The Impact of Fine-Scale Present and Historical Land Cover on Plant Diversity in Central European National Parks with Heterogeneous Landscapes
Round 1
Reviewer 1 Report
This is a very interesting study concerning the impact of land use changes and land abandonment in floristic and landscape attributes. A lot of results are produced and some of them are very interesting and can stimulate further research. However, I found very difficult to follow the part describing the statistical analyses and later on the section of the results. You have made a lot of different analyses and it is not clear how each analysis complements the rest of analyses. First of all, you should describe the statistical methods with a little bit more detail. Please, for each method explain which question is aiming to address and why you have selected the specific method (the latter only for the instances where such an explanation would help the reader). Try to structure the material and methods (the section of statistical analyses) in a similar order with that you present the results. This would help the reader to follow better both sections. This thing comprise my major concern and to be honest I had to read 2-3 times the corresponding sections and even after this I am not confident that every analysis is necessary and justified. I have some objections also about some things concerning the analyses that you can found in the detailed comments below. Also, some minor issues are included in the detailed comments.
line 37: in
line 37: with the so-called
line 38: the introduction of the so-called
Figure 1. In the map of Fig. 1 four colours are presented, while in the legend of the map three of them are included (the dark green colour is not explained).
Line 140: the number here contradicts that given in line 109. The reader cannot understand why there is this difference.
Lines 164-171: It would be better to give more details about the way that you classified certain land cover classes. Specifically, how you distinguished the closed forests from sparse forests (i.e., which was the tree cover threshold)? Similarly, further details may be needed for the classes grass-forb, meadows and fallow land.
Line 186: It is not clear how you used the categories of endangered species. Do you mean the number of species per category?
Lines 190-192: Probably this sentence was left from a template
Lines 195-196: I do not understand it. Did you apply PCoA using three different distance measures? Do you present somewhere the corresponding results?
Lines 253-257: Do you presents somewhere these results (at least in an appendix)?
Lines 286: what do these ellipses represent? This is important to report as the comparison between the different cases is understood by interpretating these ellipses!
Line 290. I do not understand this: “p < 0.001, resp. p < 0.002 by Nutrients or p < 0.003 by soil reaction”. Why you are reporting p values for two out the five EIV variables presented in the figure.
In regression trees you applied a subjective way of pruning. Why did you follow such a choice, and you did not use an objective criterion?
In line 375 you report that the main driver of species richness is the river phenomenon, while in line 345 you report that the river phenomenon has not major impacts on species composition. Is such a thing possible? How this discrepancy can be interpreted?
Figure A1: what is the category “most” in this figure
Figure A2: Which distance measure was used for the analysis? You report two distance metrics.
Line 510: 8. 8. And differences?
Author Response
Thank you for your review and helpful comments. The responses are indicated in blue.
This is a very interesting study concerning the impact of land use changes and land abandonment in floristic and landscape attributes. A lot of results are produced and some of them are very interesting and can stimulate further research. However, I found very difficult to follow the part describing the statistical analyses and later on the section of the results. You have made a lot of different analyses and it is not clear how each analysis complements the rest of analyses. First of all, you should describe the statistical methods with a little bit more detail. Please, for each method explain which question is aiming to address and why you have selected the specific method (the latter only for the instances where such an explanation would help the reader). Try to structure the material and methods (the section of statistical analyses) in a similar order with that you present the results. This would help the reader to follow better both sections. This thing comprise my major concern and to be honest I had to read 2-3 times the corresponding sections and even after this I am not confident that every analysis is necessary and justified. I have some objections also about some things concerning the analyses that you can found in the detailed comments below. Also, some minor issues are included in the detailed comments.
Yes, it is a little bit complicated, because there is partly an overlay. We greatly edited part of Methods describing statistical analyses and connected it to our questions. In addition, we removed one figure from Appendix, Fig. A1.
line 37: in
Repaired.
line 37: with the so-called
Repaired.
line 38: the introduction of the so-called
Repaired.
Figure 1. In the map of Fig. 1 four colours are presented, while in the legend of the map three of them are included (the dark green colour is not explained).
The dark green colour is the colour of forests from the background map. The background map also contains water courses in blue and settlements in grey. We don’t think it is necessary to describe these features since they concern the background map.
Line 140: the number here contradicts that given in line 109. The reader cannot understand why there is this difference.
The number in line 109 indicates historical records of the plant species. The number in line 140 indicates records from the first and second surveys together after removal of hybrids. This is stated in the lines 146-152.
Lines 164-171: It would be better to give more details about the way that you classified certain land cover classes. Specifically, how you distinguished the closed forests from sparse forests (i.e., which was the tree cover threshold)? Similarly, further details may be needed for the classes grass-forb, meadows and fallow land.
Specified.
Line 186: It is not clear how you used the categories of endangered species. Do you mean the number of species per category?
Specified.
Lines 190-192: Probably this sentence was left from a template
Thanks. Removed.
Lines 195-196: I do not understand it. Did you apply PCoA using three different distance measures? Do you present somewhere the corresponding results?
Specified.
Lines 253-257: Do you presents somewhere these results (at least in an appendix)?
One table was added, it is now Tab. 1.
Lines 286: what do these ellipses represent? This is important to report as the comparison between the different cases is understood by interpretating these ellipses!
Specified.
Line 290. I do not understand this: “p < 0.001, resp. p < 0.002 by Nutrients or p < 0.003 by soil reaction”. Why you are reporting p values for two out the five EIV variables presented in the figure.
Specified.
In regression trees you applied a subjective way of pruning. Why did you follow such a choice, and you did not use an objective criterion?
Regression trees are used to both modelling predictions (which requires large training data set) and to explain relationships. We used it for the latter, because there were only 115 grid cells within the second survey and unfortunately only 78 grid cells according to different survey methodology within the first survey.
In line 375 you report that the main driver of species richness is the river phenomenon, while in line 345 you report that the river phenomenon has not major impacts on species composition. Is such a thing possible? How this discrepancy can be interpreted?
We think that it fits very well! Species richness express the alpha diversity, here we used the simple number of all recorded species per grid cell. Also the species richest grid cells lie along the river, because many different habitats formed according to the strong river phenomenon (varying geomorphology combined with unique meso-climate) harbour many different species. However the species composition is mostly driven by distribution and the proportion of land covers, because people formed many different types of artificial or near-natural habitats/land covers which are also not too species rich.
Figure A1: what is the category “most” in this figure
We wanted to show here that the species richness of grid cells, which were surveyed only on area 72-135.99 ha (the group “most” in Fig. A1; the limit 72 ha was revealed within regression trees) is really comparable to grid cells surveyed on the whole area (136 ha). Then we added these grid cells accordingly to “inner” or “river” group. However, this figure can be confusing, therefore we removed it.
Figure A2: Which distance measure was used for the analysis? You report two distance metrics.
Hellinger transformation applied on some dissimilarity measure in called Hellinger distance.
Line 510: 8. 8. And differences?
Repaired.
Reviewer 2 Report
Researching the factors of plant diversity is necessary to preserve biodiversity. Different with almost previous study, this paper focus on the impact of present and historical land cover in fine scale. Nevertheless, there are some small problems in this paper and I recommend the author should pay more attention to my following suggestions: 1) You should describe more aspects of the human impacts and reasons on the land-cover changes, rather than only state the change of arable land in Introduction and Discussion. 2) the conclusion about N deposition isn’t supported by the data of your research. 3) Some factors you use as impacts of species richness not to explain their influences well in this paper. For example, you mention original bedrock types for geological diversity many times, but you didn’t make clear why they had influences. 4) I think you may need to describe some words and figures used in the paper more clearly, such as relative areas in Figure 2 and meso-climatic diversity.
Comments for author File: 
Comments.pdf
Author Response
Thank you for your review and helpful comments. The responses are indicated in blue.
Researching the factors of plant diversity is necessary to preserve biodiversity. Different with almost previous study, this paper focus on the impact of present and historical land cover in fine scale. Nevertheless, there are some small problems in this paper and I recommend the author should pay more attention to my following suggestions: 1) You should describe more aspects of the human impacts and reasons on the land-cover changes, rather than only state the change of arable land in Introduction and Discussion. 2) the conclusion about N deposition isn’t supported by the data of your research. 3) Some factors you use as impacts of species richness not to explain their influences well in this paper. For example, you mention original bedrock types for geological diversity many times, but you didn’t make clear why they had influences. 4) I think you may need to describe some words and figures used in the paper more clearly, such as relative areas in Figure 2 and meso-climatic diversity.
Ad 1: We have added information in the Introduction section about other impacts and reasons on land cover changes, mainly connected with population growth. The discussion section includes also changes of other land cover categories – it specifically mentions change of different forest forms (from sparse/open canopy forests to close canopy forests) as well as causes and consequences of abandonment (original lines 393-406).
Ad 2: We did not use the direct measurement of N deposition, but we used Ellenberg indicator values calibrated for the Czech flora (Chytrý et al. 2018). These bioindicator scales were based on the original work proposed by Heinz Ellenberg for the flora of Germany (Ellenberg 1992), which was routinely used to rapidly estimate site conditions from species composition, when measured values of environmental variables are not available (Diekmann 2003). This method was not used generally only in vegetation science, but also for example in invertebrate research (Horsák et al. 2007, Zhai et al. 2015). Of course, it also has some limitations (Zelený & Schaffers 2012), but we took it into account (see Methods) and we believe that it is appropriate.
Ad 3: Similarly as the dependence of plant species richness on calcareous bedrock (Cox et al. 2016) or soil pH is well known (Ellenberg 1988, Pärtel 2002, Chytrý et al. 2010) on both local and regional scale (Chytrý et al. 2003), the plant adaptation to either basic or acidic conditions influencing the plant species composition was described (e.g. Tyler 2003) as well as the impact of the connected geological diversity (Kohn & Walsh 1994; it was mentioned in Introduction). All these connected variables represent a strong abiotic environmental filter for plant species. Even directly in our study area was formerly documented the significant relationship of soil variables and species composition (Zelený & Chytrý 2007) or geological diversity (Michalcová et al. 2014), therefore we decided to remove this impact from our analyses to investigate our questions clearly. In our case of grid cells (regional scale), it seemed to be appropriate to use the geological diversity.
We slightly modified the text.
Ad 4: We used percentages of the whole grid cell area. Symbols were printed within each y-axis labels, but in addition we added the explanation into figure and analyses description.
Meso-climatic diversity corresponds with our investigated regional scale (grid cells) and describes various climatic conditions on e.g. one rock, which is on its upper south-faced part very warm, on the lower south-faced part rather mesic and on the bottom by the river cold. Such differences are much greater within river sides in the deep valley and are described directly from our study area via phenological mapping (Chytrý & Tichý 1998).
Below are minor comments:
Line 18 I couldn’t understand what is river phenomenon very well from this paper.
Short connection added into Introduction.
Line 23 I think it’s better to use the bilateral Podyjí and Thayatal National Parks, rather than NPs.
We changed the NP into national parks. We did not want to specify the national parks because we believe that these findings are true also for other national parks, based on the study of Gallego-Zamorano (2022). However, we added specification of our studied area (bilateral Podyjí and Thayatal National Parks) to line 16-17, in order to tell the reader (of the abstract) where our study was conducted.
Line 30-41 you mainly focus on the land cover changes by agriculture. But other human activities may also lead to it that you didn’t mention them clearly.
We added more information about other human activities and their impact.
Line 53 & 349 I don’t understand what is meso-climatic diversity in this paper.
Explanation above.
Line 81-83 I think you should put the small location (between the towns) in front of the large location (between the countries).
Repaired as suggested.
Line 154 I think you may want to use m2 instead of m2.
Thanks. Repaired.
Line 172 I think built-up areas will include many parts of industrial, agricultural and commercial areas.
Thanks for pointing this out – it was a mistake, built-up areas in this case represent residential and public buildings. We have amended the text accordingly
Line 182-185 & 243-244 you use original bedrock types to represent geological diversity, but I couldn’t know why and how they will affect species richness.
Explanation above.
Line 189 I think you might forget the right parenthesis in this sentence.
Thanks. Repaired.
Line 237 you may use a wrong symbol as beta.
Thanks. Repaired.
Line 253-257 I didn’t see any figures or tables that describe this result.
We added a table.
Line 259-260 I see the result of meadow is in figure 2d, that of grass-forb vegetation is in figure 2e and that of fallow land is in figure 2f.
Thanks. Repaired.
Line 259-260 I can’t obtain the result in figure 2c, because you only appear the result of field, not orchards, vineyards with trees, fields with trees and small fields, gardens and giant fields.
We did not include graphs from these land cover categories because they are not that widespread in the studied area and we did not want to overwhelm the article with so many figures.
Line 266 I couldn’t see a clear definition of relative areas.
An explanation added.
Line 268 I think the letter p should be italic as p-value.
Thanks. Repaired.
Line 283 I can not understand what are C1 and C2 in Figure 3.
Thanks. Explanation added.
Line 300-302 I couldn’t draw the conclusion that the highest impact was caused by the historical land cover, followed by geomorphology and present land cover from Figure 5.
We repeated the link from previous sentence to Tab. 1, while the link to Fig. 5 should supply the distribution and shift of species composition according to land cover groups. Another explanation added into Fig. 5 description.
Line 305 I’m not sure what the three groups of land-use represent in Figure 5.
They represent groups of grid cells based on land cover classification. Another explanation added into Fig. 5 description.
Line 324 & 329 I couldn’t know what the above number in each node means.
Explanation added.
Line 425-434 & 468-469 the conclusion that N deposition could affect species richness isn’t supported by the data of your research.
We explained it above.
Line 487 I’m confused about these four groups of cells, because you divided grid cells into three groups (fringe cells, inner cells and river cells), but there were one more group (most cells) and the different definition of fringe cells in Figure A1.
We wanted to show here that the species richness of grid cells, which were surveyed only on area 72-135.99 ha (the group “most” in Fig. A1; the limit 72 ha was revealed within regression trees) is really comparable to grid cells surveyed on the whole area (136 ha). Then we added these grid cells accordingly to “inner” or “river” group.
You are right, this figure is probably confusing. We removed it.
Line 491 I don’t think When notches do not overlap then groups differ is a complete sentence, but you use full stop after it.
This figure was removed.
Line 514 I think the references format in this paper isn’t consistent. For instance, you use pp. 281–319 to show the pages of this reference in line 575, but you use 279–302 in line 516 to do the same thing. And you make the year 2016 italic in line 647, but make the year 2021 bold in line 516 at the same time.
Line 626-629 I guess you forgot to remove the highlight.
Repaired.
Reviewer 3 Report
The manuscript “The impact of fine-scale present and historical land cover on plant diversity in central European National Parks with heterogeneous landscapes” study the effect of landscapes change on the plant diversity. The manuscript provides a better understanding of biodiversity maintaining mechanism with growing human population. This study covered long time and used better methods. Following points need author’ response
1) The manuscript revealed that river phenomenon was main driver in species composition, but in your data analysis did not find these content.
2) If regression trees analysis reveal the drivers of species composition, the author should provide more details.
Author Response
Thank you for your comments. Our responses to them are stated in bold:
1) The manuscript revealed that river phenomenon was main driver in species composition, but in your data analysis did not find these content.
We reformulated the text about the river phenomenon to better understand it.
2) If regression trees analysis reveal the drivers of species composition, the author should provide more details.
The species composition was analysed within the Distance based redundancy analysis (db-RDA), results are shown in Table 2. While regression trees were used to analyse the species richness.
We restructured the section about used analyses in Methods to better understand it.
Reviewer 4 Report
The paper is suitable for the journal and the special issue.
The abstract summarises well the paper.
The introduction is rich in references to related studies in the part of Europe where the case study is situated. However, the connection to the discussion and conclusions needs improvement from this point of view. Reference is done to other parts of Europe without properly specifying where. A comparison to the Danube flowing through Austria also with national pars would be helpful.
The methods are proper and well described. Unclear remains why the appendix figures are not integrated in the paper since they are related to the other figures.
Author Response
Thank you for your comments. Our responses to the are stated in bold:
The introduction is rich in references to related studies in the part of Europe where the case study is situated. However, the connection to the discussion and conclusions needs improvement from this point of view. Reference is done to other parts of Europe without properly specifying where. A comparison to the Danube flowing through Austria also with national pars would be helpful.
Thanks. We incorporated such studies.
The methods are proper and well described. Unclear remains why the appendix figures are not integrated in the paper since they are related to the other figures.
Thanks. The idea was to do not overwhelm the paper with many figures. Moreover, we removed the Fig. A1 from Appendix according to suggestions from reviewer 1.
Reviewer 5 Report
The article was written in a clear and precise language, and its message is quite clear. Against this background, I only have two suggestions to make:
- There are some spelling mistakes that need to be corrected before the article goes to print (e.g., "Another significant decrease was documented [...], on the contrary, these [?] was an increase within gardens and giant fields (>30 ha)" (R262-264) should probably read "[...] there was an increase [...]".
- I would personally spell out the abbreviations on their first use (e.g., "NPs" appears in the abstract, but the full spelling only appears in R80. The same applies to EIV, etc.).
Author Response
Thank you for your comments. We have edited text as suggested.
Round 2
Reviewer 1 Report
My comments-suggestions were addressed and changes were made or proper justification was provided.
Author Response
Thank you.
Reviewer 2 Report
Researching the factors of plant diversity is necessary to preserve biodiversity. Different with almost previous study, this paper focus on the impact of present and historical land cover in fine scale. However, there are still some problems in this paper and I suggest the author to pay attention to the details. My major comments are: 1) You need to supplement research data and analysis results to support your conclusion soil nutrients strongly affect the species richness even in NPs with low current N deposition. 2) The format of references should be checked carefully to keep it consistent, such as expression of page and bolding numbers of year.
Author Response
Thank you for your comments.
We have unified the format of all references.
Concerning the issue of nutrients, we believe that our conclusions are supported by our data and analyses expressed in figures 7 and A2a, which are also described in the text.
Reviewer 3 Report
Authors improve better the mamuscript and i agree with these revision. So, the manuscript could be accepted in this vesion.
Author Response
Thank you.
