The Influence of Wild Ungulates on Forest Regeneration in an Alpine National Park

Round 1

Reviewer 1 Report

 

The article presents results of monitoring forest regeneration over a period of thirty years in a context of high herbivore density. The study includes the monitoring of 168 plots in the Swiss National Park. The data set is substantial and deserves to be published. The article is relatively clear but the significance of the results needs to be better discussed in relation to the scientific literature in the field.  It requires also revision of the figures and some details in the text (see below for detailed comments).

Comments for author File: Comments.pdf

Author Response

General response: Thank you for this excellent review! In our opinion, this is a good example of how careful peer review can improve the quality of a scientific publication.

The article presents results of monitoring forest regeneration over a period of thirty years in a context of high herbivore density. The study includes the monitoring of 168 plots in the Swiss National Park. The data set is substantial and deserves to be published. The article is relatively clear but the significance of the results needs to be better discussed in relation to the scientific literature in the field. It requires also revision of the figures and some details in the text (see below for detailed comments).

Introduction

Our study shows a lack of control of regeneration by herbivores. Are there other studies that show this result or is there consensus in the literature that herbivores have a negative impact?

A broader view of the long-term effects of herbivores on regeneration would help prepare the discussion.

Response: We added a short discussion and references.

Method

2003. 38-39 The material and method is relatively clear and honest notably in describing the protocol and its evolution between 1991 and 2003.
2004. 138 Is it only the terminal shoot that was observed or are the shoots of the side branches were also observed?

Response: Thanks, we now mention that it was the apical shoot.

Table 2 Explanation for developmental stage 7 is not clear for me. Is there: height > 130 cm and BHD <7.99 cm?

Response: We clarified this in the table.

1. 144 « a height of 130 cm to a breast height diameter of 24 cm » instead of « a height of 130 cm from a breast height diameter of 24 cm

Response: We have reformulated this sentence.

1. 186 Specify all location parameters here.

Response: We added specifications of location factors.

1. 187 and L. 202 It seems that it is not the diversity of the trees as whole (including adults) but of the diversity of sapling and young trees. Could you clarify this point in the text?

Response: We clarified this point.

1. 198-208 How were differences in plot size between saplings (4-m radius) and young trees (8-m radius) taken into account in the calculations of species diversity and average development stage?
2. 204 and 205 : Is it only sapling or sapling and young trees?

Response: Our model results are only on the 4-m radius and thus about saplings, not about young trees. We have tried to make this clearer throughout the text.

Table 4: Reasons should be supported by bibliographic references.

Response: We added some references.

Table 4 – « Total number of tree » and L. 239 : Is there the number of whole tree or only the number of sapling and young tree ?

Response: We changed to «total number of saplings».

Table 4 – « A high diversity of sapling represents different important nutrients and minerals » : This point is not clear to me.

Response: We changed to «A high diversity of saplings may represent a resource of different nutrients and minerals».

1. 222-223 Did you tested interaction between species and development stage on the one hand ; and between species or development stage and topographic/location variables ?

Response: We now included an explanation why we did not include interactions in our model.

Why did you not test effect of year on browsing probabilities ?

Response: We included year as a random factor, thus accounting for differences among years.

1. 225 Could you indicate here that you have tested the correlation between the predictors?

Response: We referred to the correlations listed in table S3 in the results section.

Results

Development of the numbers of tree over time

1. 238-270

Figures 3 and 4 are not easy to read (and therefore not very suitable) due to out-layers (extreme values) that compacts main result. Tables would perhaps be more appropriate. Another solution would be to truncate y-axis or use log-scale. Moreover, Y-axis in figure 3 and 4 should be expressed as a density (number of sapling per ha) which would allow a comparison between saplings and young trees. Figures in the text should be also expressed as a density.

Response: Our main change in the revision was that we now express all count data as densities, i.e., as numbers of saplings or young trees per hectare, which of course makes a lot more sense. We thank the referee for this extremely valuable suggestion! We also log-transformed the data and thus the outliers disappeared.

I am personally bothered by the lack of a median comparison test in 3.1.1 and 3.2.2 parts. Paired nonparametric test could be appropriated to support assumption in the text.

Response: We added the factor of annual increase in numbers of saplings and young trees ± CI. If somebody wants to apply traditional null hypothesis significance testing, they can check the CI.

1. 271-294

The number of individuals between development stage should be expressed as mean or median between sampling plot, not as the total number. Ideally, value should be expressed as mean/median density. It is not clear to me how the values for stages 1-6 and 7-9 were compared as they were not measured on the same plot area.

Figure 5 is not easy to read. You should adapt y-scale for upright mountain pine and mountain ash. Moreover, Y-axis should be expressed as a density to allow comparison with other publications. Legend of years within figure 5 is shifted.

Response: We edited the text and now report densities throughout.

1. 283-284 This assessment should be correctly tested with an appropriate statistical test.

Response: We now refer to section 3.1.2. in which we gave factors of annual increase in numbers of saplings and young trees ± CI.

1. 296-299 This part is already indicated in the material and method part and should not be repeated in the results.

Response: Thanks, we deleted those sentences.

L.299-305 Is there a significant effect of species and/or development stage and/or their interaction?

Response: We give compatibility intervals throughout this section. If somebody wants to apply traditional null hypothesis significance testing, they can check the CI (see our reference 36 and https://www.nature.com/articles/d41586-019-00857-9 for a discussion on "statistical significance"). For interactions, see our comment above.

1. 303-304 Does this result only apply to the larch as indicated in the legend to figure 6?

Response: We edited the legend of Figure 6 to clarify this.

Could you propose figure that shows variables predictors and their compatibility interval (Table S2).

Response: Predictors and compatibility intervals are shown in Figure 6 and 7.

1. 315-317 If I read S1 correctly, some predictors did not include both an increase and a decrease in the probability of browsing: Eastness.z and northness.z

Response: Yes, these are the effect sizes for which the 95% CI does not overlap zero. We exchanged the header of Table S1 to make this clearer.

L.317 I do not understand why northness was an ambiguous predictor because increase of northness decreases probability of browsing in 95% compatibility interval.

L.321 I do not understand the following part of the sentence: “yet unclear”. Increase of eastness increases probability of browsing in 95% compatibility interval.

Response about L. 317 and 321: Thanks, we agree this was unclear. We rewrote the entire section.

1. 323 The interpretation of effect of slope appear delicate.

Response: Yes, but effects do not "suddenly assume the mantle of reality" only because a threshold is crossed (Oakes 1986, "Statistical inference: commentary for the social and behavioural sciences", p. 85–86). We have extensively published on statistical significance and non-significance and are happy to provide references, if needed.

Figure 7 : X-legend is shifted

Response: Thanks, corrected.

Discussion :

The ecology of the main species is not well considered and their consequences on regeneration. Larch is a heliophilous and pioneer species. This species probably regenerates in your stands in association with avalanche paths.

Response: Thanks, we added this information to the discussion.

The regeneration of this species is probably extremely heterogeneous in your plots (with the presence of very high values) and may partly explain the high regeneration values in your plots. Cembro pine is a less heliophilous species and will succeed larch in plant succession.

Response: Our impression is that in our study site, saplings and young trees of Larch are relatively evenly distributed.

Your discussion and conclusion remain focused on your site. It would be interesting to put your results in perspective with the literature on the increase of ungulates in Europe and their impacts.

Response: Because a national park with such a high protection status is very unusual in Europe, we refrain from reviewing ungulate densities and their impacts in other parts of Europe. In our view, such a review would probably have to be the length of an entire paper.

1. 352 Can you qualify the relatively small increase in herbivores?

Response: The reasons for the development of herbivore populations in the Swiss National Park is under investigation; they are currently too unclear for us to comment on them in this paper.

1. 360 Is a 4-year herbivore exclusion experiment long enough to conclude that there is no long-term effect? Other longer-term studies in other geographical contexts show an effect. I would qualify the statement here. It seems to me a pity that this point is not discussed more in view of the numerous literatures on the subject.

Response: We have now described the conclusion from the cited study more cautiously and added some more information.

1. 365-367 You don't test this assertion properly, which is unfortunate because you have the data to do so. This assertion is probably especially true for larch.
2. 392 La faible abondance des frênes m’interroge du fait que c’est une espèce fortement consommée par les herbivores.
3. 399 Could there be a link to the periodicity of avalanches in the area? A link between the size of individuals and resistance to disturbance?

Response: All these are interesting topics for further research; we now deleted the mentions of avalanches because this is not the topic of the current study.

Reviewer 2 Report

The authors present an interesting manuscript on the influence of wild ungulates on forest regeneration. Study design and research questions are clearly described. In this sense, it is easy to understand the aim of this study. The bright side of the manuscript is that to provide some useful practical details on related topic. In this context, the study contributes to different fields. Only minor concerns are raised. Therefore, I would like to make some suggestions to improve the quality of the paper as below:

Line 22: A few words about how the results contribute to further studies and different fields would be valuable. In my opinion, it is always good to finish the abstract with such a sentence.

Lines 40-42: “In strictly protected areas like national parks, such management systems are mostly not applied, and apart from providing a natural habitat, forests mostly do not have a specific function.” I do not agree with the authors. Forest provides many ecosystem services. In my opinion, authors should rephase or delete this sentence.

 

If possible, flora and fauna of the area can be explained briefly. I mean, how many plant species, mammals and birds live in the area can be explained with 1-2 sentences in Introduction or Methods (in study area subsection). This would show the biodiversity and thus importance of the area. 

Author Response

The authors present an interesting manuscript on the influence of wild ungulates on forest regeneration. Study design and research questions are clearly described. In this sense, it is easy to understand the aim of this study. The bright side of the manuscript is that to provide some useful practical details on related topic. In this context, the study contributes to different fields. Only minor concerns are raised. Therefore, I would like to make some suggestions to improve the quality of the paper as below:

Line 22: A few words about how the results contribute to further studies and different fields would be valuable. In my opinion, it is always good to finish the abstract with such a sentence.

Response: We added a short conclusion to the abstract.

Lines 40-42: “In strictly protected areas like national parks, such management systems are mostly not applied, and apart from providing a natural habitat, forests mostly do not have a specific function.” I do not agree with the authors. Forest provides many ecosystem services. In my opinion, authors should rephase or delete this sentence.

Response: We agree and deleted the sentence.

If possible, flora and fauna of the area can be explained briefly. I mean, how many plant species, mammals and birds live in the area can be explained with 1-2 sentences in Introduction or Methods (in study area subsection). This would show the biodiversity and thus importance of the area.

Response: We added some information in the beginning of the methods section.

Reviewer 3 Report

Dear authors, these are my short comments on the article:

The text (lines 60 to 68) are a description of the methods and should be in the work methods chapter.

Figures 3 and 4 are missing legends for descriptive statistics.

Figure 7 (top row) for images with exposures, the display on the x axis is unclear.

Line 442, it's a shame you don't have that information.

For discussion:

Based on your research, do you have any recommendations regarding game in national parks? Should wild game in national parks be managed or not?

Should there be some changes in laws and regulations regarding the management of wild animals in national parks?

Can you show us what would be the "normal" number of game animals per km2 in this national park or for that type of habitat?

Are there any studies in the neighboring area of the national park regarding the preferability of damage to forest trees?

Author Response

Dear authors, these are my short comments on the article:

The text (lines 60 to 68) are a description of the methods and should be in the work methods chapter.

Response: We think it is helpful for the reader if some basic information on what was done in the study is also given in the introduction.

Figures 3 and 4 are missing legends for descriptive statistics.

Response: We added a description of the boxplots in Figure 3.

Figure 7 (top row) for images with exposures, the display on the x axis is unclear.

Response: We adapted the figure.

Line 442, it's a shame you don't have that information.

Response: We agree it would have been great to have data also on the steepest terrains, but we now deleted our sentence complaining about the lack of such data. We do not extrapolate beyond our obtained data and therefore simply cannot discuss the steepest terrains.

For discussion:

Based on your research, do you have any recommendations regarding game in national parks?

Should wild game in national parks be managed or not? Should there be some changes in laws and regulations regarding the management of wild animals in national parks?

Response: We think we cannot make such general statements, because each country and each national park will have to deal with their own ecological, social, and political environments.

Can you show us what would be the "normal" number of game animals per km2 in this national park or for that type of habitat?

Response: We are not aware of any published data on what would be "normal" for this sort of habitat. We have a reference in the beginning of the discussion saying that the density of red deer in Val Trupchun is very high compared to other regions in Switzerland and neighboring countries.

Round 2

Reviewer 1 Report

I accept the manuscript in present form. Thanks you to taking my comments into account.