Answers to Reviewers Comments

ANSWERS TO REVIEWERS COMMENTS

Ricciardi, Boyer and Ollerton

Manuscript EBFI1641

Editor’s note:

Our answer and details of major changes made are listed below.

We try to condense the discussion as you suggested: however, since reviewers suggest some more analysis to be included, the total length is not significantly lower than before. We reorganize all the section to be sure that no methods are presents in the result section and no results are present in the discussion.

Reviewer #1

We have taken into account the suggestion of Reviewer 1 making the following changes:

1)We introduced a more detailed analysis about possible other factors involving in the mutualism. We have added more result involving anemone depth distribution and areas (Pag 10, rows 7-16in results, and corresponding discussion). We have also increased the description of the mutualism including the analysis of anemonefish group numerousness (Paragraph 3.3.3 and following discussion)

2)We think that Reviewer #1 have raised an important concern about the temporal relevance of our observations and the possible variations that could occur over time. We have included this aspect in the discussion (Pages16 (rows 20-24) and 17, (rows2-4)).

3)Results section has been re-organized following Reviewer suggestions. Now, Results are structured to answer directly to the 5 question proposed in the introduction. We have moved the description of the simulation methods from the result to the most appropriate Materials & Methods section (Paragraphs 2.2 and 2.3)

4)About the extension of Table 1, we are aware that this table is quite large, but we think it is the best way to present the observed interaction site-by-site. Eventually, if the Editor thinks that it’s necessary, we could think about move this table in an “Appendix” section.

5)Scatter graphs: as the plotted data don’t correspond to a mean but to a single data points, they don’t have any associated variability. We have added the best-fit line as suggested by Reviewer #3.

6)All other minor corrections indicated by the Reviewer #1 have been fixed.

Reviewer #2

We have taken into account the suggestion of Reviewer 2 making the following changes:

1)We completely agree with Reviewer 2 that many other confounding factors may influence the mutualism between anemonefish and their hosts. We did not measure the anemonefish size in our sampling, also if we are aware that it’s an important factor in the assemblage. However, our choice was motivated by the following reasons:

a)The measure of the body lenght of the animal includes necessarily the capture of all specimens inhabiting the anemones, to obtain an affordable measure. Also if it is a common procedure in anemonefish studies, we think that this could provoke an unacceptable stress in captured fish, considering that our study included the observation of more than 770 anemonefish. Moreover, it is against the BunakenNational Park rules of conduct.

b)Many studies already have focused on the topic of anemonefish size in the mutualism, see e.g. Buston (2003) Nature 24: 145-146 or Buston and Cant (2006) Oecologia 149 (2): 362-372.

However, we have included the analysis of anemonefish group numerousness (Results, paragraph 3.3.3, Pag. 12 row 21) and the following discussion (Pag. 15, rows 10-18)

Moreover, we have described the method used for the measurement of anemone areas: wehave included the description in Mat&Meth section (Pag 7, rows 1-5), Results (Pag.10. Rows 11-14) and following discussion (Pag 15, rows 6-12). Another Table (Tab. 5) has been added to show difference in areas of anemones inhabited by 2 species at the same time.

Unfortunately, we did not measure the anemone density, so we were not able to include these data in the analysis. We appreciate the suggestion and we will include this measurement in our future studies.

2)About the use of C-Score and Unique species use in EcoSim, Reviewer 2 is right indicating that we have used these 2 metrics because we use presence/absence matrices, as stated in the text (Pag. 8, row 8).Moreover, we agree that it could be possible to choose between other simulation algorithms instead the default one we have used in our analysis. We decided the use the algorithm proposed by the author of the software (Gotelli and Entsminger, 2001) because it is the most conservative null model and, following authors, “this algorithm has good Type I properties and good power for detecting non-random patterns in noisy data sets”. Moreover, we agree that form a biological point of view some other constraints seem to be appropriate, but software authors also affirm that “Using equiprobable of proportional rows sums inflates the probability of a Type I error (falsely rejecting the null hypothesis when it is true)”. So, we are confident that using the most conservative model can provide the most affordable results.

We modified the Mat&Met section adding some more details about statistical methods. We agree that an intra-site analysis of the anemones depth could be interesting, but we think that our sampling protocol was not designed to test differences in depth of anemone species depending on the locality, for this reason we have decided to not include this analysis in our paper.

3)See answer no. 2 to Reviewer 1.

4)See answer no. 3 to Reviewer 1.

5)Materials and Methods section has been completely re-organized following reviewer(s) indications.

6)We agree with the reviewer that this is an interesting point. However, also if our feeling is that that the anemonefish diversity is a consequence of the sea anemone diversity, we have not found any bibliographic reference that can support this hypothesis.

7)We discuss phylogenetic relationship between fish species (Pag. 14 (rows 23,24) and 15, rows 1-3)

Moreover, all minor updates suggested by the Reviewer have been changed following his/her indications, including:

-a wider map of the area, including whole Indonesia;

-a rationale for the sampling sites (Pag 6, 14-16)

-Fig. 4 and 5 has been corrected indicating the number of observations;

-English has been revised.

Reviewer #3

Here listed answers to the reviewer’s 3 suggestions:

-We agree that many other arguments could be added to the introduction, as this mutualism is one of the most studied in the reef ecology. However, as also Reviewer 2 indicated in his introduction, we believe that our introduction contain enough information also for not-specialist readers and that much other information could be easily found out having a look on the literature cited.

-We modified the SSR equation, as it contained a mistake in the previous version;

-We add best fit lines to Fig. 4 and 5, as requested by the reviewer;

-Clarified what we did mean indicating the anemonefish species in charge of defence (Pag 11, row 19-20)

-We deeply modified the Result section, moving the description of methods in the appropriate section, as suggested also by the others reviewers

-As suggested by the reviewer, we have introduced in our analysis the influence of the anemone size (end of Paragraph 2.1 and 3.1, with following discussion

-We reorganized the Results following the 5 questions present in the introduction

-We modified the Fig. 2 as suggested.