Drought Superimposes the Positive Effect of Silver Fir on Water Relations of European Beech in Mature Forest Stands

Round 1

Reviewer 1 Report

This manuscript compares the water relations of two forest stands with different composition, under natural and reduced water availability, using sap flow measurements. The Authors aim to understand if mixed stands are more or less water use efficient than pure stands and if stand resilience to water shortages can be increased by changing composition. The experiment was well planned (apart from a few drawbacks) and the topic of the paper is important to forest management in Central Europe and is well in the scope of this journal.

The methods, although not novel, appear to be sound and support the results. The paper is generally well-written, with just a few typos or grammatical errors that can easily be corrected (eg: lines 19; 155; 440). I am not sure if the way the authors chose to present their data is the most suitable to allow an easy comprehension of the results.

My main concerns are related to the number of trees sampled for sap flow measurements. How can the authors be sure that sampling 3 trees per treatment is enough to overcome the large uncertainty arising from between-tree variability in sap flow?

Comments and suggested edits:

Line 37:

Line 74: “in mixed plots during.” Sentence reads like something is missing.

Lines 99, 112, 156 and 312: why is supplementary data available on Tree Physiology online?

Line 106: Please specify if one or more roofs were installed in each stand.

Line 122 states that SWC was measured continuously at 10 cm depth. However, Figure 2 shows correlations between sap flow and soil moisture at different depths (80, 50, 25, 15 and 10 cm). What’s the source of this data?

Figure 1: In 2017, SWC in the pure beech rain excluded stand increased steadily after the re-wetting events, whilst in the other plots the patterns of variation were similar. Also, tree water use (pure beech) was similar between treatment and control after re-wetting. How do the authors explain this result other than by the longer transpiration period of fires (lines 444-445)?

Also, is it possible to combine SWC and rainfall data in the same chart?

Line 214 (Table 1): Gray shaded IDs are absent, it’s not clear which trees belong to each plot.

Lines 217-218 (Table 2 and 3): Is it possible that the experiment started too late to fully characterize the seasonal variation of fir water use (ie, missed the peak)?

Lines 324-329: Please link comments about evapotranspiration (and throughout the text also) to a figure/table or give values in the text.

Line 344: delete s in years.

Lines346-350: Can it also be reflecting effects of interspecific competition and the higher tree density in the mixed plot?

Line 378: “competitive reduction for beeches of resources”. This sentence is odd, consider rephrasing.

Lines 378-379,  386-387 and 405-406. In the same area, 0.1 ha, the pure stand has 36 beeches and the mixed stand has 33 beeches plus 19 firs. Even if beech and fir are competing for resources at different depths, beech number is similar and likely with no effect on intraspecific competition.

Table 3, Figure 4: Why are the June values (daily sap flow; stand water use and stand sap flow) for the pure beech control missing?

Figures need editing as the colour scheme is not always straightforward. In Figures 3 and 4, especially, it’s difficult to single out the variables.

Lines 457-469: I suggest moving this text to results section, detailing the origin of the values used in scheme I.

Author Response

Please see the attachment. 

Author Response File: Author Response.pdf

Reviewer 2 Report

Review of Manuscript Forests 60260

This is a fairly detailed report of the effects of stand composition and drought on water use by beech and silver fir trees. The experiment is part of a larger project on German forest management and in general was well conducted and analyzed. I have a few criticisms and suggestions for improvement. Grammatical/English corrections are listed at the end of my review.

I noticed that the pure stand subplot has 36 trees compared to the mixed plot with 52 trees. Yet the authors did not address this difference in tree density. How might that have affected their results? For instance, was light interception different due to the increased density in the mixed plot or was it a function of the differences in canopy structure between the beech and fir, or both?

Also, while I understand the complexity and cost of the infrastructure, statistically, the authors are performing an unreplicated study, as there is only one pure beech plot and one mixed beech-fir plot. Yet, there is no discussion of the implications for deriving their conclusions based on the lack of replication. I wish the authors would address this. For example, they can’t really conclude that the differences and similarities found are due to the composition of species in the plots because other confounding factors, such as soil type, location, history, and so on, are present. While I think the results make sense as presented, they need to address this issue nonetheless.

The authors measured soil moisture content at a depth of 10 cm, but report later that beech take up most water from a depth ranging between 20 and 80 cm. Fir, with their “taproot”, depend on even deeper water sources. Why didn’t the authors measure soil moisture at several depths and most importantly, deeper down, considering that these trees subsist on soil moisture from those depths?

I was glad to see the section where the authors compared their SFDs to previous studies, as this gives the reader more confidence in the results presented here. But has anyone in Germany thought about whether they need to calibrate their sapflow sensors? The initial calibration provided by Granier may not hold up among different species, especially if there are diel variations in stem water content (e.g., Changes in stem water content influence sap flux density measurements with thermal dissipation probes; By: Vergeynst, Lidewei L.; Vandegehuchte, Maurits W.; McGuire, Mary Anne; et al. TREES-STRUCTURE AND FUNCTION   Volume: 28   Issue: 3   Pages: 949-955   Published: JUN 2014). If the authors are going to make conclusions across species, it would behoove them to make sure that there are no calibration errors contributing to making an incorrect conclusion.

In the very first sentence of the introduction the authors state “Forest ecosystems are regarded rather insensitive to alterations of environmental factors….”. I take exception to this overly general statement. At Coweeta Hydrologic Lab, an insect infestation in the summer drastically and rapidly altered water use and nutrient cycling in highly detectable ways. That would not suggest to me that this ecosystem was “insensitive”. Cold spells at Hubbard Brook stimulate losses of N from the ecosystem, again not insensitive. And I could go on and on. I strongly suggest the authors rethink this opening statement and come up with a better one. Furthermore, the second half of the sentence, which is supposed to serve as a justification for the first part, doesn’t really make sense to me. Having an ecosystem respond more sensitively to perturbations, so as to restore stability/function, would make more sense than not responding, which would tend to drive an ecosystem far from steady-state conditions and could possibly result in a state shift.

It is not clear to me how the Spearman correlation coefficients were obtained. It is unusual to see a plot of “rho” versus time of day. Were correlations done by the hour, over the days of the experiment, and then plotted versus time on the graph? If so, is a bar graph more appropriate? Please better explain how these data were obtained. Normally I would have expected that variations in diel variations in VPD, for instance, would be the variations used to model their influence on SFD. I can’t figure out how one can get all these correlations across a day.

On line 312 the authors mention that supplementary material can be found in Tree Physiology. But as I understand it, this manuscript is headed for FORESTS, so perhaps this is a typo?

I agree that the differences in rooting behavior may be the key to the differences found in water use and response to drought. I’m not so sure about the hydraulic lift and its importance. Is there any evidence that that occurred in these plots?  Also, the authors are quick to assume that drought induces root cavitation which prevents recovery of transpiration in the beech after rewatering. But it could also be that the imposed drought caused the death of fine roots, which would be the most active root fraction with respect to water uptake, and that the failure to recover could have been due to a lag time before fine root production resumed. I think the authors need to include that as a possible mechanism in addition to just embolism.

In conclusion, this is a well done study, which could be improved for readers by addressing some of my criticisms and comments above.

Below, I detail some grammatical/English corrections that I think the authors should do:

Line 20 – change “adaptation” to “adaptive”

Line 21, revise as: “…measures, e.g., the introduction….”

It is not well explained here why German land managers want to reduce water use in forest stands. What is their goal? Increased wood production or decreased water consumption, or both?

Line 24 – it is “drier”, not “dryer”

Lines 29 and 39 – should be “beeches’” with an apostrophe to indicate the plural possessive mode

Line 40 – replace “not more” with “less”

Line 66 – insert “as” before “less”

Line 71 – put an apostrophe after “trees” as it is possessive in the plural

Line 74 – take out “during”

Lines 79-85 – this section is poorly worded. The hypotheses need substantial revision to become grammatically correct. They should be written in the future tense, as they are proposals of what will happen, not what has happened. It is also not clear that these are the best hypotheses, given how the discussion proceeds, or in the right order. Nonetheless, here are my suggestions:

Hypothesis 1: Transpiration in the mixed plot will be smaller than that in the pure beech plot because of the more isohydric behavior of the fir.

Hypothesis 2: Rain exclusion will affect the more shallowly rooted beech more so than the deep, tap-rooted firs which have more access to moisture in deeper soil layers.

(Perhaps this should be the first hypothesis and hypothesis 1 should be #2?).

Hypothesis 3: If hypotheses 1 and 2 are true, then rain exclusion will have less effect on the mixed beech-fir plot than the pure beech plot because of differences in water acquisition strategies of the beech and fir and also because of differences in crown architecture, which will alter rainfall interception and evaporation from the stand.      

Throughout the paper, the authors fail to separate units from their values (e.g., they write 100cm, which should properly be 100 cm). Please go throughout the paper and put a spaces between values and their units.

Line 106 – take out the comma after “both” – not necessary.

Line 230 – insert “to” before “only”

Line 254 – “exemplary” is a confusing term. What do you mean by this? I interpret that word to

mean “excellent”. Excellent for what?          

In the Discussion, first paragraph, the authors refer to mixed and pure plots, then later insert the species names to refer to those same plots. Pick one way to do it and be consistent.

Line 349 – replace “not more” with “less”

Line 420 – replace “explicable” with “explainable”

Line 445 – don’t cite an author directly using a number, rather, use their name and then reference the number of that citation. No one is named “68!

Supplemental Figure 2 – you don’t need to have a key that includes both the symbol and the line. Redundant. Take out the line keys.

What do the shaded areas mean in Supplemental Figures 3 and 4 mean? Not explained in figure legends.

Author Response

Please see the attachment. 

Author Response File: Author Response.pdf