Effects of Bark Stripping on Timber Production and Structure of Norway Spruce Forests in Relation to Climatic Factors
, Zdeněk Vacek 1, Rostislav Linda 1, Stanislav Vacek 1, Petr Marada 2,3, Václav Šimůnek 1 and František Havránek 2Round 1
Reviewer 1 Report
The manuscript “Effects of bark stripping on timber production and structure of Norway spruce forests in relation to climatic factors” addresses the impacts of bark stripping on the production, structure, and health of young Norway spruce forest stands. It combines long-term field data from four sites, 16 plots, in the Czech Republic, under different levels of exposure to bark stripping by sika deer with several statistical and modelling approaches.
Two climatic factors were assessed mainly in relation to the radial growth (diameter increment). It is a bit misleading and therefore, I do not think it is enough to mention in the title. Likewise, the authors mention some economic aspects both in the introduction and discussion, but they were not analysed in this paper.
Regarding the statistical and modelling approaches, the authors merely offered a reference and did not describe nor explain why did they choose these methods. For example, what does DendroClim do exactly? Please describe the required parameters a little and the main processes involved. Also, please mention your sample size for each of these methods.
Additionally, the authors mention observations in the Results section that are not further elaborated in the Discussions section. For example: “Crown closure was in the range of 77.3–93.9 and showed higher variability on HD plots”. Why is that, how come?
Therefore, my main concerns with the manuscript relate to the appropriateness of the modelling and the significance and quality of the presentation. Please improve your text, word choices, and eliminate grammar errors.
Line 33: Missing word: “This increase in is not”.
Lines 42-43: Missing word: “Browsed seedlings generally show significant deceleration in growth [26,27], extreme causes of repeated browsing of young trees could be fatal [28].
Just for clarity purposes, please reference the climatic data you used between 1961 and 2017 in Chapter 2.1 as well.
Lines 135-141 are exactly the same as lines 142-148.
Line 153: I am assuming that “using a different year smoothing spline” means that you applied the smoothing spline for a different year and used it for the current year. Please clarify your methods a bit better?
Line 159: “For the modelling of diameter increment in relation to climate characteristics (monthly precipitation and temperature) the DendroClim software was used [64]”. Eight lines later you write: “In order to determine the response of climatic factors, radial growth (diameter increment) values were tested for correlation with climatic data”. If these are separate things, please explain the difference, otherwise, these sentences are only the reverse of each other.
Line 169: “Analyses were processed in Statistica 12 (StatSoft, Tulsa) and R software [64]. Three-dimensional plot was made using Gnuplot 5.2.”. Please explain why were precisely these software and programs used and to what purpose? Also, please mention your sample size.
Line 179: A stand index from 0.49 to 0.97 on HD, compared to a stand index from 0.81 to 0.99 on LD, is not quite a “small difference” as you call it, especially since you continue by writing that “considerable differences were found in the mean DHB of LD (20.03 cm (±3.42 SD)), compared to HD (17.49 cm (±1.37 SD))”.
Table 2: Does N refer to the tree planting density from 40+ years ago, the tree density from 2018, or the tree density including mortality?
Figure 4: The overall averages for the ring width index seem to go in the same spot. One might think that the presence of deer could be overlooked. Also, there are years where the ring width index for damaged trees is noticeably higher than that of healthy trees. Why is that?
Lines 280-283: Please move this part to the Methods and Materials section and briefly refer to it here.
Figure 6: This is a flashy figure and while it is appropriate, the main difference stays within a closer range for the chosen parameters.
Lines 320-329: This part would be a great introductory phrase to your objective.
Lines 371-382: This part could also be in the introduction, but shortened.
Line 398: Except for this line, you don’t really explain your results. Please also explain your results in terms of the studied area and not only in the context of other studies.
Author Response
Dear Reviewer 1,
thank you for your beneficial comments. We have tried to answer all your suggestions. The whole manuscript was also checked by native English speaker. Detailed answers are written below behind your questions.
Sincerely,
Jan Cukor, corresponding author
Author Response File: 
Author Response.docx
Reviewer 2 Report
The submitted manuscript analyzes the impact of sika deer bark stripping on Norway spruce forest production and stand structure in the context of climate in the Czech Republic. With growing global ungulate populations, research on ungulate damage impacts to forest growth and yield are critical to future forest management. Given the population growth of the naturalized sika deer in the Czech Republic and central/western Europe, quantification of their present impacts provides a means for projecting future impacts on forests not only in their present range but also in new areas to which they may spread.
General Comments
The authors have clearly spent a lot of time thinking about potential impacts of sika deer damage to forest stand structure, as testified by the numerous production, structural, and diversity metrics that are presented in the manuscript. Also, I thought the inclusion of climate factors in the analysis of deer damage effects on radial growth to be an appropriate and an interesting interaction, given climate's overwhelming importance in dendrochronology and the impacts of herbivory on developing trees.
The manuscript had several issues with English grammar and syntax that, in some sentences, made it difficult to infer the authors' meaning. I would encourage the authors to seek assistance with English stylistics while conducting revisions. Further, the Methods section lacks some required information in regards to field and analytical methods. See specific comments for other recommended revisions/considerations.
Predominantly, I am concerned with the choice of statistical methods. It appears as though each of the 16 PRPs are treated as an individual sample, as forest stand metrics and diversity metrics are calculated for each PRP and the number of PRPs with "significant" differences (as demonstrated by Monte Carlo simulations) are counted, and the area density type (i.e. HD vs. LD) with the greater number of "significant" plots is considered the more important. I have never personally used this approach nor seen it used in a research article. Also, the authors make statistical comparisons in the results based solely on means and standard deviations. A more appropriate method of comparing differences in stand and diversity metrics would have been to compare the means (or medians) of each metric between the LD and HD areas, acknowledging the hierarchy of the data (i.e. 4 PRPs in each of the 4 areas) with the inclusion of random effects in formal models. This would provide a robust estimation of deer density effects on the forest structure. Also, PRPs were only 100 m2 (10 m x 10 m) and, from the information provided in the manuscript on the counts of trees of each damage type and the stem densities, it appears as though each PRP only contained 10-20 trees. Although Norway spruce plantations have rather low variation, this is still a rather low number of trees upon which to perform this type of analysis. In the future, I would encourage the authors to consider using larger plots or collecting a greater number of plots in studies of this type.
Specific Comments
Lines 13-14: I would add the Latin names immediately after the common names of sika deer and Norway spruce. If this is done, the Cervus nippon should be removed from the Keywords.
Lines 36-37: Instead of saying that sika deer were "...introduced from Japan to Europe ca. 150 years ago", inserting the range of years (e.g. 1860-1870) instead of the years before present would provide more information.
Lines 48-49: "The intensity of bark stripping is not only determined by population density of deer, but as well by forest stand characteristics..." Provide the reader one or two examples of forest stand characteristics that bark stripping is known to impact.
Line 62: Scotland is part of the United Kingdom. Both of your references are for localities in Scotland, so "United Kingdom" should be removed from the sentence, or references to studies (if any) in the remainder of the United Kingdom should be added.
Figure 1: In the inset map depicting Horni Kozolupy and Rochlov hunting districts, "PRP 8-12" in Rochlov hunting district is incorrectly labeled and, I believe, should be "PRP 9-12."
Line 96: The abbreviation "resp." is unclear.
Lines 105-109: This information on the deer harvested in each of the hunting districts provides an index of population size and would be better placed somewhere in lines 80-84 where the "game influence" of each district is initially introduced.
Lines 111-121: The authors never explicitly mention in what year(s) their field data were collected. Increment cores were collected in 2017 (line 122): was this also the year when field data sampling occurred? The authors need to clearly state when field sampling occurred.
Lines 126-129: How did the authors determine the years in which bark stripping damage and stem rot occurred? Was this estimated through dendrochronology or did the authors collect field data on bark damage and stem rot annually? The authors need to clarify this point.
Lines 141-148: These sentences duplicate information in lines 135-141 and should be removed.
Line 165: The exact type of multiple comparisons (e.g. Dunn, Bonferonni, etc.) used after the Kruskal-Wallis test should be denoted.
Lines 177-189: The authors provide several summary statistics for low deer density and high deer density stands and describe "statistical differences." However, the authors do not formally test for statistical differences in metrics between areas. The authors should consider performing statistical tests to distinguish the HD/LD areas, as these would provide an indication if stand metrics truly vary with deer density.
Table 2 and Table 3: Instead of making the metric abbreviations the column titles, I would use the actual metric names to avoid confusion (e.g. use "Basal Area" as the column title instead of "G").
Lines 220-221: The PRPs and density descriptions are mismatched. In the rest of the paper, LD plots are PRPs 1-8, and HD plots are PRPs 9-16. However, they are switched here. Also, the inclusion of both PRP numbers and LD/HD in each mention are redundant: choosing one or the other and using it throughout the paper would be a better alternative.
Lines 280-283: This information would fit better in the Methods discussion of the model (lines 165-167). Also, I believe given the results that the link function used was a "negative inverse" link instead of only an "inverse" link, as the authors have described. The model should also have indicated that the hierarchy of the data in that 4 PRPs are nested in an area and 8 PRPs each are nested in a hunting district (or at least the authors should have mentioned in the Methods that they considered this). The inclusion of area or hunting district as a random effect in their generalized linear model would have been an appropriate step.
Table 4: This table should be formatted similarly to Tables 2 and 3.
Lines 330-333: The long list of authors' names is somewhat tedious. This sentence should be restructured to use only the reference numbers.
Lines 350-352: The authors should augment this information by describing, in a paragraph, known impacts of sika deer on forests in different localities.
Author Response
Dear Reviewer 2,
thank you for your beneficial comments. We have tried to answer all your suggestions. The whole manuscript was also checked by native English speaker. Detailed answers are written below behind your questions.
Sincerely,
Jan Cukor, corresponding author
Author Response File: Author Response.docx
Reviewer 3 Report
Summary
In this manuscript, the authors evaluate the effects of bark stripping by sika deer on Norway spruce plantations, and relate these effects to climatic variables. They underline differences in tree growth and vertical heterogeneity of the stands according to stripping pressure, as well as increased susceptibility to extreme climatic conditions for damaged trees.
General comments
This research seems like a relevant addition to the current knowledge on the subject, for instance by showing long term results in an area with very high ungulate pressure.
It adds dendrochronological data to the bark stripping corpus, which is interesting, but would gain from better presenting these results.
Major English revision is necessary.
The variance is seldom shown throughout the results. This is crucial to properly assess the soundness of the results, and should be shown systematically where relevant.
The methods seem sound, but need more detail in some sections:
Clearly state in the methods that the stands are plantations
Clearly show when the bark stripping took place. Was the damage done constantly over the history of the stand? When was the fence removed? Why is the 10-23 years shown in the abstract but not in the main text?
Explain clearly how the number of hunted deer relates to their density. There could have been changes in hunting licences or multiple other factors influencing the relationship between these two variables.
Each of the two areas (LD and HD) are split into two localities. While this is normal when studying natural browser populations, it does call for mixed models in order to account for the correlation that could ensue (site effect). There seems to be a difference between the sites belonging to the four localities when looking at figure 7B, which further legitimates the use of mixed models.
The indices used are not detailed in the manuscript, and some come from reports or articles in other languages than English: there is no way for me to evaluate the soundness of these indices.
Regressions are shown, but their differences are not compared. Are they significantly different?
The response variables vary between DBH, height, and volume. If the reason behind these changes are not clearly explained, it looks like fishing for statistical significance.
The analysis of relations between growth, stripping damage and climatic variables is interesting and a nice addition to the manuscript. There is potential to further link it to climate change, and this would gain from being discussed in more length rather than only mentionned in the conclusion.
Specific comments:
Line | Comment |
20 | 10-23 years old shown in abstract, but not in text. Add to results to show when the stripping took place. Need for more clarity regarding stripping time/duration. While it could have happened consistently over time, the dendrochronological methods suggest that “tree-ring widths with the occurrence of bark stripping damage (year) and stem rot were measured…” (line 126), which suggests more precision for timing of stripping. |
21 | Define “standard diameter increment” |
27 | Add “Norway spruce, Picea abies |
37 | 6 references needed here? Select most relevant. Applicable elsewhere also. |
45 | Reference 30 irrelevant here. |
47-48 | Show other reasons, such as reduction in available resources in winter |
48 | Reference 31 irrelevant here. |
62 | No reference for Czech Republic? Seems necessary to show relevance in local context. |
65 | Show clearly whether the increase in hunted deer is linked to increase in population or only change in hunting restrictions. Confounding factors could include changes in hunting regulations/quotas. |
72-73 | I fail to see the real difference between objectives 1 and 3. Add interactions in objective 1 to avoid repetition. |
79 | State that the stands are plantations |
86 | How long was the fence there? When exactly was it removed? |
Fig1 | The map is too busy. The inset with the whole of Europe could be removed. |
113 | Clearly define the variables: · How was the position measured? · How was the crown projection measured? |
114 | State DBH height at first mention (here) rather than at line 118 |
117 | Reference in Czech, I can’t evaluate the method. |
119 | How were the categories defined? Were there clear patterns suggesting these 3 categories? Is it based on literature? |
122-126 | There are 60 cores and 16 sites, so the number of cores per site is inconsistent. How were the cored trees selected and distributed within and amongst sites? The random selection mentioned suggests that there could be no stratification between sites. |
135-140 | The indices need to be defined in text. I don’t understand the language for most of the references, so I can’t evaluate the relevance of the indices. |
141-148 | First part of paragraph repeated here. |
165 | Explain why Kruskal-Wallis. Non-normality? |
Table 2 | There should be a measure of variance for DBH, h, v, HDR, CPA. |
Table 2 | The methods do not show how crown closure was measured/calculated |
196 | Why was the “small damage” not included in the analysis? |
196 | Show the detailed Näslund function |
200-201 | a and b have not been defined. |
Fig2 | Are the two statistically different? Once the “small damage” is added, are the three statistically different? This is needed in the analysis. |
205 | Explain why the response variables shifts between height, crown diameter and DBH. There needs to be an explanation as to the relevance of these particular response variables. |
Fig3 | The DBH values for largely damaged trees seems to stop at 25 cm. Why is the line going beyond? It suggests extrapolation beyond the data. Is the R2 value of 0.12 high enough to compare the statistical difference between the two groups? |
215-223 | I can’t understand these results as the indices have not been detailed. It seems odd that the vertical structure is moderately to strongly varied on HD plots, but that height differentiation is not. |
232-233 | Use only one of the two terms: regional standard chronology or local standard chronology. |
234 | What is the level of significance? How was the significance measured? |
236 | Did the populations subside or not? The introduction suggests considerable increases in population over the last decades, and this paragraph suggests that the “impact of game attack” subsided. If the populations did increase, then what brought the reduction in game attack? |
Fig4 | This figure is confusing. Too many lines, hard to follow. The Ring Width Index needs to be defined. |
Fig4 | Show when the stripping occurred on the figure (if possible, again there is no information in the text as to when it occurred). |
250 | Show the significance levels, not just the correlation. |
270 | Refer to comment at line 205 |
280 | This model does not look at the effects of stem rot, but of “% circumference damage”. If you want to model rot, use rot as your explanatory variable. 62.1% of the damaged trees were showing rot (line 262), so the correlation is not perfect. |
281 | Discussing the impact of rot on stem volume recovery for wood production would be interesting. |
290 | Where does the 1/3 come from? The inflexion point needs to be clearly defined, and it is not obvious on figure 6. |
Fig6 | The 2 coloured scales are repetitive on this figure. The whole figure could actually be avoided. |
319 | It is hard to follow the trail of thought in the discussion. Importantly, the discussion should focus on how the results from this study relate to other findings in the literature. It generally fails to do this. |
320-336 | Belongs to the introduction and repeats some of the material from there. |
320-325 | Too many references, some too old to show what is stated in text (especially lines 324 and 325). |
337-345 | Repeating results. This is not for the discussion |
341-342 | Need reference for this statement. |
349 | This is the first mention that the stripping damage was “long-term”. This needs to be clear in all the previous sections. |
349-352 | Belongs in introduction |
353-357 | Need more references than #86-88. This concept goes well beyond the local context and there are multiple global studies looking at it. |
358-366 | Careful when comparing to literature on other species. It can be relevant, but needs to be stated clearly. |
360 | Say clearly that ref #23 looks at red deer |
361-363 | How do you know that the current DBH in this study correspond to the preferred DBH for bark stripping? You stated previously that the most stripping diminished before the end of the study. |
364 | Ref #90: this is not the original reference. |
367 | Say clearly that ref #35 refers to red deer. |
375 | Add reference for the “50 cm2” |
383-404 | The section on climate needs a clearer storyline. For example, it could gain from contextualising it with climate change (clearer in the discussion section with references instead of a brief mention in the conclusion). If your results show higher vulnerability to drought for the stripped trees, for instance, then climate change bringing more drought (although I don’t know what’s predicted for the Czech Republic) could further increase the consequences of stripping. |
413 | “early stages of growth”: there was a fence in the first years, so how do you define early stages of growth? |
426-437 | Missing sections |
Author Response
Dear Reviewer 3,
Thank you for your beneficial comments. We have tried to answer all your suggestions. The whole manuscript was also checked by native English speaker. Detailed answers are written below behind your questions. All changes in the original version were yellowed for highlighting the changes.
Sincerely,
Jan Cukor, corresponding author
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
Firstly, I thank the authors for their thorough answer and I agree with their input.
The manuscript “Effects of bark stripping on timber production and structure of Norway spruce forests in relation to climatic factors” has improved significantly. The methods, tools, and software are now described with sufficient details and the analysis is well presented. The results are now well elaborated and explained in the context of existing research. The text is now clear and concise.
Apart from these general remarks, I only have the following minor remarks:
Line 57: “Severe infection of multiple trees could cause the disintegration of the entire stand could due to…”.
Line 63: “…with a high population density…”
Line 66: “whereas the European populations…”
Line 166: “…for a difference time windows window…”.
Line 198: “…the amount of trees per hectare were was comparable…”. The quantifier “the amount” does not fit with the countable “trees”, maybe consider “the number”.
Line 221: “…showed a low…”
Line 254: “…shows their the high rate…among them.”
Line 289: “…trees between 10-23 years old…"
Line 316: “a rapid decrease…respect to the presented/investigated surface plots, were was illustrated…”
Line 320: Instead of “Similar declinations were”, maybe “A similar decline was”.
Line 358: “…volume of only…”
Line 359: Please also write the last standard deviation in parenthesis and stay consistent throughout the paper.
Line 377: “…the horizontal structure of forest stands were was…”
Line 381: “…red deer [21], however,/ although Welch et al. [32] also showed…”
Line 383: “…study: the actual mean…”
Line 404: “…factors, the radial growth…”
Line 437: “…for the last 20 years…”, “…with the an average…”
Line 439: “…has a significant negative…”
Other than that, there are only a few commas that are either missing or incorrectly placed, but these don’t change the overall quality of the paper.
Author Response
Dear Reviewer 1,
thank you for all your comments. The suggestions were improved throughout the manuscript. All changes are highlighted by the yellow color.
Sincerely,
Jan Cukor, corresponding author
Author Response File: Author Response.docx
Reviewer 2 Report
Authors, thank you for your diligence in assimilating my suggestions into your manuscript and for addressing my comments. I found the revised version of your manuscript to be much improved. I have a few small comments which I hope you will consider.
Specific comments:
Line 28 (Keywords): I believe another reviewer recommended that you add the Latin name of Norway spruce to the Keywords. Generally, Latin and common names should be paired with one another not only in the body of the article but also in the Abstract. You have done this in the Abstract. Therefore, you do not need the Latin names in the Keywords and should remove Picea abies from the Keywords.
Line 65: I know often there are no accepted common names for some species, but I would recommend adding the common names (with the already mentioned Latin names and authorities placed afterward) for these two tree species and for any other species throughout the paper (if possible).
Line 145: The equation would be better displayed if placed on a line by itself.
Line 346: Add the Latin name and naming authority for red deer at first mention. Also check to make sure first mentions of all species include the Latin name and naming authority.
Lines 355-356: I find the structure of this sentence awkward. While your findings may be supported by statistics, they are certainly not proven by statistics. I recommend restructuring this sentence or at least changing the word "proven" to "supported."
Lines 373-374: While deer damage of smaller trees was likely a major factor causing canopy and stocking variability, it was also likely not the only reason. This sentence should be generalized to say that deer preference was likely a major contributing factor in the observed variability.
Author Response
Dear Reviewer 2,
thank you for all your suggestions. All changes were improved throughout the manuscript. Changes are highlighted by yellow color.
Sincerely,
Jan Cukor, corresponding author
Author Response File: Author Response.docx
Reviewer 3 Report
The authors adjusted for the vast majority of the comments from all reviewers. The current version of the manuscript is greatly improved in my view.
The written language is considerably better, but is still not good enough for publication. There were too many mistakes and bad formulations for me to point them individually.
The authors failed to fix one of the major points brought up by reviewer 2 and myself: they refuse to use mixed effects models. I do not agree with their rationale and ask them to reconsider. See attached text for details regarding this and for other comments.
Comments for author File: Comments.docx
Author Response
Dear Reviewer 3,
thank you for all your suggestions. All changes were improved throughout the manuscript. Changes are highlighted by yellow color.
Sincerely,
Jan Cukor, corresponding author
Author Response File: Author Response.docx
