Do Salamanders Limit the Abundance of Groundwater Invertebrates in Subterranean Habitats?

Round 1

Reviewer 1 Report

Dear Authors,

I found your manuscript a fine contribution to the field of speleobiology, addressing a question which has rarely (if ever) been addressed. I applaud to your extensive and long year field work and for assembling such a big data set. In my opinion the manuscript is generally fine, though I have some suggestions for improvements. I write my major suggestions here below, and have marked all other minor ones as comments in the attached pdf of you manuscript.

1. I suggest changing the title. The term "subterranean superficial habitats" is misleading in your case. Firstly it is an obsolete term, it is now generally refered to as "shallow subterranean habitat" (SSH); see this book:
   Culver, D.C., Pipan, T., 2014. Shallow Subterranean Habitats. Oxford University Press, New York, USA. Also, SSH is generally considered not deeper than 10m from the surface. Judging from Figure 1d, you also sampled much deeper within caves, so beyond the SSH. I suggest to use ecotone, transitional habitat or just entrance sections instead of SSH. Please see my suggestions for alternative titles in the pdf.
2. I think it would be useful if you can add a list of sampled subterranean objects/caves with details such as coordinates, type of habitat, how many pools/site were checked, year, season (date) and, etc. Such a table can go in the supplementary materials. These are crucial info for the repeatability of the study, but are now missing.
3. Data analyses. A) Please consider including time as a covariate or as a random effect in your models. Irregular sampling across such a long time period (2012-2019) might have induced additional variation in your data. Abundances could vary between years and seasons alone. It seems you did not consider this in your models. B) You should report on the results of the random effect of "cave id" in your Table 1. C) Please consider changing the boxplots in Figure 2a-c according to my suggestions written in the pdf. 
4. Many Niphargus species, including those appearing in entrance sections to caves are not just detritivores (shreaders) but also predators of other invertebrates. I argue for this more in detail in the pdf. Please rewrite parts related to Niphargus feeding habits and its trophic role and your result interpretation with this info in mind.
5. Discussion of results for Monolistra is not sufficient. You should state your interpretation of the significant effect of water depth (also for Niphargus) and surface area. To me both variables together describe microhabitat size/volume. It is logical that the greater the habitat that was inspected, the more individuals were found. Please add this discussion.
6. Finally, I think the authors should provide alternative explanations for the observed pattern of animal distribution. Predation of salamander larvae is not the only mechanism that could have produced such a pattern. If you think this is not true, state your arguments; how your sampling design eliminates other explanations for the observed pattern. Please add a paragraph to the discussion explaining this. This seems the most important shortcomming of the manuscript to me.

Throughout the text there is quite a few language, spelling and formatting mistakes. I corrected all that I have observed. I am not a native speaker though. There are also a few references which can be replaced with more recent ones, see the pdf.

I hope you find my suggestions useful and that they will help you improve the manuscript.

All the best and good luck in your future research activities

Comments for author File: Comments.pdf

Author Response

Dear Authors,

I found your manuscript a fine contribution to the field of speleobiology, addressing a question which has rarely (if ever) been addressed. I applaud to your extensive and long year field work and for assembling such a big data set. In my opinion the manuscript is generally fine, though I have some suggestions for improvements. I write my major suggestions here below, and have marked all other minor ones as comments in the attached pdf of you manuscript.

We thank the reviewer for the positive comments.

I suggest changing the title. The term "subterranean superficial habitats" is misleading in your case. Firstly it is an obsolete term, it is now generally refered to as "shallow subterranean habitat" (SSH); see this book:

Culver, D.C., Pipan, T., 2014. Shallow Subterranean Habitats. Oxford University Press, New York, USA. Also, SSH is generally considered not deeper than 10m from the surface. Judging from Figure 1d, you also sampled much deeper within caves, so beyond the SSH. I suggest to use ecotone, transitional habitat or just entrance sections instead of SSH. Please see my suggestions for alternative titles in the pdf.

We thank the reviewer; we tried to change the title according to his suggestion and to the suggestion of reviewer 2. How to change the title was highly debated among us. As the reviewer correctly point out we sampled also deep sectors. At the end we chose to speak more generally about subterranean habitats; we go form the entrance (0 meters) to 370 m from the entrance. All the sites that we sampled can be referred as a kind of subterranean environment (border, shallow, transitional, and deep).

I think it would be useful if you can add a list of sampled subterranean objects/caves with details such as coordinates, type of habitat, how many pools/site were checked, year, season (date) and, etc. Such a table can go in the supplementary materials. These are crucial info for the repeatability of the study, but are now missing.

Following the suggestion, we provided a supplementary table with all the requested information and the data used in the analyses. In the previous version we made confusion between sampling occasions and sites (pools, sectors of streams) surveyed. We performed 386 visual samplings, surveying 117 sites (pools and distinct subterranean stream sectors) within 17 natural caves and 24 artificial cavities. All the data collected at each site in each sampling is provided in the Supplementary Table 1.

Data analyses. A) Please consider including time as a covariate or as a random effect in your models. Irregular sampling across such a long time period (2012-2019) might have induced additional variation in your data. Abundances could vary between years and seasons alone. It seems you did not consider this in your models. B) You should report on the results of the random effect of "cave id" in your Table 1. C) Please consider changing the boxplots in Figure 2a-c according to my suggestions written in the pdf.

We changed the analyses and the section “Statsitcal analyses” following these suggestions and the suggestions of Editor and reviewer 2, so some parts are a bit different compared to the modifications suggested here and in the pdf attached, but always conceptually consistent to solve the issues raised in the whole revision process. Random effects variance and SD were always far < 0.01.

Many Niphargus species, including those appearing in entrance sections to caves are not just detritivores (shreaders) but also predators of other invertebrates. I argue for this more in detail in the pdf. Please rewrite parts related to Niphargus feeding habits and its trophic role and your result interpretation with this info in mind.

We followed the suggestion and we changed the text here and also in the discussion.

Discussion of results for Monolistra is not sufficient. You should state your interpretation of the significant effect of water depth (also for Niphargus) and surface area. To me both variables together describe microhabitat size/volume. It is logical that the greater the habitat that was inspected, the more individuals were found. Please add this discussion.

We added the suggested consideration to discussion and we improved the interpretation of the results not strictly linked with the salamanders. The use of the area as covariate was intended to take into account the variation linked to the sampling effort. The water depth is not so strongly related with the sampling effort but also with direct environmental factors occurring in the sites (water flow, debris

Finally, I think the authors should provide alternative explanations for the observed pattern of animal distribution. Predation of salamander larvae is not the only mechanism that could have produced such a pattern. If you think this is not true, state your arguments; how your sampling design eliminates other explanations for the observed pattern. Please add a paragraph to the discussion explaining this. This seems the most important shortcomming of the manuscript to me.

Following this suggestion, and incorporating the suggestions of reviewer 2 we provide the alternative explanations that can be linked to the observed patterns.

Throughout the text there is quite a few language, spelling and formatting mistakes. I corrected all that I have observed. I am not a native speaker though. There are also a few references which can be replaced with more recent ones, see the pdf.

We are really grateful to the reviewer for all the precious comments; we modified the text accordingly. Some parts have been modified also according to the comments of the other reviewers. We did not change the name of Monolistra coeca in M. caeca as at least the species that we sampled is called effectively “coeca”, even if it is strange; we checked the literature regarding it.

I hope you find my suggestions useful and that they will help you improve the manuscript.

All the best and good luck in your future research activities

The suggestions were quite useful and allowed a significant improvement of the manuscript.

Reviewer 2 Report

Review of the manuscript:

Title: Are salamanders predator of cave-adapted 2 invertebrates in subterranean superficial habitats? 

Authors: Raoul Manenti, Enrico Lunghi, Benedetta Barzaghi, Andrea Melotto, Mattia Falaschi, Gentile Francesco Ficetola and 6 

This paper focuses on salamanders as crucial predators of hypogean invertebrates in various groundwater habitats in Italy. As underlined by the authors this study constitutes the first one investigating the impact of facultative cave-breeding salamanders on groundwater fauna. Although I consider that a such study is needed in groundwater ecology, and I appreciate the extensive fieldwork effort, the manuscript is highly speculative. In my opinion, in this form, based on the proposed methodology, the manuscript has an improper rationale.

Overall, the English is good, but there are places where the phrasing could be improved. I would suggest to English proof-read the manuscript.

Please find bellow comments and suggestions that in my opinion might help in your future endeavours related to this manuscript. 

Title

In the context of no direct evidence proving that salamanders pray on the sampled aquatic invertebrates, the title is misleading. The title should have a clear, precise scientific meaning. Re-writing the title by ensuring that it is informative and appropriate would benefit the carried research.

Abstract

Lines 23-24: The authors claim they sampled 369 microhabitats. I think they meant sampling sites; it is doubtful that there are 369 groundwater microhabitats. I think it is important to be precise regarding the types of microhabitats you sampled (i.e. interstitial, pools, the stream benthos…). 

Lines 28-31: The conclusion is speculative. The used methodology does not provide direct assessments that may support such conclusion.

Keywords

Selected keywords should define the carried research. This research does not involve an assessment of the subterranean trophic structure, so terms as cascade, top-down should be avoided. Also detritus is mentioned without being specifically assessed. No original information is presented dealing with the aforementioned keywords, so I would suggest their removal. 

Troglobiont: this study does not deal with terrestrial specialised cave fauna. I would suggest its removal.

Stygofauna: as emphasised by authors this term refers to fauna adapted to live in groundwater, thus restricted to groundwater. However, given that species of the analysed genera are not indicated and some of the species belonging to these genera might also be stygophilous, I would suggest to use the term hypogean. This term is widely accepted as ecological category based on species occurrence in groundwater and not based on species adaptations that sometimes are difficult to assess for aquatic invertebrate fauna. NB: Depigmentation and blindness are not characteristic for cave dwelling species only. 

Introduction

Lines 58-59: Please be careful when describing the subterranean environment. The environmental variables are not limited, but the variation of underground microclimate is weak compared to surface environments. 

Materials and Methods

2.1. The sampling design is not described in a way that ensures it’s repeatability. 

• The manuscript does not provide a map showing the sampling sites. Moreover, a proper description of the sampling area is not given. It should be explained why it’s important to monitor this area.
• The types of sampled habitats are not listed and described in enough detail.
• How the sampling has been performed? Line 108: you mentioned repeated sampling and in line 120 you mentioned visual encounter surveys. If you sampled the invertebrate fauna, how did you do it? The permits obtained for sampling should be mentioned. If you didn’t sample the aquatic fauna: how did you manage to count it, given that some of them are highly mobile and small?
• It is not clear when the sampling has been performed (i.e in what season/s the samples were gathered/observed; all sampling sites were sampled every year?; How many invertebrate samples where analysed?; How many samples/microhabitat were considered?; how many samples/site were considered?). I know the answers to those questions might be highly variable, but it’s important to take into account this variability when choosing the most appropriate statistical approach for data exploration and analysis. 
• Lines 142-144: in aquatic ecology, at least basic physico-chemical parameters should be considered (i.e. water temperature; electrical conductivity, pH) to be related with species presence/absence. Many studies dealing with groundwater fauna emphasise on the important role of these parameters, correlated with surface environmental features (i.e. rainfall, temperature, …) in shaping species distribution. 

2.2. Statistical analysis. For data exploration, authors use boxplots. I think for data exploration they should avoid data transformation and further data transformation, if needed, should be justified based on the exploratory analysis. 

GLLMs: Nothing is mentioned on the pre-work required before running the multiple regression (i.e. checking the relationship between each independent variable and the dependent variable + the relationship among dependent variables using scatterplots, correlations/simple regressions). The risk of potential multicollinearity is not addressed (i.e. VIF). Further, I would suggest to use the likelihood ratio tests for negative binomial models to assess the contribution of each variable to the model. 

Results

The entire dataset should be provided: sampling site, sampling date, sampling habitat, number of sampled points within the same sampling habitat, identified taxa, no of individuals … This would be useful in understanding how the sampling was conducted and how the statistical analysis were adapted to the dataset.

The statistical approach does not allow to compare results among microhabitats with different pressures, among years...

Discussion

The main idea is that fire salamanders are predators of groundwater invertebrates. This statement is speculative, given that stomach content analysis has not been performed. 

There are several other reasons that might explain the dynamics of groundwater invertebrates that should be also discussed as potential factors leading to low abundances:

• some groundwater invertebrates are able to roam the interstices in sediments or the fissures in limestone, so their absence might be also related to their ability to avoid/escape potential adverse pressures;
• Invertebrate interspecies and intraspecies interactions. For instance in niphargids canibalism has been documented;
• environmental pressures. 

With my best regards.

Author Response

This paper focuses on salamanders as crucial predators of hypogean invertebrates in various groundwater habitats in Italy. As underlined by the authors this study constitutes the first one investigating the impact of facultative cave-breeding salamanders on groundwater fauna. Although I consider that a such study is needed in groundwater ecology, and I appreciate the extensive fieldwork effort, the manuscript is highly speculative. In my opinion, in this form, based on the proposed methodology, the manuscript has an improper rationale.

We tried to avoid speculations and we used also suggestions of the other reviewers to improve the rationale of the study.

Overall, the English is good, but there are places where the phrasing could be improved. I would suggest to English proof-read the manuscript.

We improved the English spelling in the manuscript following the suggestions of all the reviewers and by an attentive revision of the manuscript.

Please find bellow comments and suggestions that in my opinion might help in your future endeavours related to this manuscript.

We thank the reviewer for the suggestions; when possible they were all used to increase the manuscript quality.

Title

In the context of no direct evidence proving that salamanders pray on the sampled aquatic invertebrates, the title is misleading. The title should have a clear, precise scientific meaning. Re-writing the title by ensuring that it is informative and appropriate would benefit the carried research.

We used this suggestion and that of reviewer 1 to make the title more adherent to the results and more informative.

Abstract

Lines 23-24: The authors claim they sampled 369 microhabitats. I think they meant sampling sites; it is doubtful that there are 369 groundwater microhabitats. I think it is important to be precise regarding the types of microhabitats you sampled (i.e. interstitial, pools, the stream benthos…).

We changed the sentence and we clarified the dataset that consists of 386 sampling occasions performed on 117 sites belonging to 41 distinct caves.

Lines 28-31: The conclusion is speculative. The used methodology does not provide direct assessments that may support such conclusion.

We changed the sentence in order to be more adherent to the negative relationship that we detected between fire salamander occurrence and the abundance of stygofauna.

Keywords

Selected keywords should define the carried research. This research does not involve an assessment of the subterranean trophic structure, so terms as cascade, top-down should be avoided. Also detritus is mentioned without being specifically assessed. No original information is presented dealing with the aforementioned keywords, so I would suggest their removal.

Troglobiont: this study does not deal with terrestrial specialised cave fauna. I would suggest its removal.

Stygofauna: as emphasised by authors this term refers to fauna adapted to live in groundwater, thus restricted to groundwater. However, given that species of the analysed genera are not indicated and some of the species belonging to these genera might also be stygophilous, I would suggest to use the term hypogean. This term is widely accepted as ecological category based on species occurrence in groundwater and not based on species adaptations that sometimes are difficult to assess for aquatic invertebrate fauna. NB: Depigmentation and blindness are not characteristic for cave dwelling species only.

Following the suggestions, we changed the keywords, and we added hypogean.

Introduction

Lines 58-59: Please be careful when describing the subterranean environment. The environmental variables are not limited, but the variation of underground microclimate is weak compared to surface environments.

Following the suggestion, we changed the sentence considering that subterranean environments, compared to surface one can be considered quite more simple (Poulson & White, 1969 – Science).

Materials and Methods

2.1. The sampling design is not described in a way that ensures it’s repeatability.

We provided the full dataset and we better described the sampling design

The manuscript does not provide a map showing the sampling sites. Moreover, a proper description of the sampling area is not given. It should be explained why it’s important to monitor this area.

We included the complete dataset with the coordinates of all the sites. A map would be not so informative we guess as basically will cover all the North of Italy. However, if the Editors thinks that it is appropriate we can add this. We expressly did not emphasize the importance of the area as is not more or less important than every other areas of the world where surface salamanders enter subterranean habitats and interact with groundwater fauna.

The types of sampled habitats are not listed and described in enough detail.

How the sampling has been performed? Line 108: you mentioned repeated sampling and in line 120 you mentioned visual encounter surveys. If you sampled the invertebrate fauna, how did you do it? The permits obtained for sampling should be mentioned. If you didn’t sample the aquatic fauna: how did you manage to count it, given that some of them are highly mobile and small?

For the analyses showed in this study we only used data from visual encounter surveys. We better stated this in the text. Visual Encounter Surveys are simple surveys performed by just watching the ACTIVE organisms that occur in a certain site. This technique does not involve animals collecting and manipulation and thus does not need to be authorised. Moreover, cave aquatic invertebrate fauna is not protected in Italy except for some regions; in our case for example for some caves of Friuli even if we would we couldn’t provide data of net samplings as we did not do them. Visual surveys were performed by lightening an electric torch over the site; consider that studies performed on fire salamander larvae comparing the number of individuals active and the number of individuals sheltered showed that in caves most of the larvae do not shelter as they are already in complete darkness. About the three genera of invertebrates considered in our study, after saying that someone could underline that lightening them can be more disturbing that netting and avoiding this should be insert in laws for their conservation once that they will be established, is important to note that:

• they are not so much mobile, especially when lightened, except Niphargus that move a lot and makes itself more visible
• they are not so small
• they are easily visible and countable in small sites with not so much microhabitat differentiation as a cave pool. It is just matter to look for them. White, depigmented aquatic organisms stand out on darker substrates when we look for them, and can be easily counted. Consider that even speleologists once trained to look for cave planarians, started to find them easily in places where they passed through for many years without noticing nothing.

Moreover, these invertebrates can easily use interstices and small shelters that should be considered as a different habitat (we improved discussion on this point considering the comments of reviewer 1), while we wanted to assess the number of active individuals in habitats with or without fire salamanders larvae but comparable to each other and in theory exploitable by the larvae. We clarified a bit the text following the suggestion. Finally, the statistical approach if abundance analysis is recommended for contexts in which detectability is low or unknown (Barket et al., 2017, see References).

It is not clear when the sampling has been performed (i.e in what season/s the samples were gathered/observed; all sampling sites were sampled every year?; How many invertebrate samples where analysed?; How many samples/microhabitat were considered?; how many samples/site were considered?). I know the answers to those questions might be highly variable, but it’s important to take into account this variability when choosing the most appropriate statistical approach for data exploration and analysis.

We agree with the reviewer that more variability should be considered in the analyses as also pointed out by the Editor in the specific comments. We underline that the sampling was not thought for the analysis showed, but we are reusing/using a number of data to put a new stone of knowledge on the understanding of the role of epigean salamanders for cave-adapted invertebrates thanks to the occasion of this interesting Special Issue. We included the time and the number of visits that each site in the analyses. We proved a detailed dataset with all the data and we better clarified the text.

Lines 142-144: in aquatic ecology, at least basic physico-chemical parameters should be considered (i.e. water temperature; electrical conductivity, pH) to be related with species presence/absence. Many studies dealing with groundwater fauna emphasise on the important role of these parameters, correlated with surface environmental features (i.e. rainfall, temperature, …) in shaping species distribution.

We agree with the reviewer; unfortunately, we do not have such parameters for all the samplings/sites for the reason stated above. Having these parameters will be useful in future studies to understand better that patterns of activity and distribution of the cave invertebrates considered, even if the data available for from other studies in the same karst areas  that we sampled show scarce subterranean variability and scarce heterogeneity (Pezzoli, 1996 e 2000). However, we are confident that including chemical parameters, will not change the relationship observed between the occurrence of salamanders and invertebrates.

2.2. Statistical analysis. For data exploration, authors use boxplots. I think for data exploration they should avoid data transformation and further data transformation, if needed, should be justified based on the exploratory analysis.

We thank the reviewer for the suggestion. We provide now the graphs without data transformation.

GLLMs: Nothing is mentioned on the pre-work required before running the multiple regression (i.e. checking the relationship between each independent variable and the dependent variable + the relationship among dependent variables using scatterplots, correlations/simple regressions). The risk of potential multicollinearity is not addressed (i.e. VIF). Further, I would suggest to use the likelihood ratio tests for negative binomial models to assess the contribution of each variable to the model.

Following the suggestion, we tested multicollinearity through VIF analysis on a simple linear model with Poisson distribution including all the variables. For each variable VIF was strongly below 5 and not even reach 2.  We added a likelihood ratio test to assess variables significance. We also briefly state the exploratory analyses made on the variables prior to the multiple regression.

Results

The entire dataset should be provided: sampling site, sampling date, sampling habitat, number of sampled points within the same sampling habitat, identified taxa, no of individuals … This would be useful in understanding how the sampling was conducted and how the statistical analysis were adapted to the dataset.

Following the suggestion, we provided the entire dataset.

The statistical approach does not allow to compare results among microhabitats with different pressures, among years...

We could also have had a different approach in the use of the data available, but we chose uniformity among fire salamander larvae occurrence/absence to respond to a specific question related to salamanders. We agree with the reviewer that it could be interesting to assess the different pressures played by different microhabitats in stygobiont fauna and we hope to perform such study with specific investigations.

Discussion

The main idea is that fire salamanders are predators of groundwater invertebrates. This statement is speculative, given that stomach content analysis has not been performed.

Following the suggestion, we modified the text.

There are several other reasons that might explain the dynamics of groundwater invertebrates that should be also discussed as potential factors leading to low abundances: some groundwater invertebrates are able to roam the interstices in sediments or the fissures in limestone, so their absence might be also related to their ability to avoid/escape potential adverse pressures; Invertebrate interspecies and intraspecies interactions. For instance in niphargids canibalism has been documented;

environmental pressures.

With my best regards.

Following also the suggestions of reviewer 1, we integrated these explanations in the discussion. We thank the reviewer for the constructive comments.

Reviewer 3 Report

Dear authors,

Overall, this is a clear, concise, and well-written manuscript. The introduction is relevant and theory-based. The methods are generally appropriate, although clarification of a few details you need to add. Overall, the results are clear and compelling with few exceptions which have been added to the corrections. Also, the use of the word "underground" has more than a single interpretation. You make a systematic contribution to the research literature in this area of investigation. The longitudinal nature of the work is a welcome addition. Overall, this is a high-quality manuscript that has implications for the theoretical basis and research development in cave ecology.

Specific comments and corrections were incorporated in the pdf form of the MS.

Best wishes and regards.

Comments for author File: Comments.pdf

Author Response

Dear authors,

Overall, this is a clear, concise, and well-written manuscript. The introduction is relevant and theory-based. The methods are generally appropriate, although clarification of a few details you need to add. Overall, the results are clear and compelling with few exceptions which have been added to the corrections. Also, the use of the word "underground" has more than a single interpretation. You make a systematic contribution to the research literature in this area of investigation. The longitudinal nature of the work is a welcome addition. Overall, this is a high-quality manuscript that has implications for the theoretical basis and research developent in cave ecology.

We thank the reviewer for the positive comments.

Specific comments and corrections were incorporated in the pdf form of the MS.

Best wishes and regards.

We are grateful for the suggestions; we modified the text according. Some of the changes do not occur as the text has been modified following also the suggestions of the Editor and of the other two reviewers.

Round 2

Reviewer 2 Report

Review of the manuscript:

Title: Do salamanders limit the abundance of cave-adapted invertebrates in subterranean habitats? 

Authors: Raoul Manenti, Enrico Lunghi, Benedetta Barzaghi, Andrea Melotto, Mattia Falaschi, Gentile Francesco Ficetola

The authors have done a substantial work to improve the quality of the manuscript. In particular, the language is improved and additional information has been added. The structure of the manuscript was also improved. After reading the manuscript I think that a minor revision is necessary. 

Below are some major comments and suggestions. Specific major and minor comments are detailed along the entire MS, attached to this page.

Major suggestions:

Materials and Methods

I still think a map with the sampling region would be to the benefit of the MS, without specifically providing the name of the sampling site; something that helps the reader to easily understand your sampling design, given its heterogeneity (i.e. colours according to year; symbols according to microhabitats…). 

Results

Before the GLMMs results, please explore (in words) if there is a trend/are peculiarities over time, or between sampling sites, or between microhabitats. For instance I noticed in your supplements that in 2016 and in almost all 2015 samples, no salamanders have been observed. Is there an explanation? 

Supplementary materials: 

• Please ad an a in the column’s name Est - East; 
• Please ad units for area, distance from entrance, depth; 
• Please explain/give a short description of site typology (i.e. Interface artificial pool; Subterranean artificial pool are, for instance, unclear for me).

Stay safe and healthy! With my best regards.

Comments for author File: Comments.pdf

Author Response

The authors have done a substantial work to improve the quality of the manuscript. In particular, the language is improved and additional information has been added. The structure of the manuscript was also improved. After reading the manuscript I think that a minor revision is necessary.

Below are some major comments and suggestions. Specific major and minor comments are detailed along the entire MS, attached to this page.

We thank the reviewer for the positive comments.

Major suggestions:

Materials and Methods

I still think a map with the sampling region would be to the benefit of the MS, without specifically providing the name of the sampling site; something that helps the reader to easily understand your sampling design, given its heterogeneity (i.e. colours according to year; symbols according to microhabitats…).

Following the suggestion, we added a map. In the main text we show a map differentiating from artificial and natural caves. In Supplementary material we provide a map with the year of sampling for each cavity. In case of sampling in multiple years the last one prevails.

Results

Before the GLMMs results, please explore (in words) if there is a trend/are peculiarities over time, or between sampling sites, or between microhabitats. For instance I noticed in your supplements that in 2016 and in almost all 2015 samples, no salamanders have been observed. Is there an explanation?

Following the suggestion, we added a short sentence. There is not a particular explanation, but the data considered for these years belong to caves in which the fire salamander does not breed. We did not observe particular changes across time, at least comparing the data considered in this study.

Supplementary materials:

Please ad an a in the column’s name Est - East;

Please ad units for area, distance from entrance, depth;

Please explain/give a short description of site typology (i.e. Interface artificial pool; Subterranean artificial pool are, for instance, unclear for me).

We performed the suggested changes. We performed also quite all the changes suggested in the pdf. We also modified the figures and the title following the reviewer suggestion.

Stay safe and healthy! With my best regards.

Thank you!!!The same for you!