Review of "Contrasting water use strategies of beech trees along two hillslopes with different slope and climate" submitted for publication in HESS by Ginevra Fabiani

Ginevra Fabiani and co-workers present a very interesting study on ecohydrological adaptation options of beech trees. They have conducted intense data collection along topographic sequences at two different sites (Luxembourg Ardennes and Italian Apennines) for about three vegetation seasons. Most noteworthy is the combination of local differences in water pools (stream, groundwater, rain, soil water and xylem water) approached quantitatively with monitoring sensors and samples analysed for stable water isotopes, and along such a topographic transect. The data are truly impressive and I can only congratulate the authors for the nice study. 

Without diminishing my high regard for this work, I still think that the manuscript deserves substantial revisions to really work out its strengths and novelties. This is especially so since the presented data and analysis comes with substantial overlap to their previous study looking at the Luxembourg site only (Fabiani et al. 2021). Although properly cited, it took me a while to become clear about this. 

Since the manuscript tries to reach quite far (water use strategies, physiological responses as a function of water availability, topographic and climate effects…) I am not quite convinced that the scope of the presented analyses is really suitable to make this case.  Although I think to be quite within the ecohydrological bubble dealing with beeches and water uptake, I found it difficult to understand your argumentation lines to work out the research questions or how these questions really relate to this extraordinary dataset about state dynamics, fluxes and isotopic compositions. One central issue to me stands with your term of “water limitation”, which would always define in terms of potentials not pools (see also Novick et al. 2022).

Moreover the study setup has made some of the analyses rather difficult to substantiate: 

While the non-overlapping temporal coverage might be of minor importance with the climate in Tuscany not strongly coupled to the climate in western Luxembourg, 2019 and 2020 have been strongly characterised by a multi-year drought across Europe (e.g. Rakovec et al. 2022) to which especially older forest stands could not easily adapt at many places. This setting is also visible in the data table 2, where even Tuscany received more rain in 2021 than Luxembourg in 2020. To what degree could this affect your conclusions?

With respect to the two selected sites I see the issues of a) the slopes are facing in opposite direction (entailing not only different exposition but also that the sap-flow sensors have to be placed at opposite sides, see e.g. Renner et al. 2016 in the same Luxembourgish catchment), b) geologically and pedologically contrasting conditions (shallow gravelly, silty soils on slate vs. sandy loams on sandstone), and c) topographically different settings (slope steepness and length differ, leading to questions about how well the delineated sections footslope, midslope and hilltop/plateau can be equally defined at both sites).

Maybe these are all elements you have thought of, but which I could not find sufficiently addressed in the manuscript. Since in the end your analyses boil down to statistical comparisons between the relative differences along each transect and between both sites in the 2+1 years, I am not really convinced that this really corroborates water use strategies (as suggested in the title), nor that the conceptualisation of ecohydrological process (fig. 9) is really the lesson to learn. There are few places with more intense and long-lasting monitoring of ecohydrological compartments than the Weierbach and we have come to much more precise concepts already. Moreover, we studied root water uptake and sap flow under contrasting geology/soil close to the Weierbach and did not see much effect on sap flow despite pronounced differences in soil moisture dynamics (Jackisch et al. 2019, and we unfortunately lacked isotopic analyses).

With respect to the analyses, I find it difficult to follow your conclusions, too. E.g. the average temperature appears to compare the timeseries March 2019-Oct 2020 (Lux) with the timeseries Jan 2021-Oct 2021 (It). Hence the population of the variables are neither aligned to hydrological years nor to the same period. Does this cause the differences you work out in the averages and maybe even in your distributions/statistical tests? Are the slightly higher midlope sap velocities  in the Weierbach catchment consistent in both years and across the active flow field (not just a matter of the sampled fraction)? What are effects of the data subset and what are really things we can attribute to the physical/physiological reality?

I suggest to revise the manuscript to really work out its biggest strengths: The combined data for the different ecohydrological state dynamics and isotope rations. I suspect that a slightly more detailed analysis can hold so much more interesting results which avoid most of the issues pointed out above. This should also help to clearly differentiate this manuscript from its 2021 predecessor. Further, I suggest to revise the title, since water use strategies would in my view require even more detailed physiology-related data and analyses. 

Even if you disagree with my suggestion, I think that the manuscript can substantially gain from a much more intense and well-structured argumentation about the relationships between REW (relative extractable water), VPD (vapour pressure deficit), sap velocity and isotopic characteristics. All these aspects are largely lacking in your introduction. When they appear in the methods section, they are described without much context and interrelation. In the end, the paper appears to be quite lengthy for few simple mixed effects models with different time series and with the rather broad conceptual sketch in the end. I doubt that this was your intended takeaway.

I will attach my handwritten comments to your paper as pdf (simply due to time-constraints, please accept my apologies). I did not note much to the discussion and conclusions since I suppose that they will be largely reworked anyways. I sincerely hope that you will receive my suggestions as constructive remarks helping to advance your manuscript. I can only repeat my sincere regards for your work and I really think that it deserves to be presented in a more rigorously worked out and condensed manner.

All the best.

Conrad

Fabiani, G., Schoppach, R., Penna, D., and Klaus, J.: Transpiration patterns and water use strategies of beech and oak trees along a hillslope, Ecohydrology, 15, https://doi.org/10.1002/eco.2382, 2022.

Jackisch, C., Knoblauch, S., Blume, T., Zehe, E., and Hassler, S. K.: Estimates of tree root water uptake from soil moisture profile dynamics, Biogeosciences Discuss, 2019, 1–25, https://doi.org/10.5194/bg-2019-466, 2019.

Novick, K. A., Ficklin, D. L., Baldocchi, D., Davis, K. J., Ghezzehei, T. A., Konings, A. G., MacBean, N., Raoult, N., Scott, R. L., Shi, Y., Sulman, B. N., and Wood, J. D.: Confronting the water potential information gap, Nat Geosci, 15, 158–164, https://doi.org/10.1038/s41561-022-00909-2, 2022.

Rakovec, O., Samaniego, L., Hari, V., Markonis, Y., Moravec, V., Thober, S., Hanel, M., and Kumar, R.: The 2018–2020 Multi-Year Drought Sets a New Benchmark in Europe, Earth’s Future, 10, https://doi.org/10.1029/2021ef002394, 2022.

Renner, M., Hassler, S. K., Blume, T., Weiler, M., Hildebrandt, A., Guderle, M., Schymanski, S. J., and Kleidon, A.: Dominant controls of transpiration along a hillslope transect inferred from ecohydrological measurements and thermodynamic limits, Hydrol. Earth Syst. Sci., 20, 2063–2083, https://doi.org/10.5194/hess-20-2063-2016, 2016.

This manuscript presents an interesting comparative study for water use strategies of beech trees along two climatically and topographically differing hillslope transects. It is accompanied by two extensive data sets encompassing sap velocities and stable water isotopic signatures of precipitation, soil- and xylem water.

Overall, the paper is well written its storyline is coherent. However, before I can recommend the paper for final acceptance, there are some points that should be addressed.

line 95: Are you talking of a species "space oak"? Then please provide a scientific name, because I'm unable to find anything by that name. Or is "space" used as an adjective? Then please rephrase before other people also start looking for "space oak".

line 97: Could you provide a scientific name for those "Scotch broom bushes"?

line 137: Could you also add the information at which xylem depths the two measurement points of your SFM1 sensors ended up after the installation?

Fig. 1: It would be nice to see both countries within the same overview map, which should also include a scale bar.

In case one overview map for both countries does not seem feasible to you, you should at least add scale bars to both countries' outlines.

line 143: As far as I know, the wounding correction causes a linear scaling of the original measurements.

In Section 2.6 you normalize all sap velocities. So this step here seems superfluous... 

line 149: How exactly did you determine the sapwood-hardwood boundary? For beech that should not be so obvious. Or ist it?

In Fabiani2022a you wrote that  "Beech tends to develop HW only after decades [...];thus, we sampled cores for beech with an average length of 7 cm which was split into two, representative for outer and inner SW." Assuming we are talking about the same sampling campaign, does that mean, that you simply treated the outer 3.5 cm as the relevant outer sapwood and ignored the rest of your cores?

line 179: Previously you said, that the bags already have been  heat sealed. How did you inflate the previously heat sealed bags?

line 194: How did you determine the daily mean velocity of one tree if your sensors are giving you values for outer and inner xylem? According to Fabiani2022b you averaged the two values - I think you should also mention that somewhere in this paper.

line 204: I suppose you are talking about precipitation, but I would suggest to actually state that here.

line 214: Your method section does not mention how you determined wood moisture content - where do those values come from?

Fig. 2: Please add the respective years to your x-axes.

Fig. 3: Instead of using three colors that are very similar to your location color codes, I would suggest to introduce a new color coding for Weierbach2019, Weierbach2020 and Lecciona2021 (e.g. light purple, dark purple and yellow).

Fig. 5: In line 131 you said that there was only one tree per location at the Weierbach catchment. So where does your "average deviation between trees at the same location" in this figure come from?

Fig. 6: The current combination of subfigure headers is a bit awkward. I would suggest to have the three months as column headers once at the very top row of your figure. Then each subfigure could have a color coded header (see my suggestion for Fig.3) to indicate Weierbach2019, Weierbach2021 and Lecciona2021.

Fig. 7: In this Figure, the 2019 above all figure columns is just wrong, since the second row surely refers to 2020 and the third to 2021. I would suggets to follow my hints to Fig. 6. Generally, I think that Fig.6 and Fig.7 take more space in the main part of the paper than they should. Please consider moving one of them to the appendix.

line 292: Since all soil water signatures in Fig.8 are displayed in the same color it is impossible to see that there.

line 293: I would expect to find heavier summer precipitation somewhere along the "upper right" part of the LMWL but not above it. Can you rule out any issues with the soil water isotope measurements? Or could there be an issue with the data basis of your LMWL? Where (coordinates, elevation), when (time period, sampling intervals) and how (type of sampler, measures against evaporation) were the isotopic signatures of precipitation measured? All that might be crucial information for future users of your impressive data sets.

Fig. 8: In this figure, you are using different colors for your three topographic locations than in the previous figures. Instead, rain and xylem are occupying your midslope and footslope colors. Please consider to harmonize your use of colors across all figures. This will facilitate a more intuitive understanding of your visualizations.

line 317: Now that you brought the topic up: Are there any observations available of what happend in 2021 in the Weierbach catchment?

line 319: I guess you meant to talk about "lateral subsurface flow" (flow is missing)

line 357: Did you account for different levels of solar radiation, when you diagnosed an incomplete recovery of sap velocities after rainfall in September?

line 367-396: The trees might be adapted to capitalize water sources from different soil depths (or other potential sources) and in case these sources have distinct isotopic signatures, the isotopic signatures of xylem water may be evidence for that adaption. But a specific adaption "to capitalize water sources with different isotopic composition" seems quite unlikely to me...

line 375: Could contrasting measurement techniques for soil and xylem water isotopic signatures be a reason for that mismatch?

Fig. A2: This doesn't have to be a figure. A table might be more efficient to convey the displayed information.

Fig. A3: Why are you showing one average over all soil moisture measurements for the Weierbach, but location averages for the Lecciona catchment? Once again, where do your wood mositure contents come from? If you derived them from your xylem cores - why don't they have confidence intervals like your xylem water isotopic measurements?. Also: I think the left y-axis label is clipped (i-dot and the top of the d seem to be missing).

Fig. AA & A5: The symbol sizes between the two Figures vary unexplicably. Have you considerd to merge the two figures with two subfigures each into one with four subfigures? Personally, I find those time series more intersting than Fig.7 or even Fig. 8. in the main section of the paper. Maybe you can swap them with one of them?