Referees noted a few technical points especially with respect to soil moisture limitation thresholds that - in addition to numerous suggestions to add important detail to the measurements - appear to be comprehensively considered in the response letter. I thank the authors and referees for attention to detail and look forward to reviewing the revised manuscript.

Sincerely,
Paul C. Stoy

The manuscript was reasonably well-designed and written if not perhaps a bit standard with limited new insights into the response of sapflux to environmental variables in dryland ecosystems. The discussion was a weak point; mechanistic rigor could be added by noting important lags in the soil-vegetation-atmosphere system that must be taken into account to understand hysteresis behavior (e.g. Bohrer et al. http://onlinelibrary.wiley.com/doi/10.1029/2005WR004181/full or Matheny et al., http://onlinelibrary.wiley.com/doi/10.1002/2014JG002623/full). The discussion of hysteresis loops is interesting but must be framed more mechanistically to have full impact. I also note that the discussion of the decoupling coefficient was somewhat lacking after its introduction in the Methods section nor is it clear how u* was calculated for atmospheric conductance calculations. Please address these and the following minor comments and send a revised manuscript for consideration of publication.

Sincerely,
Paul C. Stoy

Minor comments:
(All line numbers correspond to the Author's Response manuscript.)

Quantify 'sizeable' on line 455.

'and thus less sensitive' on line 461.

I'm not sure what the resultant for line 489 is but I assume that it will be improved by BG editorial services. (The same holds for line 509.)

502: usually around midday

In equation 2 how was friction velocity measured? (and wind speed for that matter, which is described in the following subsection 2.3 that would be better off before the description of the decoupling coefficient.)

On line 594, why these definitions of the seasons?

on line 598, these REW/VWC relationships should approximately hold for different soil types: do the references relate to the soil types found at the site? On line 604 I don't understand how this would vary from year to year.

Should equation 5 be corrected for the clumpiness of shrubs on the landscape? The leaf area index applies for the area near the shrub, but I'm assuming that the space between the shrubs has marginal vegetation which should be accounted for when presenting results on a per-area basis.

'Range of daily' should be 'The range of daily' in a few instances in the Results section.

683: peaked

687 wasn't an effective argument: why would you want to remove the influence of Js from the data? Perhaps its diurnal behavior to help understand lag relationships.

On line 766 the counterclockwise hysteresis loop is consistent with capacitance in the soil-plant-atmosphere system: it takes time for water to move up and expand vascular elements during the transition from night to day.

Please justify the statement 'In semi-arid regions, low VWC restricts plant transpiration more than VPD'. I'd assume that they are both important but perhaps at different time scales with VPD playing an important role during the mid-day depression even under high VWC, but low VWC is correlated to low VPD indicating that both water supply via VWC and demand via VPD are important in this case.

I note that the decoupling coefficient was not discussed following its description in the Methods section.

In Figure 6 please avoid using red and green simultaneously if possible (or at least choose different symbols so our red-green colorblind colleagues can distinguish them).

The revised manuscript by Zha et al. represents a modest improvement, but still requires substantial work. The methods and results sections are largely sound with some important avenues for further improvement noted below. I still struggle with the Introduction and Discussion (which is barely longer than three pages). Both require improvement for the manuscript to be publishable in Biogeosciences. Statements like “In general, VWC has an influence on physiological processes of plants in water-limited ecosystems (Lei et al., 2010; She et al., 2013)” do not lend confidence that this manuscript synthesizes existing knowledge effectively. The problem is that fundamental references regarding the response of leaf and canopy conductance to water availability and micrometeorological drivers is largely missing. As a consequence, the manuscript has a weak foundation and results are presented as surprising because existing knowledge has not been synthesized.

At a minimum, please read (and cite if you choose):
Jarvis and McNaughton 1986:
(http://www.sciencedirect.com/science/article/pii/S0065250408601191)
Jarvis 1976: (http://rstb.royalsocietypublishing.org/content/royptb/273/927/593.full.pdf)
Oren et al. 1999:
(http://onlinelibrary.wiley.com/store/10.1046/j.1365-3040.1999.00513.x/asset/j.1365-3040.1999.00513.x.pdf?v=1&t=j2glnk7z&s=3c262f2ce665cf0835b5c17c02962ae104087e06)
Koerner 1995: (https://link.springer.com/chapter/10.1007/978-3-642-79354-7_22#page-1)

I note that Jarvis and McNaughton is cited with respect to the decoupling coefficient on page 8, but this information was never synthesized in a meaningful way the Discussion.

A number of specific passages require rethinking. A selection:
“Changes in stomatal conductance and, thus, transpiration may equally affect plant water use efficiency (Pacala et al., 2001; Vilagrosa et al., 2003)” is an odd statement given that transpiration is part of the equation for water use efficiency. I recommend re-wording.

On line 98 it need not be only mid-day.

The sentence on line 105 should be cut. It seems like the authors are surprised that we know quite a bit about the controls over transpiration. The fundamental literature is largely not cited as noted. Including it seems in Artemisia species. One example: “Soil water content, in combination with other environmental factors, may have a significant influence on sap-flow rate” Of course it does when it is limiting! It’s like the introduction was written with not cognizance of fundamental plant ecohydrology.

On line 187, these are the normal climatological season definitions.

“Our finding regarding lower sensitivity in Js to environmental factors (Rs, T and VPD) during dry periods was consistent with an earlier study of boreal grasslands (Zha et al., 2010)” has little meaning. How, and why was this paper written by the authors selected for a vague comparison?

The passage on line 247 could mean any number of things (“Soil water was shown to modify the response of Js to environmental factors (Fig. 4).” Given the specific subsequent sentences I recommend removing it.

On line 278, “The effect of…”

“In general, VWC has an influence on physiological processes of plants in water-limited 302 ecosystems (Lei et al., 2010; She et al., 2013).” is not informative.

On line 304, the literature is almost entirely cherry-picked to reflect papers of the authors rather than more relevant papers in the sapflux literature. Why are boreal grasslands and (on line 299) Scots pine chosen for comparison?

Line 334 reiterates my point about foundational plant physiology, “According to O'Brien et al. (2004), diurnal variation in Rs could cause change in the diurnal variation in the consumption of water.” Yes, of course. Plants respond to photosynthetically active radiation, which comprises the major component of incident shortwave radiation.

The paragraph beginning on line 338 could still be written in a way that reflects that we know how stomata respond to environmental stimuli for decades or longer. Researchers have just discovered that stomatal responses to VPD are controlled by de novo synthesis of abscisic acid (McAdam et al., 2015, http://onlinelibrary.wiley.com/doi/10.1111/pce.12633/full).

No meaningful discussion of the differences between 2013 and 2014 are presented in the Discussion.

How certain are the values presented in Table 2 (i.e. what is the representative uncertainty)?

The manuscript represents an improvement but requires further minor edits. Comments are based on the draft with tracked edits and focus on the introduction and discussion sections.

53: Js was just defined in the previous paragraph. (see also line 61 "sap-flow" with hyphen and other abbreviation inconsistency likewise elsewhere; please be consistent with abbreviations.)

55: 'ahead of' is colloquial.

'escaped' on line 62 is too strong a word. Every plant dreams of escaping drought limitation but arid species can't. (I do note that it is used in Briolette et al., but for reproduction timing).

Line 68 can be cut because no models are involved.

Line 91 needs re-wording

Lines 100-105 should be moved perhaps to the methods, they are a distraction here.

The paragraph on line 122 is actually informative. This is effective background that educates the reader and motivates the study. (line 127 can be moved to the next paragraph to "funnel" the flow of arguments toward the importance of studying A. odorosica (i.e. undo the change on line 151). Note that it may help to write A. odorosica henceforth for brevity.

what is 'steep' on line 345? ('steep' is qualitative).

sentences like that on 357 provide nice comparisons. But after that, is Picea crassifolia a dryland species? (note bolding/lack of italics in "Caragana korshinskii".

The gray banding in Fig. 2 is inconsistent if it refers to VWC or REW values below the indicated thresholds (unless there is a time-integrated aspect that is best to describe in the legend).

Sincerely,
Paul C. Stoy

The re-revised manuscript makes the requested improvements and is acceptable for publication in Biogeosciences.

Sincerely,
Paul C. Sto