Experimental Examination of Vegetative Propagation Methods of Nothofagus antarctica (G. Forst.) Oerst. for Restoration of Fire-Damaged Forest in Torres del Paine National Park, Chile
Round 1
Reviewer 1 Report
Review
Experimental examination of vegetative propagation methods of Nothofagus.
Cafourek J et al.
The authors have conducted a trial of the effects of various treatments/substrates/times of collecting on vegetative propagation of cuttings of Nothofagus. The trial is well designed, carried out, analysed and interpreted. The writing is excellent. I have a few specific comments set out below; one is about the analyses as presented and interpreted. I am sure the authors can add some text explaining more clearly what they have done, and how the two sets of analyses presented relate to each other.
Specific comments
Line 142: substitute ‘number’ for ‘amount’ to read ‘rooted cuttings of the total number of cuttings’
Line 140 – 161: data analysis. This set out well, describing the pathway the investigators took to arrive at what they are showing in the paper. Their design was a 3-way factorial, and they explain how they handled the interactions here. What I didn’t understand is why the analyses shown in Appendices 5 – 8 are presented; do they add to or provide information not apparent in the previous 4 analyses. This needs some explanation here. If the analyses are not providing new information, it would be better to omit them.
Line 164: same as above, replace ‘amount’ with ‘number’
Figures: it would help data presentation and interpretation if the significance of the factors (shown on the X-axis) was indicated above each box plot using the *, ** and *** conventions for P <0.05, <0.01, 0.001 respectively. These P values are shown in the Appendices A1, A2 and A3 respectively, but showing them on the Figures would be simpler for the reader. If this can be done with the graphing program used, it would be very helpful. So for Fig. 1 for example, substrate B gets a ** above the box plot, substrate C gets a *** (P values taken from Appendix 1). This can be repeated for each of Figs. 2 – 9.
Lines 199 – 200: reference to Appendix A5. See comment above about data analysis. Does A5 tell us anything more than what the analysis in Appendix 1 tells us?
Lines 198 – 200: same comment as above
Lines 215: A6 analysis shows a P < 0.05 effect for stimulator, which does not show up in A2 (no stimulator significant). This is confusing; or is A6 the overall test for stimulator, and A2 the result of individual comparisons? This can happen in analyses; an overall test of a factor can be significant, but individual comparisons of levels of the factor fail to show this. So this is possibly what is going on here.
Line 257: ‘As shown in A3 and A7, period and substrate are shown as non-significant ..’. In A3, substrate B appears as significant, but in A7, substrate overall is not significant. Confusing again.
Appendices: the response variable for each analysis is not shown here in the Appendix title (it is given out in the text). It would help readers to have this information given in the title to each Appendix.
Author Response
Thank you for your time reviewing our proposed manuscript “Experimental examination of vegetative propagation methods of Nothofagus antarctica (G. Forst.) Oerst. for restoration of fire-damaged forest in Torres del Paine National Park, Chile” for publication in Forests and giving us a chance for manuscript revision.
We appreciate the quality of the reviews' insightful and constructive comments which helped us to make considerable improvements to the quality of the work as a whole. We have taken into account all comments and have responded as clearly as possible in a point-by-point response format bellow.
Reviewer 1
The authors have conducted a trial of the effects of various treatments/substrates/times of collecting on vegetative propagation of cuttings of Nothofagus. The trial is well designed, carried out, analysed and interpreted. The writing is excellent. I have a few specific comments set out below; one is about the analyses as presented and interpreted. I am sure the authors can add some text explaining more clearly what they have done, and how the two sets of analyses presented relate to each other.
Response:
Specific comments
Line 142: substitute ‘number’ for ‘amount’ to read ‘rooted cuttings of the total number of cuttings’
Response: amount substituted for number as suggested
Line 140 – 161: data analysis. This set out well, describing the pathway the investigators took to arrive at what they are showing in the paper. Their design was a 3-way factorial, and they explain how they handled the interactions here. What I didn’t understand is why the analyses shown in Appendices 5 – 8 are presented; do they add to or provide information not apparent in the previous 4 analyses. This needs some explanation here. If the analyses are not providing new information, it would be better to omit them.
Response: Appendices 1 to 4 present results of model fits, which include Wald test of significance of the differences between reference levels of each factor and the remaining levels of the same factor. These Wald tests were not primarily used to assess the significance of any factor as a whole.
For that purpose, LR-test were used, and these are presented in Appendices 5 to 8. The use or LR-test might look unnecessary at the first glance, but the contrary is true. It can be demonstrated by comparing the results of Wald tests in Appendix 3 (Substrate B differs from A at the alpha level equal to 0.05) and Appendix 7 (none of the factors was detected as significant, including the Substrate). The reason is that the performance of a number of Wald tests and rejection of H0 (saying the factor is generally not significant) in case that any of these single tests rejects the H0 (particular factor level differs from the chosen reference factor level) increases the probability of type I error (one rejects the H0, when it is actually true) and the P value of the particular Wald test (comparing the particular level to the reference) does not correspond to the P value of the overall significance test. If Wald test were to be used for the general factor significance testing a compensation for this P value increase should be applied at the same time (e.g. Bonferroni). However, we like many other analysts or statisticians, consider the LR-test as superior approach for overall factor significance assessment.
We do apologize for this obvious misunderstanding because we have just discovered, that the manuscript (after English revision) changed in a way complicating or even preventing the correct understanding of why and how we used the two tests.
Line 164: same as above, replace ‘amount’ with ‘number’
Response: amount substituted for number as suggested
Figures: it would help data presentation and interpretation if the significance of the factors (shown on the X-axis) was indicated above each box plot using the *, ** and *** conventions for P <0.05, <0.01, 0.001 respectively. These P values are shown in the Appendices A1, A2 and A3 respectively, but showing them on the Figures would be simpler for the reader. If this can be done with the graphing program used, it would be very helpful. So, for Fig. 1 for example, substrate B gets a ** above the box plot, substrate C gets a *** (P values taken from Appendix 1). This can be repeated for each of Figs. 2 – 9.
Response: We did not use any indication of significance here, because the box plots do not serve for the significance testing and we do not want to confuse the readers by still another type of ‘significance’. The presented box plots correspond to just one-dimensional slices of data (only one out of the possible three factors is addressed in each of the boxplots). However, the two types of significance tests (see above) depend on the model structure in terms of the factors included.
Following your suggestion, we would also need to indicate the reference level of each factor to which these ‘significance tests’ would refer. This means that the reference level would never get any asterisk.
Overall, we prefer not to follow your suggestion.
Lines 199 – 200: reference to Appendix A5. See comment above about data analysis. Does A5 tell us anything more than what the analysis in Appendix 1 tells us?
Response: Kindly see or above explanation concerning the different purpose of Wald and LR-test used in the respective groups of Appendices.
Lines 198 – 200: same comment as above
Response: Kindly see or above explanation concerning the different purpose of Wald and LR-test used in the respective groups of Appendices.
Line 215: A6 analysis shows a P < 0.05 effect for stimulator, which does not show up in A2 (no stimulator significant). This is confusing; or is A6 the overall test for stimulator, and A2 the result of individual comparisons?
Response: Both, your observations and also conclusions are absolutely correct. The LR-test for the overall factor significance is known to have larger power i.e. it is capable to detect actually significant factors in more cases that the Wald test would eventually do (after their P values were corrected for multiple comparisons).
This can happen in analyses; an overall test of a factor can be significant, but individual comparisons of levels of the factor fail to show this. So this is possibly what is going on here.¨
Response: Correct.
Line 257: ‘As shown in A3 and A7, period and substrate are shown as non-significant ..’. In A3, substrate B appears as significant, but in A7, substrate overall is not significant. Confusing again.
Response: We hope this is now also clear, after the above explanations, and revisions of the text.
Appendices: the response variable for each analysis is not shown here in the Appendix title (it is given out in the text). It would help readers to have this information given in the title to each Appendix.
Response: Yes, this has been fixed in the revised version of the manuscript.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
All comments are shown in the text of the paper. In order to repeat the same or similar research, you need to describe in more detail the materials and methods of work (nursery location, description of containers, types and characteristics of peat and perlite used in the research, their physical, chemical and biological properties, add exact dates of observation cuttings, ... ). Explain why you used a mixture of only two substrates and not more (maybe three). Explain why you also considered callused cuttings vital because they are sometimes not an indicator of survival or rooting. Missing citation 51 in the article. Consider whether you need additions from A1 to A8 in the article because they only burden the work and increase the number of pages.
Comments for author File: Comments.pdf
Author Response
Thank you for your time reviewing our proposed manuscript “Experimental examination of vegetative propagation methods of Nothofagus antarctica (G. Forst.) Oerst. for restoration of fire-damaged forest in Torres del Paine National Park, Chile” for publication in Forests and giving us a chance for manuscript revision.
We appreciate the quality of the reviews' insightful and constructive comments which helped us to make considerable improvements to the quality of the work as a whole. We have taken into account all comments and have responded as clearly as possible in a point-by-point response format bellow.
All comments are shown in the text of the paper. In order to repeat the same or similar research, you need to describe in more detail the materials and methods of work (nursery location, description of containers, types and characteristics of peat and perlite used in the research, their physical, chemical and biological properties, add exact dates of observation cuttings, ... ). Explain why you used a mixture of only two substrates and not more (maybe three). Explain why you also considered callused cuttings vital because they are sometimes not an indicator of survival or rooting. Missing citation 51 in the article. Consider whether you need additions from A1 to A8 in the article because they only burden the work and increase the number of pages.
Response: Many thanks for your valuable comments, see separate responses bellow.
Line 92: coordinates, altitude
Response: added
Line 94: coordinates, altitude
Response: added
Line 110: volume one hole of containers?
Response: added
Line 112: Add the characteristics of perlite and peat there are many on the market. This was important so that the research could be repeated.
Response: We used common commercial peat and perlite to get as close as possible to the operating conditions in forest nurseries.
Line 114: Why such a combination of two substrates (perlite and peat mixture)? Describe?
Response: Based on our long-term experiences these two substrates work best.
Line 132: Why you transplanted callused cuttings and classify as live? Describe?
Response: lightly coloured callus section is sign of future successful rooting, therefore it was classified as “survivors”. Added into text.
Line 133: number of holes, volume one hole of containers?
Response: 50 holes with volumes of 100 ml, added into text.
Line 136: delete, duplicate
Response: Deleted.
Line 137: date?
Response: Exact date was added.
Line 138: Causes of mortality?
Response: It was already mentioned in Discussion (lines 308-314)
Line 222: No open circles in figure 4.
Response: deleted from Figure description.
Line 291: delete, duplicate
Response: deleted
Line 311: Peat, especially white, does not contain pathogenic fungi because it is sterile. Mushrooms later come to the nursery in various ways, explain in your research?
Response: We used common commercial unsterilized black peat. Most of the dead cuttings were attacked by rot.
Line 318: 21
Response: missing number of reference was added.
Line 336: And plant phytohormone too.
Response: Plant phytohormones were added.
Lines 360-364: Merge reference 54 in ono sentences.
Response: Merged.
Lines 473, 477, 542, 544, 578: Page numbers?
Response: missing page numbers were added.
Lines 570-572: Missing reference 51 in article?
Response: In Table one, there was this reference incorrectly marked as number 21, corrected
Author Response File: Author Response.pdf
