Quick et al. present a study of tree water use under elevated CO2 in a FACE experiment. BIFoR is a unique FACE experiment in that it is located in an old growth forest and could potentially offer better insight into the future of global forests than previous FACE experiments in young plantation forests. The question of plant-water relations under high CO2 is a critical one and includes processes that are often uncertain so data such as those presented in this paper are extremely valuable.

While it is obvious that a lot of effort has gone into data collection, the paper is very poorly written, with little focus or in depth analysis of what was measured. While two hypotheses are raised in the introduction, these neither follow from the rest of the introduction nor are they followed through in the rest of the paper. The statistical analysis is lacking and it is hard to understand what the conclusions of this study are. I hope that my comments can help improve the manuscript so that this very valuable data can be published.

Major comments:

• Abstract - this is very detailed and it is unclear what the focus of the paper is. Details of measurements do not generally belong in an abstract and distract from the main point of the paper.
• Introduction - this starts very abruptly and the different sections feel very disjointed, without a clear storyline. In general, paper introductions do not need subsections and removing the headers and joining the paragraphs logically might help the reader.
• Methods - there are a lot of details on the measurements at the site, perhaps too many for the main body of the paper, but little to no details on the statistical analysis of the data. From the results section it is apparent that some linear regressions have been performed, although these are not documented in the methods and it is unclear how and why these particular analyses were chosen. It is unclear if the data have been tested for normality or any other assumptions of the methods employed.
• Statistical analyses - the study contains only very basic stats. There is no attempt made to test whether any of the observed differences are significant or not. Most importantly, since ecological processes are complex and many factors contribute to water use, as highlighted by the authors themselves, it is unlikely that a simple linear regression is the best model to use. Some sort of mixed effect or general linear model that can take into account the multiple factors involved might give a better picture of the response of tree water use to CO2. 
• Initial conditions - this is the case with any manipulative experiment, but one always has to wonder about the effect of initial conditions in each treatment on the actual treatment effect. As a minimum, initial conditions should be shown and discussed. Initial conditions could also be included in any more complex statistical model involved.
• Discussion & interpretation - this is almost entirely lacking. The last two sections are in some part a discussion, but they are very superficial, do not discuss the implications of the findings or any limitations of the study.

Minor comments:

L 14  “diurnal (i.e. daylight)” maybe use daytime to avoid confusion? 

L 47 This is a rather abrupt start to the introduction - why large trees?

L 50 unclear what the relaxed xylem is

L 60 wet deposition of?

L 71 photosynthesis is the name of the process even when the pigments are not green

L 71 slightly odd references to model predictions, especially Guerreri et al. there is a lot of literature out there on model predictions of CO2 fertilization effect

L 110 Do you mean drought?

L114 What is meant here by location - geographical coordinates, altitude, slope?

L 141 This sentence is very convoluted and it is unclear what separation refers to

L 156 monthly distributions of what? Significant differences between what and what?

L 182 is 1840 the plantation year?

L 318 It is unclear how these characteristics were used

L 335 I’m not sure this section and level of detail are necessary

L 477 It is unclear why there is a model fitted to each day rather than across days. It would also be good to see R2 values

L 482 ‘smaller’, ‘similar’ Were there any statistical tests done to see if these differences were significant?

Figure 7 It is hard to tell what the difference between a) and b) and c) and d) is and why these could not just be one plot. 

Figures 7 and 8 missing statistical values such as R2 or p

Figure 9 Would be good to show where the statistical differences are

L 549 So what does this figure show us?

L 564 Since you have precipitation data, could you test this deduction?

L 591 this information belongs in the methods

Sections 3.4 and 3.5 are these meant to be a discussion?

This study evaluates the possible effects of enhanced CO2 (eCO2) on tree-water use, the measurements were conducted at one of the FACE experiments on mature forests. Here, the authors defined tree-water use as the daytime accumulated sapflow (TWU) and proposed two hypotheses: H1 A detectable eCO2 treatment effect on TWU is present and H2 TWU is greater in the presence of FACE infrastructure. In my opinion the hypotheses are relevant and could provide valuable information that could improve our understanding on ecosystem responses to environmental change and on eCO2 experimental design.

Unfortunately, the sections of the manuscript feel unconnected due to inclusion of irrelevant information, which could easily be moved to the supplementary material. The hypotheses are not properly tested, so the interpretation of the results and the conclusions are mostly based on visual interpretation of the plots.

The data used in this study is incredibly rare and valuable, so I really hope that the authors can make use of the following comments to better leverage these data.

Major Comments:

Abstract:

There is an unnecessary amount of detail here, it distracts the reader.

It is not clear to me how this info is going to contribute to the development of more “realistic dynamic vegetation models”

Introduction:

The sections inside the introduction feel disconnected and instead of leading the reader to the research questions they feel like tangents.

Not a single mention to the Huber value or water use efficiency, which are key concepts to connect CO2 and water in trees.

Methods:

Almost half of this manuscript is used to describe the methods. It is too much in my opinion. If the authors feel like the figures in this section are really necessary, they could be moved to the appendix section.

There is no mention to any statistical test to evaluate the hypotheses.

I don’t really think TWU is a good metric to detect effects of eCO2 in only 5 years of experiment if the trees are 180 years old, the change of sapwood area of trees under aCO2 and eCO2 would be negligible. However, if measurements of Huber value and Leaf area index are available, TWU can be converted to transpiration, enhancing the differences between treatments if any.

There is no connection on how soil moisture and throughfall are going to be used complement the analysis.

Results and Discussion

There is almost no discussion at all. The results are not put in context, it is hard to tell how the reported values compare to other studies. A relative change between aCO2 and eCO2 would be very useful here.

It is well known that sap flow escalates with sapwood area (proportional to Rb), and Rb to Ac, I don’t really see novelty on reporting linear models on these relationships. How these regressions help with the evaluation of the hypotheses?

All the interpretations to the results rely on visual inspection, I believe this is not enough to evaluate if your hypotheses are true.

It is very interesting to see here interception, throughfall, and soil moisture measurements, however these results appear disconnected from the other results.

The word “significant” is used to described key results of the study, however there is no statistical test associated.

You state that hypothesis 2 is true, however this is hard to tell without proper stats. It is unlikely that the effect of the infrastructure alone is bigger than the effect of the eCO2. If that were true, we could see the combined effect of infrastructure + eCO2 effect in the eCO2 treatment as well.

Minor comments

L 70 “…could be beneficial for individual tree Productivity” hard to tell without considering respiration and temperature increase. These ideas would be helpful to contextualize better.

L 155 how did you test for significant differences?

L 335 Is this relevant?

Figure 9. unreadable

L 570 how can you tell there are” significant differences” without any test? Why do you attribute similar general responses to daylength? What about temperature or VPD?

L 585 “Influence of … and herbivory on water usage dynamics” is misleading, you just point out that an attack of herbivory happened.