|This manuscript is somewhat improved over the version that I read previously. The paper makes it clear that the authors carried out a significant amount of work, both in the field, and in data processing and analysis.|
However, I do have some concerns about the work done, and about the way in which it has been written up.
In terms of the work done, I have concerns about the stemflow collecting system used by the authors. Despite their assurances, the photographs supplied with the paper show quite wide gaps between the stem and the stemflow collar, that may have allowed rain or released throughfall from branches above to enter the stemflow collars and be erroneously counted as stemflow. I would like to have seen the authors make some estimate of how large this error could be.
The authors still say nothing about wind speed or about the extent to which the rain had an oblique approach angle. They seem overly concerned about leaf architecture and not sufficiently concerned about the possible effects of oblique, wind-driven rain on the field measurements. Likewise, what was the effect of wind in dislodging drops that might otherwise have become stemflow? Considerations of this kind cause me to wish that the authors could be somewhat more cautious in their conclusions, and at least acknowledge the potential effects of variables that they do not consider in their analysis.
The authors have been more circumspect following my previous review, and have backed away from claiming that stemflow is critical to drought survival, now stating that it ‘might be’ important. However, they are still in my view insufficiently careful with their argument. For instance, on page 3, in their Introduction, the authors claim that ‘Stemflow delivers precipitation directly into the root zone…’ (line 42). But of course this is often not the case, and instead the stemflow arrives at a litter layer beneath the plant where, in addition, the soil is sometimes hydrophobic. The authors need to be more careful (especially as they have no evidence of soil moisture changes caused by stemflow) in making claims of this kind. Some fraction of the stemflow likely reaches the root zone, but just how much does so is an important question that should not be overlooked. The authors also become rather enthusiastic in line 531, where they argue that efficient stemflow collection might be of ‘great’ importance – or perhaps this should be just of ‘some’ importance, until we have some actual evidence.
Given the very large literature on stemflow, I also think that the authors are too sweeping on page 3 where they claim (line 64) that ‘..previous studies have usually ignored stemflow…’. This would be news to the authors of hundreds of papers on stemflow.
Written English is still poor in places. Especially in section 2.1 (study area), the authors use past tense inappropriately (e.g. line 145 should say that the shrubs ‘are’ multi-stemmed, not ‘were’; this error occurs repeatedly in the whole top half of page 7). In general, ‘grew’ should be ‘grow’, ‘was’ should be ‘are’, and so on. The authors still use the SI metric system carelessly. For instance, ’20-cm-diameter’ (line 167) is incorrect, and should be ’20 cm diameter’ ; similar errors occur in line 204, line 417, and elsewhere.
The authors need to proof-read their work carefully. For instance, they often refer to ‘defoliated and manually defoliated shrubs’ (e.g. line 112, and again in lines 117-118), when they mean ‘foliated and manually defoliated’. This becomes quite confusing. The authors sometimes refer to the shrubs competing for ‘lights’ (e.g. line 181, line 520) but this should simply be ‘light’.
My overall feeling about the paper is that the work is generally of good standard, but that the authors should again check their paper carefully for errors; give thought and comment to the quality of the field data (e.g. was throughfall counted as stemflow?). At the same time, if possible, the manuscript should be shortened as the Discussion in particular is quite long and a little repetitive.