Driving Mechanisms and Changes in Dominant Forest Tree Taxa in Europe Under Climate Change
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsDear authors,
This manuscript presents an attempt to predict different scenarios of the impact of global climate change on the distribution of three main tree species in Europe. You have chosen a fairly interesting topic for research, and analyzed a data array that has allowed you to put forward a number of hypotheses about the future of European forests. However, this manuscript in its current form cannot be published in Forests, and you need to conduct a serious revision, especially in the Materials and Methods and Results sections. You can read the detailed comments below.
Abstract:
1. It should be a little shorter (about 200 words in total, please check https://www.mdpi.com/journal/forests/instructions);
2. There is a doubling of the starting words "However", at the end of the abstract.
3. Personally, I didn't know the abbreviations "SSP126" and "SSP585", and I think they should be removed from the abstract. In my opinion, it is better to just write "optimistic" and "pessimistic", respectively. After this, describe these abbreviations and their meaning in the introduction of the article. This will make the beginning of your work more friendly to a broad audience.
Overall, the introduction lacks a clear description of what similar studies were done by the authors before, and what are the strengths of your work. What is its novelty? Also, I would like to see the methods and data highlighted in this section. What models exactly did you build/use, and which climate/ecological data was used?
Line 38 "[1] or [2,3], or [4–6]." - the meaning of these quotes is not clear here - they are not given proper explanations in the text, please clarify or remove some of them.
The materials and methods need significant revision, because now they are written very vaguely, and it will be impossible to reproduce this work methodologically from this description.
2.1. What was the principle for selecting relevant datasets? In what coordinates exactly was the search carried out? Please provide more detailed information about your search in databases.
"253,242, 389,110, and 374,833, respectively" - what is the meaning of these numbers? Please correct this sentence.
Also, please clarify whether one record corresponds to one individual tree or not?
"Norway spruce, pedunculate oak, and European beech" - Please provide the Latin names of these species in brackets in italics.
2.2. From the first sentence of this section it is not at all clear what factors are being considered - they need to be explained in this sentence in short. No criteria for searching for climate and land-use data are specified: what parameters were used to search the databases? The section seems unclear.
"for 2100 with a 5′ spatial resolution" - I do not understand the meaning of this part;
2.3. This section is also written in an unclear way. On lines 132-135, the authors introduce the concept of "strong correlation" but note that it can lead to overestimation of the model. And then they claim that "less important factors" were not taken into account in the model. It turns out that only those factors that led to their overestimation were taken into account in the models. Please formulate clear selection criteria here. Also, in paragraph 2.3. or in additional paragraphs there should be a full description of these seven models - how exactly they were formulated, or references to the corresponding references should be given. Now these important details are also missing.
Overall, the Results section needs significant revision before publication.
It remains unclear to me which of the seven models the authors chose to build their hypotheses, and I strongly recommend that you make a separate, first section of the results regarding this. Otherwise, everything that happens further in the text of the results becomes even more incomprehensible.
3.1. Line 147 "distribution(Figure 1)." - missing space between words, please, double-check such errors throughout the full text.
What is this concept of "importance value"? It was not addressed in the materials and methods or in the chapter. Please introduce it first and provide a reference or formula for it.
Figure 1 needs serious revision. Firstly, I don't understand the concept of circular plots - instead you could try to draw all this information as a single heat map and not confuse the reader with dots on inclined lines on circular plots. The second point is that the figures are very small and need to be enlarged. The captions should also be enlarged several times. The figure as a whole has very poor resolution and the captions "bio15" themselves should be deciphered either on the figures themselves or on the legend to this figure. Also, I am confused by the fact that all three figures use different scales - A (0-0.25); B (0-0.30); C (0-0.7). Together, this can distort the reader's perception and worsen their experience.
3.2. This section is completely unclear to me. It starts with a description of current trends in the distribution of three species. And then, without discussing how positive and negative scenarios were predicted, or what data are hidden behind them, the reader is immediately given heat maps of tree distribution according to predictions.
Also, the presentation of Figure 2 is very counterintuitive - without proper subplot captions like "Picea abies, SSP126", etc., it is difficult to understand which species is which and where the real and predicted data are. What are the "High" and "Low" values? Is it the density of trees per square kilometer or another unit? The authors do not provide this information at all.
The last sentence (lines 181-182) states that fragmentation will increase under climate change, but this judgment remains quite unfounded, because the authors did not identify any formal criterion for clustering the regions under consideration. Like the previous figure, this one is presented at a low quality and small scale. More space needs to be given to these figures to be displayed adequately.
3.3. Like the previous section, this section is not clear to me.
In the very first sentence, there is reference to supplementary table S4, which should clarify what "suitable" and "unsuitable" habitats mean. However, table S4 gives several parameters: "The maximum training sensitivity plus specificity (MSS) logistic threshold for the tree taxa", which in no way explains these definitions.
All the same comments apply to figure 3 as to the previous ones - low resolution, very small-scale labels in kilometers, and the figure itself, like all the others, is pushed far to the right, toward the edge of the page. A better rearrangement and improvement of this figure is needed. Figure 4 is generally uninformative in its current separate presentation from Figure 5. They can both easily be combined by simply showing all three quantities in Figure 4: current habitat area, SSP126 area, and SSP585 area.
3.4. I have the least complaints about the results of this division, but I do not understand the decision to use different shades of blue and red to represent different tree species. I would suggest building six separate subplots (3 species x 2 scenarios), similar to the ones you built above.
Overall, the discussion section is well written by the authors and slightly improved my overall impression of the quality of the manuscript. However, it did not make any clearer to me whether the authors' results are consistent with previous research in this area, and in what way they extend it. Please add relevant references to your discussion about this point.
I don't have any major complaints about the conclusion, except that the authors should introduce bullet points regarding what was shown for the first time in this article compared to previous ones in this field, to emphasize the significance of this study.
I have some questions regarding the data availability statement. Can you explicitly provide links to the datasets you use to reconstruct your model? And is it possible for you to share the parameters of the model used to predict the distribution of forests? Currently, due to the lack of initial data and data regarding the models used for prediction, your work does not look convincing.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors build species distribution models for three different tree taxa under three different climate scenarios in Europe. In general, the paper is well written and the topic is timely. Some additional detail on the methods used to complete the study, and a deeper discussion of some of the surprising results, would improve the manuscript.
Subsections 2.1 and 2.2 detail the parameters that you used to construct the species distribution models. How do the time periods of the three different tree records from GBIF compare to the time periods of the climate and land factors you got from the other sources for the “current” time period? Similarly, you chose over two thousand records of each of the three tree types to analyze – did you take climate, topography, and land use data from the same locations as those tree records? A 5’ spatial resolution is mentioned for the climate source – are the land use and topography sources at the same resolution?
Line 118 – you note the usage of two GCMs to test the accuracy of future climate information. Is that analysis only in the referenced work, or did you perform this analysis? That distinction should be made more clearly here.
In subsection 2.3 (line 126) you give some information regarding the model construction and the procedure you followed for acquiring the final species distribution model. This paragraph should provide more details. For instance, how did you choose what result was considered positive versus false positive when building the ROC curve or calculating true skill statistic? How did you assign importance values (and the cutoff for minimum importance) for the various parameters? What data points or sample points were used to extract model results?
One missing feature of this overall analysis is an uncertainty estimate. The headline result is future suitable habitat area, which you present in 7 significant digits down to the square kilometer. While you give an AUC and a TSS cutoff, there are not enough additional details for the reader to gauge your confidence in the MSS logistic threshold’s “yes or no” pronouncements.
Figure 2 presents habitat suitability for the three tree types in current and future scenarios. Since at least SSP1 and SSP5 change over time, what year(s) does the figure represent? Are Figures 3 and 4 the same time period(s) as Figure 2?
Figure 6 presents information similar to Figure 5 and, while beautiful, may not be necessary. Shading from yellow to red for decreases is clear, but the blue is difficult to discern. Perhaps green-blue-purple?
There are some surprises in the results that may be caused by the pathway this analysis took. For instance, your model predicted Norway spruce to populate southern and central Spain in the future under SSP5, while the climate there is predicted (in the IPCC report) to get warmer and drier. The SDM for Norway spruce places great emphasis on temperature but not moisture, however there aren’t many Norway spruce trees in dry areas. How does your methodology build that physical factor into the final SDM? Would your method result in an SDM that would place cactus-type plants in swampy areas because other environmental or geographic factors are favorable?
Line 305 – Some climate scenarios predict changes in moisture at 25% or more for SSP5. Depending on one’s perspective, that may be a greater change than a temperature change of 2oC. For clarity, I suggest removing the “which” portion of this sentence.
Some sentences in the abstract could be improved. I suggest in Line 19, end the sentence after SSP585, and start a new sentence with Based. Line 21 try using “variables potentially influencing the distribution of the main tree taxa found ….”. The next sentence (Line 22) should then say that your study determined that the main factors driving…
I did not specifically focus on typos or other minor language suggestions in the manuscript, though I did notice a few candidates:
Lines 170 and 171 – I suggest removing the second sentence’s “however”.
Line 218 – …greater than the changes under SSP126.
Line 238 and Line 243 disagree slightly. It says only pedunculate oak habitat area decreased, but then says European beech area decreased under SSP126.
Line 345 contains one extra ) parenthesis.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe manuscript by Zhou et al. presents an interesting study that investigates the simulation evolution of forest ecosystems (tree taxa) in Europe under climate change for three representative tree species: Picea abies (Norway spruce), Quercus robur (Pedunculate oak), and Fagus sylvatica (European beech). The authors employ two contrasting climate change scenarios—an optimistic and a pessimistic one—to model forest distribution, incorporating climatic, topographic, and land-use variables.
The introduction and objectives are well-structured and clearly articulated, providing a strong foundation for the study. It could be interesting to explain more about the terminology "tree taxa" because it is in the title (or change the title). However, the methodology section would benefit from the inclusion of a summary table detailing occurrence data and including some of the supplementary tables in the text, which would enhance transparency and reproducibility. The description of the modeling approach is relatively concise; while this may reflect the simplicity of the model used, a more comprehensive explanation would aid readers unfamiliar with its structure and assumptions. A point of concern is the use of the optimistic scenario, which, in my opinion, may already reflect current conditions in certain regions. This limits its utility for future projections and reduces the model’s relevance in anticipating ecosystem shifts (possibility of using more scenarios?).
The figures of the results are great and perfectly describe the modeling data predictions. However, increasing their size would improve readability and allow for better appreciation of specific places described. Figure 5 requires the addition of significance levels and error metrics to support the robustness of the findings.
The discussion and conclusions are generally appropriate and well-reasoned. Nonetheless, given the reliance on predictive modeling, a stronger emphasis on current empirical data would enhance the credibility and contextual grounding of the study.
No major typographical or language errors were identified. Overall, the manuscript presents a straightforward, concise study with clear scientific value, though there are opportunities to strengthen the methodological and interpretative aspects.
Thank you for your work.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsDear Authors,
After revisions, your manuscript has improved.
Now I think your work can be published in Forests.
A few points that need to be corrected in the process of finalizing the manuscript are below.
1. The captions to Table 1 should decipher the abbreviations of the model names - CTA, FDA, GBM, GLM, MAXNET, RF. The reader should be able to read tables and figures as easily as possible, and their captions should be self-explanatory. It is also important to indicate how you have made integrative models in the materials and methods section.
2. I like the new figure 1. However, I would advise you to re-arrange it - place figures A, B, C under each other and increase their size for better perception.
3. In the text captions to Figures 2–5, please indicate explicitly which scenario is negative and which is positive, because only abbreviations are provided now.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors have addressed the comments and suggestions satisfactorily. I still think Figure 5 (formerly Figure 6) has three blue colors (for "increasing") that all look very similar to each other, but if your stronger eyes have no difficulty I will not insist upon changes. I look forward to seeing other researchers' reactions to the methods and results presented in the paper.
Possible typo or copy/paste issues:
Line 21: ...three tree taxa. Suitable habitats...
Line 25: remove the word "two"
Line 124: Slope is capitalized, also there is a missing space near parenthesis.
Line 177/178: "...using three statistical metrics:...", but only AUC is listed in this sentence.
Line 217 is not a full sentence, maybe remove it, and add "optimistic" and "pessimistic" to the following sentences
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
