about 6 m tall trees. Yet in the scenarios tested here it is not unreasonable to imagine that the deep water reservoir (TAWdeep) encompasses a depth that exceeds the height of the trees. Therefore I wonder if and how longer transfer times between the deep reservoir and the canopy would affect the results presented.
Fig3, panel a. For consistency, the legend should read “Measured”
L350-353: is there a figure showing this? It is not obvious from Fig. 3.
L355: where are fig. G1 and G8?
L369: what is P12_gs?
L389: lent? I would advise the use of gradual here.
L435-443: the “rare root strategy” might also corresponds to a “rare root activity” strategy, by which roots are not necessarily rare at depth, but only a rare fraction of them are actively involved in the process of water uptake depending on the actual local water potential conditions there are exposed to. Indeed, and I understand the inherent constraints of the modelling exercise presented and discussed her, the soil water potential in the deep layer and the corresponding root density within it is are both considered homogeneous, when obviously, this is cannot be the case, particularly in karstic environments. Therefore, one can hypothesize that many roots are grown in deep cracks and pockets of sediments of varying water retentions (discrete uptake sites), under a foraging strategy aimed at maximizing chances of accessing water resources, but that only a few of these local environments can sustain the tree water demand over long periods of time, resulting in a low proportion of the roots present at depth being involved in uptake at any given time during a drought episode, hence the apparent “rare root strategy”. Therefore, I believe that it would add value to the discussion to consider how the rare root hypothesis may correspond to mechanisms under which active water uptake sites are used by the trees depending on the depletion/availability of water at these discrete uptake sites, and how such mechanisms may be regulated through intra-plant signalling or as a result of increased water-holding capacity / altered hydraulic conductivity of the rhizosphere.
Interestingly I think, the process described here is also not dissimilar to the downward progressing uptake front as shallow water is progressively taken up, as observed and modelled by Doussan et al. in the early 2000s and which by way of consequence of the architectural architecture of the root systems investigated (both geometrically and hydraulically), resulted in a decreasing number of uptake sites as drought proceeded.
L456: yes, this is what the paper focusses on but this should me mentioned explicitly earlier.
L460: closing bracket and full stop missing here.
L461: should read “uses” here.
L461: should read “temporally restricted”.
L497-500: I am unsure what the point is as it is rather clear to me that any survival strategy will represent departure from optimal metabolism (?)
L519 & 523-524: yes, cf my comments on L435-443.
L527 : and also root mucilage mediated mechanisms.
L534-547 and Table 2: these are results of the study that should more naturally be presented under the “results” section than here in the discussion.
Table 2: These proportions are also in line with figures reported by Doody and Benyon (2010), although these authors did not present measurements made on an annual basis.
L558: was this really the objective? My understanding is that the paper investigates the role of fine root vertical distribution on deep water uptake during drought.
L568-569: this alone is a very significant and important result.