Influence of the Declaration of Protected Natural Areas on the Evolution of Forest Fires in Collective Lands in Galicia (Spain)

Round 1

Reviewer 1 Report

You have identified an important problem that deserves more discussion.  Unfortunately without a much more detailed introduction, the reader is left wondering about the answers to several questions, and wondering if many of your assertions are really supported by evidence in the literature.

It would be a good idea to state near the start that all over the world, studies show that arson is responsible for most fires where the cause is known.

Other authors in your bibliography looked at socioeconomic and political patterns in wildfires in Galicia.  If you believe you are the first to consider patterns related to conservation policy, you should say so. More generally you should explain all your motivations for conducting this study.

Some of your cited authors see a terrorist motivation for some arsonists.  There is an extensive literature about the psychology of arsonists and of terrorists.  You should probably mention at least briefly relevant findings from that literature. 

Are there any indications for what the arsonists in Galicia think they are gaining?  If this is arson as political vengeance as you seem to think, do they make announcements? Are they acting in groups?  Is there politically motivated arson in neighboring Portugal?

Studies indicate several arsonists are mentally ill which complicates an already complicated interpretation.  Land protection may lead to less attention to who is walking on the land, creating more opportunities for arsonists with no political motivations.

Your suggestions for improving conservation policy seem to be based on previous policy discussions which you should cite.  How can your results from this study be used to improve conservation policy or to lower arson in the future – or at least what further data added to yours would help?

7    Council Directive 92/43   =>   1992 European Union Council Directive 92/43

8-10      of protected land, which, according to several scientific studies, seems to be shown by the occurrence of arson fires in 10 these areas   =>    of land affected by the directive.  Some studies have indicated that some of this opposition has found expression in subsequent arson fires in these areas.  

You need to explain in the abstract why 2004 is significant

130-135  How lack of management fits in the discussion.

144-153     =>  Using a single statistic for the amount of fire is problematic.  In some cases firefighting extinguishes a fire before it burns all of the area that the arsonist apparently intended to affect and the arsonist returns to set new fires.  In other cases where the fire meets the arsonist’s goals, there is only one arson attempt.  Therefore we use two complementary measures –  fire rates (Nº of fires/100 ha) and area burned by fire (ha).

165-166    It is important to justify why you are including fires where the coded cause was ‘unknown’.

169-170     For this reason, we have chosen to use annual data; a greater disaggregation of the data does not contribute anything.     =>      We have chosen to use annual data; a greater disaggregation of the data does not contribute anything.

203    so the weather is not a determining factor in their occurrence     =>        and weather is not a determining factor in their occurrence  

366   It is not clear 

398 - 499   Needs  rewriting.

524-5    why would the fires certainly be lower?  (more valuable land is less likely to burn? – citation needed)

526-528   this statement appears unsupported

Author Response

Dear Sir or Madam,

We would like to thank you for your time and kind contributions concerning this paper. In the next lines we will try to give an answer to your notes and comments:

8-10      of protected land, which, according to several scientific studies, seems to be shown by the occurrence of arson fires in 10 these areas   =>    of land affected by the directive.  Some studies have indicated that some of this opposition has found expression in subsequent arson fires in these areas.  

Thank you very much for your recommendations. We have amended both sentences.

You need to explain in the abstract why 2004 is significant

Thank you. We have modified the abstract.

130-135  How lack of management fits in the discussion.

We hope that the changes made have helped to better explain our arguments.

144-153     =>  Using a single statistic for the amount of fire is problematic.  In some cases firefighting extinguishes a fire before it burns all of the area that the arsonist apparently intended to affect and the arsonist returns to set new fires.  In other cases where the fire meets the arsonist’s goals, there is only one arson attempt.  Therefore we use two complementary measures –  fire rates (Nº of fires/100 ha) and area burned by fire (ha).

Thank you very much for your kind advice. We have replaced the original text with the sentence you recommended.

165-166    It is important to justify why you are including fires where the coded cause was ‘unknown’.

Thank you for your advice. We have included a brief explanation in the paper. Actually,  from professional experience gained in the Forest Fire Defence Service of the province of Orense (Galicia), we know that the coding "unknown" only means that the cause of the fire has not been investigated, even if it is known that the fire was due to anthropogenic reasons.

169-170     For this reason, we have chosen to use annual data; a greater disaggregation of the data does not contribute anything.     =>      We have chosen to use annual data; a greater disaggregation of the data does not contribute anything.

Thank you very much for your kind advice. We have replaced the original text with the sentence you recommended.

203    so the weather is not a determining factor in their occurrence     =>        and weather is not a determining factor in their occurrence  

As in previous cases, we have replaced the original text with the sentence you have recommended. Thank you very much.

366   It is not clear 

Our intention was to say that if there are significant differences in the SACs but not in the MVMC-SACs, these differences must be due to the private properties in the SACs.

398 - 499   Needs  rewriting.

We have rewritten this section. We hope it is now better.

524-5    why would the fires certainly be lower?  (more valuable land is less likely to burn? – citation needed)

We have made some changes to the paper. We hope we have improved the text.

526-528   this statement appears unsupported

We hope to have answered this with the figure in appendix A, which shows the protected area by municipality and the evolution of the population between 2001 and 2014.

We sincerely thank you for your suggestions, which have contributed to a significant improvement in our work.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments 

1     The paper is actual and brings a new and necessary view of the relationship with the ecological network Natura 2000. It clearly indicates the existence of problems and attitudes towards the primary activity. The structure of the paper is clear and well written. The introduction is clear and indicates what the authors wanted to do and present. The chapter on results is clearly and concisely written, also in the chapter of discussions and conclusions the author is focused on the interpretation of the results. The literature chapter is fine.

Suggestions 

1.      In the chapter Material and methods, line 186 - page 5 - I suggest that Table 1 be transferred to the Appendix for better visibility and clarity.

2.      In the discussion, it would be useful to add what are the criteria (conservation policies rules) that limit primary activities and are related to the Natura 2000 Ecological Network? In addition, it would be good to state what has changed compared to the previous way of management?

Comments for author File: Comments.pdf

Author Response

Dear Sir or Madam,

We would like to thank you for your time and kind contributions concerning this paper. In the next lines we will try to give an answer to your notes and comments:

Comments 

1     The paper is actual and brings a new and necessary view of the relationship with the ecological network Natura 2000. It clearly indicates the existence of problems and attitudes towards the primary activity. The structure of the paper is clear and well written. The introduction is clear and indicates what the authors wanted to do and present. The chapter on results is clearly and concisely written, also in the chapter of discussions and conclusions the author is focused on the interpretation of the results. The literature chapter is fine.

Suggestions 

1. In the chapter Material and methods, line 186 - page 5 - I suggest that Table 1 be transferred to the Appendix for better visibility and clarity.

 Thank you for your advice. We have moved the table 1 to the appendix A.

2. In the discussion, it would be useful to add what are the criteria (conservation policies rules) that limit primary activities and are related to the Natura 2000 Ecological Network? In addition, it would be good to state what has changed compared to the previous way of management?

Thank you very much for your advice. We have made some modifications to the paper which we hope we have explained better.

We sincerely thank you for your suggestions, which have contributed to a significant improvement in our work.

Author Response File: Author Response.docx

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

In this Manuscript, Rodriguez et al. look at how the declaration of protected areas and its intersection with communal private property affected the fire regime and the occurrence of arson. While this is a very interesting question, the authors do a poor job of clearly laying out the knowledge gap and the mechanisms which lead them to their hypothesis. Their presentation of the methods and results is highly disorganised and very difficult to follow and demonstrates a lack of understanding of standardised fire activity metrics. Further the language is verbose and difficult to follow (at one point a run-on sentence lasted an entire paragraph!). I appreciate that english is probably not the first language of the authors, but a higher standard of language is needed for this paper to be understandable. This paper seems like it should be an early draft rather than a version submitted for review to a peer-reviewed journal. I do hope the authors substantially rewrite this paper (and refine some of the analyses), as I’m interested in the results, but in my mind this paper is not of high enough quality to be considered for peer review.

I outline the broad issues with this manuscript below, followed by specific comments.

The introduction is very hard to follow. It does not clearly define the problem or the knowledge gap. It does not discuss in any substantial way the literature around forest management, fire, and land protection (there needs to be more of an explanation of the content of some of the references). It goes into a lot of detail about policy and social opposition, which in my view is not necessary given it seems to be an ecological study. Also the role of intentional fire and the effect of protected areas is not given much focus in the introduction, but it seems that this is the actual focus of the study. 

The Methods are incredibly disorganised and very difficult to follow. The background on Galicia is good, but the subsection “Materials and Methods” (which should be renamed, maybe broken up into a geospatial analysis and statistical analysis subsection) is extremely difficult to understand. You should start with an overview of your research question and methods for data extraction and statistical analysis. Then go into the details with a paragraph for each. I also find it highly problematic that you did not incorporate climate data, I detail why below.

The results section is similarly disorganised and an information dump. I recommend splitting the results into subsections and starting each one with a summary of its findings. The tables are impossible to interpret.  I recommend converting everything to figures and annotating these figures based on statistical significance. The tables showing the t–test results can be supplementary material. Also a lot of the terminology for the response variables is difficult to remember/understand. I would recommend the authors familiarise themselves with more standard terminology used in remote sensing/geospatial fire science (e.g. avg. area burnt per fire or number of fires per hectare), but at the very least they need a table describing what each of the variables means and how it is calculated.

The Discussion covers interesting topics, and is easier to read than the other sections, but given the lack of clarity in the rest of the paper it is difficult to robustly evaluate. 

Specific comments:

• L. 31 This definition of SACs is very important but verbose and hard to follow.

• L. 42 Can you discuss some of the socioecological mechanisms with which the declaration of protected areas may affect ecosystem function?

• L. 49 This intro is missing a paragraph broadly describing the link between fire regimes, fuels, and vegetation management. Given that this is a journal about forests, more background on fire is required.

• L. 50 I think this is a key paragraph, as you are introducing the link between protected areas, forest management , and fire, which is, as far as I can tell, the focus of this study. However this paragraph is extremely verbose and unclear. You need to clearly articulate the mechanisms through which protected areas might affect forest management and in turn wildfire activity.

• L. 73 What does mid-term analysis mean?

• L. 77 Why would this occur, you need to discuss the mechanisms that would lead to an increase in arson on this land.

• L. So is the focus of this paper on how protected areas might reduce arson or how they might change forest management and fire regimes. I’m confused.

• L. 109 This paragraph may be more relevant in the introduction, as it is helping contextualise the research question.

• L. 123 What do you mean by the “Galician Mountain”

• L.124 How do the comuneros manage the land? Do they neglect it, do they manage it for forestry. It sounds like you are trying to describe this in the next sentence but I can't understand. I think this is a really important factor if you are looking at fire regimes.

Figure 2 Can you overlay these? I think that woul be more interesting.

• L. 138/Table 1 You go into an unnecessary amount of detail about these data sets, only discuss the specific data you plan to analyse.

• L. 151 Despite you saying you’ve previously mentioned it, it wasn’t clear to me before this point that you are focusing on intentionally lit fires.

• L. 155 This is not a good enough justification as to why you didn’t include climate data. How do we know that fire activity post 2004 isn’t worse due to worsening fire weather with climate change as opposed to because of the declaration of SACs? This is especially true if you’re looking at something like area burnt. This needs to be addressed in the analysis. I highly recommend incorporating some sort of fire weather index.

Figure 3 is good, but you need to give a better, more concise overview of this process in the text.

• L. 192 Where does the figure caption end and the text begin?

• L. 198 What was you8r sampling unit? Per Year? Per parcel? You need more details.

• L. 199 What do you mean by intentionality?

• L. 203 This paragraph, along with a broad overview of the methodological approach, belongs at the beginning of the Methods.

• L. 212 This information should be incorporated into the previous paragraph.

• L. 216 Start this paragraph with some take home message about your results.

• Table 2 This should be a figure, there is so much information here that it is difficult to tease out what this means.

• Figure 6 I don’t think the maps here are necessary, the figure is too complicated. Also you need some sort of indication of variability on this figure.

• L. 247 This belongs in the Discussion.

• Figure 8 Where are the yellow dots on the figure? They should tell us the comparison between fire rate and

• L. 329 What were these intended positive effects, I don;t think you’ve made that clear.

• L. 338 Yes climate and weather were very important factors that were not accounted for.

Author Response

In this Manuscript, Rodriguez et al. look at how the declaration of protected areas and its intersection with communal private property affected the fire regime and the occurrence of arson. While this is a very interesting question, the authors do a poor job of clearly laying out the knowledge gap and the mechanisms which lead them to their hypothesis. Their presentation of the methods and results is highly disorganised and very difficult to follow and demonstrates a lack of understanding of standardised fire activity metrics. Further the language is verbose and difficult to follow (at one point a run-on sentence lasted an entire paragraph!). I appreciate that english is probably not the first language of the authors, but a higher standard of language is needed for this paper to be understandable. This paper seems like it should be an early draft rather than a version submitted for review to a peer-reviewed journal. I do hope the authors substantially rewrite this paper (and refine some of the analyses), as I’m interested in the results, but in my mind this paper is not of high enough quality to be considered for peer review.

I outline the broad issues with this manuscript below, followed by specific comments.

The introduction is very hard to follow. It does not clearly define the problem or the knowledge gap. It does not discuss in any substantial way the literature around forest management, fire, and land protection (there needs to be more of an explanation of the content of some of the references). It goes into a lot of detail about policy and social opposition, which in my view is not necessary given it seems to be an ecological study. Also the role of intentional fire and the effect of protected areas is not given much focus in the introduction, but it seems that this is the actual focus of the study. 

The Methods are incredibly disorganised and very difficult to follow. The background on Galicia is good, but the subsection “Materials and Methods” (which should be renamed, maybe broken up into a geospatial analysis and statistical analysis subsection) is extremely difficult to understand. You should start with an overview of your research question and methods for data extraction and statistical analysis. Then go into the details with a paragraph for each. I also find it highly problematic that you did not incorporate climate data, I detail why below.

The results section is similarly disorganised and an information dump. I recommend splitting the results into subsections and starting each one with a summary of its findings. The tables are impossible to interpret.  I recommend converting everything to figures and annotating these figures based on statistical significance. The tables showing the t–test results can be supplementary material. Also a lot of the terminology for the response variables is difficult to remember/understand. I would recommend the authors familiarise themselves with more standard terminology used in remote sensing/geospatial fire science (e.g. avg. area burnt per fire or number of fires per hectare), but at the very least they need a table describing what each of the variables means and how it is calculated.

The Discussion covers interesting topics, and is easier to read than the other sections, but given the lack of clarity in the rest of the paper it is difficult to robustly evaluate. 

Specific comments:

• 31 This definition of SACs is very important but verbose and hard to follow.

 You are right. We have amended the text. We hope it is better now. Thank you very much.

• 42 Can you discuss some of the socioecological mechanisms with which the declaration of protected areas may affect ecosystem function?

Excuse us, but we did not want to talk about socio-ecological mechanisms. We simply wanted to check whether the fire regime in the SAC was different from that observed in other territorial units.

• 49 This intro is missing a paragraph broadly describing the link between fire regimes, fuels, and vegetation management. Given that this is a journal about forests, more background on fire is required.

Thank you for your comment. We have improved the paper following your advice.

• 50 I think this is a key paragraph, as you are introducing the link between protected areas, forest management , and fire, which is, as far as I can tell, the focus of this study. However this paragraph is extremely verbose and unclear. You need to clearly articulate the mechanisms through which protected areas might affect forest management and in turn wildfire activity.

We have modified the text. We hope it is now better.

• 73 What does mid-term analysis mean?

 We have removed the term, as it could be misleading.

• 77 Why would this occur, you need to discuss the mechanisms that would lead to an increase in arson on this land.

Due to the social opposition to the Natura 2000 network in several areas, it is likely that the number of fires in areas such as Galicia, where the percentage of arson is very high, will increase after classification as a SAC. 

• So is the focus of this paper on how protected areas might reduce arson or how they might change forest management and fire regimes. I’m confused.

 We hope the article is clearer with the modifications we have made.

• 109 This paragraph may be more relevant in the introduction, as it is helping contextualise the research question.

We consider your comment. In any case, we believe that it is more interesting to keep this section in the study area because it is an important type of property in this region and that does not exist in others.

• 123 What do you mean by the “Galician Mountain”

It was a mistake. We have corrected it. Thanks.

• 124 How do the comuneros manage the land? Do they neglect it, do they manage it for forestry. It sounds like you are trying to describe this in the next sentence but I can't understand. I think this is a really important factor if you are looking at fire regimes.

We have included in the paper a couple of facts that make the management of MCVMs difficult: the fact that only 30 % of the MCVMs have a management project, and that the percentage of the area with certified management is not higher than 7.5 %. These factors, among others, hinder the management of MCVMs and can lead to the emergence of various conflicts between community members.

• Figure 2 Can you overlay these? I think that woul be more interesting.

Please excuse us, but a reviewer asked us to do so. Initially we had done it as you indicated.

• 138/Table 1 You go into an unnecessary amount of detail about these data sets, only discuss the specific data you plan to analyse.

We have modified the text. We hope it is now better.

• 151 Despite you saying you’ve previously mentioned it, it wasn’t clear to me before this point that you are focusing on intentionally lit fires.

We have modified the text. We hope it is now better.

• 155 This is not a good enough justification as to why you didn’t include climate data. How do we know that fire activity post 2004 isn’t worse due to worsening fire weather with climate change as opposed to because of the declaration of SACs? This is especially true if you’re looking at something like area burnt. This needs to be addressed in the analysis. I highly recommend incorporating some sort of fire weather index.

In Galicia, most forest fires are human-caused fires (arson fires), so climatic factors usually only affect the behaviour of the fire. In any case, we have included some climatic data. 

• Figure 3 is good, but you need to give a better, more concise overview of this process in the text.

We have modified the text. We hope it is now better.

• 192 Where does the figure caption end and the text begin?

Thank you. We have corrected it.

• 198 What was you8r sampling unit? Per Year? Per parcel? You need more details.

 I don't know if we understand you correctly. We are not sampling. We have taken into account all intentional forest fires that occurred between 1999 and 2014.

• 199 What do you mean by intentionality?

Yes, thank you, that was a mistake. We are actually referring to the percentage of forest fires caused by man. We have corrected it.

• 203 This paragraph, along with a broad overview of the methodological approach, belongs at the beginning of the Methods.

We do not know if you are referring to the caption of the image, as the lines do not match. If so, we thought it was a good idea to explain it that way, for clarity.

• 212 This information should be incorporated into the previous paragraph.

 Yes. You are right. We have corrected it. Thank you.

• 216 Start this paragraph with some take home message about your results.

 We have included a short introductory paragraph.

• Table 2 This should be a figure, there is so much information here that it is difficult to tease out what this means.

 We have replaced that table with Figure 7. We hope that it is now more understandable.

• Figure 6 I don’t think the maps here are necessary, the figure is too complicated. Also you need some sort of indication of variability on this figure.

 We have removed the maps.

• 247 This belongs in the Discussion.

We have included this text in the discussion. Thank you very much for your help.

• Figure 8 Where are the yellow dots on the figure? They should tell us the comparison between fire rate and

 We have modified the graph. We had made a mistake. Thank you very much.

• 329 What were these intended positive effects, I don;t think you’ve made that clear.

We understand that if there are still arson fires in the Natura 2000 network, then the desired objectives have not been achieved. That is what we mean. 

• 338 Yes climate and weather were very important factors that were not accounted for.

We have carried out a small climatic study of the region.

Author Response File: Author Response.docx

Reviewer 2 Report

This paper sought to evaluate whether wildfire occurrences in Natura 2000 Network’s Special Areas of Conservation in Galicia, Spain, had changed since the declaration of those reserves since 2004. Statistical tests of counts of intentional wildfires and area burned by intentional wildfires, based on a small number of annual average observations before and after the declaration, show conflicting results. The authors take the conflicting results as evidence that the Natura 2000 Network’s SACs have not succeeded in reducing wildfire. I have several major comments and some specific comments about parts of the text.

Major Comments

1. It is not clear whether wildfire occurrence is a performance measure for SACs, or how wildfires impact performance measures. The text does provide a brief overview of the goals of the Natura 2000 Network, namely habitat and species protection. But the authors do not link habitat integrity or species population changes to wildfires in this study, so it is not clear that changes in wildfire are relevant or have threatened protection goals. Is there evidence that can be offered about the links between wildfire and habitat and species protection in Galicia?

2. The statistical results are based on annual averages across the Galicia-wide aggregates of MVMC, SAC, MVMC-SAC, and MVMC-Non-SAC. Is it possible that disaggregating in time (to months, for example) could yield different findings?

3. Related to 2, although climate variables could be the same within years Galicia-wide, what is less likely to be the same are the biophysical, human population, and wildfire management variables associated with SAC and non-SAC areas of Galicia. Without controlling for those factors, it could be that averages reported and compared between the various MVMC and SAC units do not represent the same populations. I would suggest that a higher level statistical analysis, a regression- based approach that controls for the biophysical, human, and wildfire management factors, could yield more definitive results on wildfire activity.

4. It could be argued that area burned, which did decline on SAC land units at a greater rate than non-SAC units, is a better measure of wildfire changes that would be relevant to Natura 2000 goals. In the limit, one large 10 000-hectare wildfire would likely be far more damaging to habitats and imperil species to a greater degree than 1000 1-hectare wildfires. Yet the paper seems to consider the two measures (count changes, area burned changes) as being equivalent in importance. What is the basis for treating them equivalently?
5. Specific Comments

• In the abstract, the authors use two abbreviations to represent Special Areas of Conservation: SAC and ZEC. The text needs to be consistent on how to refer to these land units. (Figure 8 also uses “ZEC”.)
• Was normality of the statistical distributions evaluated? The authors could try at least a logarithmic transformation of the variables (counts, area burned, and their rates per unit area) and see if the differences across land types are robust to this transformation.
• Lines 122-123: what is the “Galician mountain”? do you mean Galician wildland areas, or Galician rural areas?
• Line 195: what is MVMX?
• Line 221: is it true that arson-ignited wildfires represent only 9.95% of the total? That figure does not align with the values shown in Table 3.
• Table 2: what is “Med” in the second and third to last rows? Median? Average? In the column headers, what are “Inc” and “Inc Int.”? Do you really mean “count” and “count of intentional”?
• In Figure 6: (i) The legends would be more readable if the words were spelled out. (ii) What are the numbers on the outsides of the various maps of Galicia in this figure? They are not referred to in the text or in the figure caption or the legends, so they seem to be a distraction.
• Figure 8 needs to replace “ZEC” with “SAC”.
• Lines 411-419: The text here ventures into opinion that is not appropriate for an academic paper, including statements unsupported by the research that was carried out in this study.
• Line 432: The inclusion of the word “Germanic” seems to be unnecessary.
• Lines 438-446: The text offers opinions not supported by the research that was carried out for this study.

Author Response

This paper sought to evaluate whether wildfire occurrences in Natura 2000 Network’s Special Areas of Conservation in Galicia, Spain, had changed since the declaration of those reserves since 2004. Statistical tests of counts of intentional wildfires and area burned by intentional wildfires, based on a small number of annual average observations before and after the declaration, show conflicting results. The authors take the conflicting results as evidence that the Natura 2000 Network’s SACs have not succeeded in reducing wildfire. I have several major comments and some specific comments about parts of the text.

Major Comments

1. It is not clear whether wildfire occurrence is a performance measure for SACs, or how wildfires impact performance measures. The text does provide a brief overview of the goals of the Natura 2000 Network, namely habitat and species protection. But the authors do not link habitat integrity or species population changes to wildfires in this study, so it is not clear that changes in wildfire are relevant or have threatened protection goals. Is there evidence that can be offered about the links between wildfire and habitat and species protection in Galicia?

Natural fires in Galicia are very infrequent. In fact, most forest fires are arson. Taking this into account, it could be concluded that an increase in the number of intentional forest fires in SACs is due to a rejection of the Natura 2000 Network. There are of course other possible factors, such as neighbourhood conflicts, for example.

2. The statistical results are based on annual averages across the Galicia-wide aggregates of MVMC, SAC, MVMC-SAC, and MVMC-Non-SAC. Is it possible that disaggregating in time (to months, for example) could yield different findings?

In Galicia, fires occur mainly between February and April and during the summer months. However, this is only due to the state of the vegetation. In both cases they are arson attacks. For this reason, we believe that we should make an effort to analyse social conflicts, and that disaggregating the data temporally would not provide very relevant information.

3. Related to 2, although climate variables could be the same within years Galicia-wide, what is less likely to be the same are the biophysical, human population, and wildfire management variables associated with SAC and non-SAC areas of Galicia. Without controlling for those factors, it could be that averages reported and compared between the various MVMC and SAC units do not represent the same populations. I would suggest that a higher level statistical analysis, a regression- based approach that controls for the biophysical, human, and wildfire management factors, could yield more definitive results on wildfire activity.

Thank you for your help, but we did not want to include additional factors in our analysis. We would like to emphasis the fact that most of the forest fires in Galicia are intentional fires. We therefore believe that we should focus on social conflicts.

4. It could be argued that area burned, which did decline on SAC land units at a greater rate than non-SAC units, is a better measure of wildfire changes that would be relevant to Natura 2000 goals. In the limit, one large 10 000-hectare wildfire would likely be far more damaging to habitats and imperil species to a greater degree than 1000 1-hectare wildfires. Yet the paper seems to consider the two measures (count changes, area burned changes) as being equivalent in importance. What is the basis for treating them equivalently?

Natural fires are very infrequent in Galicia. Almost all of them are arson fires. Therefore, the occurrence of wildfires is incompatible with the Natura 2000 Network.

5. Specific Comments

• In the abstract, the authors use two abbreviations to represent Special Areas of Conservation: SAC and ZEC. The text needs to be consistent on how to refer to these land units. (Figure 8 also uses “ZEC”.)

We have already corrected it. Thank you very much.

• Was normality of the statistical distributions evaluated? The authors could try at least a logarithmic transformation of the variables (counts, area burned, and their rates per unit area) and see if the differences across land types are robust to this transformation.

We value your proposal. We consider that in our case, due to the type of questions we wanted to answer (differences between groups) and the type of variables we were using (area and percentage), we believe that changes in scale would not entail changes in the results.

• Lines 122-123: what is the “Galician mountain”? do you mean Galician wildland areas, or Galician rural areas?

Yes, sorry. “Galician rural áreas”

• Line 195: what is MVMX?

We have already corrected it. Thank you very much for your feedback.

• Line 221: is it true that arson-ignited wildfires represent only 9.95% of the total? That figure does not align with the values shown in Table 3.

We refer to the fact that 9.95% of forest fires caused by man occur in SACs. We have amended the text. It was probably not very well explained. Thank you very much for your help.

• Table 2: what is “Med” in the second and third to last rows? Median? Average? In the column headers, what are “Inc” and “Inc Int.”? Do you really mean “count” and “count of intentional”?

We have replaced that table with Figure 7. We hope that it is now more understandable.

• In Figure 6: (i) The legends would be more readable if the words were spelled out. (ii) What are the numbers on the outsides of the various maps of Galicia in this figure? They are not referred to in the text or in the figure caption or the legends, so they seem to be a distraction.

The numbers on the outside of the maps are the geographical coordinates. The reference system is mentioned on the maps above. However, for clarity we have made an allusion to it in the image caption.

• Figure 8 needs to replace “ZEC” with “SAC”.

We have already corrected it. Thank you very much for your feedback.

• Lines 411-419: The text here ventures into opinion that is not appropriate for an academic paper, including statements unsupported by the research that was carried out in this study.

We have completed the text, in order to explain it better.

• Line 432: The inclusion of the word “Germanic” seems to be unnecessary.

We have already corrected it. Thank you very much.

• Lines 438-446: The text offers opinions not supported by the research that was carried out for this study.

Along these lines, we agree with other authors who have argued that the Natura 2000 network must take into account local communities and respect their traditional ways of using the land. This is particularly important in Galicia, in our opinion, as in this region more than 90% of forest fires are man-made fires.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

As before, the research presented here sought to evaluate whether the designation of special areas of conservation under Natura 2000 had an impact on wildfire in Galicia, specifically the area burned and the numbers of intentionally ignited wildfires. The authors undertook text revisions, in part in response to my comments (and likely partly in response to another reviewer’s comments). Revision efforts were focused on replacing terms (‘ZEC’ with ‘SAC’, ‘intentional fire’ with ‘arson’) and refining presentation of results and conclusions.

Nevertheless, the authors did not do an adequate job in responding to the primary concerns: why or whether wildfire occurrence should be an important performance measure in Natura 2000 land units in Spain, whether their statistical results would be different if data were evaluated at higher frequency (e.g., monthly rather than annual data), and why area burned and number of fires should be considered equally valid measures of wildfire impact. Although the authors acknowledged in the text the potential importance of confounding variables when carrying out their study (e.g., not accounting for hazardous fuels levels, which might differ between SAC and non-SAC lands), they did not undertake any scientific analyses to quantify the potential effects of such confounders.

My specific comments were partly addressed, but in addressing especially the problems of terminology inconsistency and opacity, authors were not completely successful: one example is that “ZEC” (instead of “SAC”) still appears in several figures of the revised manuscript. Furthermore, in the revision process, the authors introduced new text inconsistencies; one example is that term “intentional” was partially but not completely replaced with “arson” in many places in the text, figures and tables and their respective captions and titles.