The Effects of Landscape Heterogeneity on East China Anuran Communities: Identifying Spatial Scales in an Urbanizing Landscape
Round 1
Reviewer 1 Report
I think, it is interesting and well written manuscript. The elaboration of the article is careful. The methodology developed is sufficiently clear and leaves no doubt as to how the survey was developed. The results section is well elaborated, supported by the methodology, as well as the discussion of the results.
However I have some formal recommendations for the authors. For this reason, it seems to me that this article is ready to be published with minor structural or formal revision.
Congratulations for the work.
ALbert Montori
Reference format in the paper (Introduction, Mat & Meth., Results and Dicussion) is not correct. Solve this please.
Lines 69-70: Shanghai is the city with the highest levels of urbanization in China. Really? or is one of the most...
Line 102: support some amphibian populations in Shanghai.
Line 112: We chose 1-km as the landscape size. We refer to 1km2 squares? 1 km diameter circles? 1 km radius? Explain it.
Line 120: 44 survey wetlands or 44 1x1 km2 survey wetlands or 1km circle. If you explain this before it is not needed to include it in this line
Line 142 con-stant constant
Line 144: PRESENCE or Presence see line 148 to uniformize it
Table 1: Area Metrics (configurational heterogeneity), Aggregation Metrics (configurational heterogeneity) ans Diversity Metrics (compositional heterogeneity), in bold please.
Line 222; surveyed, and Bufo surveyed. Bufo...
Table 2: SOrry, I don't understand the use of Naïve in STE. Could explain me please?
Table 3; I think that its not needed write Table 3a, b, c and d, but if the authors prefer indicate it for better posterior references in the text, please reduces the interline space or locate at left of 500m, 1000m 1500m and 2000m of radius respectively.
In table 5 I belive that horizontal lines or pointed lines to separe items is needed. I dubt if i.e. ENN_MN is for Total anuran abundance (is it) or Anuran species richness. Moreover, I believe that Table 5 could translate to Supl. Mat. It is too much long. Other solution could be show only significant results.
Table 4 is located after Table 5. Correct it. However this table is also too much long to facilitates the readig of the work. I propouse move to Suppl. Mat.
Line 327: We hypothesized this is caused by a combination of two factors. On the one hand, in cities, anurans might adapt to fragmentation by developing better dispersal abilities, as has been shown for a “urban adapter” On the other hand, or cities might preselect for species with good dispersal abilities.. I like these ideas.
References
Authors need to correct all reference format.
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report
General comments
This study seeks to infer the spatial scale at which urban landscape features influence anuran richness and abundance by means of three model species distributed across extant aquatic habitats in Shanghai, China. The authors did a commendable job in establishing a modelling framework linking observed amphibian estimates to various scales and factors, and the results are fundamentally interesting. I nevertheless think several major revisions will be required before it can be published. Most notably, it should be better caveated as a more exploratory study in several places, especially working towards the applicability of the results in a conservation management context. The presentation of the results would furthermore need to be revised in order to improve general readability. Several smaller clarifications are furthermore needed where information is missing (see comments below).
I furthermore provided several suggestions where I think some language improvements might be beneficial. These can of course be implemented on own accord (I am also not a native English speaker) and are not exhaustive. The article would therefore benefit from an additional proofreading.
Major comments
The idea that the study “highlights the importance of identifying the most suitable spatial scale in urban environments” (as stated in the abstract and the discussion) is partially true, but somewhat hyperbolic and open to misinterpretation. Although it provides a noteworthy step in identifying spatial scales and the landscape context broadly relevant in amphibian conservation planning in urban areas (e.g. compositional and configurational heterogeneity), the study is still rather exploratory and fundamental in nature, as is also acknowledged in the introduction. It is therefore still quite a few steps away from effectively informing conservation policy. For instance, the results are likely species and area specific (L 318-319), and yielded small differences at different spatial scales (L 333-334). I think the manuscript currently lacks a good discussion on where to go from here. What future studies would be needed towards a better mechanistic understanding of the environmental factors at play and improved amphibian conservation management in urban areas? For instance, since constructing diverse habitats patch, neighbouring habitats etc. (L 405-406) might be very beneficial, it could be worthwhile to create such patches and see how diversity patterns are affected. In addition, several comments on the present conservation status and policies in the study area might provide some valuable context here for comparison with other study systems.
For general readability and interpretability of the results, I would strongly suggest placing the extensive results listed in – especially – tables 3 and 4 and its various sub-tables (a, b, c, etc.) to the supplement. These results can better be summarized in a written form in the results section (as is mostly already done) with reference to the extended tables in the supplement. Alternatively, the various model outputs (e.g. coefficients and significant predictors per species and distance class) could furthermore be summarized in a graphical form. Also, the summed Akaike weights are not very clearly placed in the tables (perhaps better italicized) and are redundant in several instances that only yielded one model within the AICc cut-off.
In addition, the data availability statement currently lists no original data or associated data sources.
Minor comments
L 46-48: This is slightly overstated. Better to mention attention for the subject has increased in recent years (with increasing urbanization).
L 85-87: This question reads confusing in its present form. Although from the context it becomes clear where you want to go, I would suggest simplifying/rephrasing here.
L 103: The version number of the software is rather redundant here. It might be better to state what the date and/or source was of the underlying satellite images that were used as the basis for site selection.
L 120-121: Although this map provides a good indication of the general position and spacing between the study sites, please also provide the GPS coordinates in the appendix to make the study better replicable. Of course, if you feel this gives away too much vulnerable information that could adversely affect some species, it can be omitted. But please clarify in either case.
L 203: Indeed, a collinearity threshold of 0.7 is widely used but does not correspond to the threshold that was used for variable selection as stated in the caption of figure 1. Please shortly clarify the choice for 0.6 and/or edit in case of a misspelling.
L 231-232: As it seems the detection probability of M. fissipes exceeds the number of surveys that were performed per site (as stated in the methods L 130-131), this is slightly contradictory to the statement in L 226-229. Were the detection probabilities additionally used to correct the species diversity response variable (L 195-198) for potential misdetection of species? Please clarify or underpin why (not).
L 325-326: These references indeed link diversity patterns to far greater spatial scales. However, since these distances (>2km) were not taken into account in the present study, I fail to see how these can be compared.
Language/in-text suggestions
L 16: “cities”
L 19: Suggest rephrasing to “diversity patterns”
L 24-25: “..large spatial scales, i.e. 1500 m and 2000 m, best predicted anuran relative abundance and richness”
L 32: “anuran relative abundance”
L 42-44: Slightly confusing - I suggest rephrasing to “by modifying land use and landscape heterogeneity”
L 49: “urban areas”
L 50: “cities”
L 54-56: If I understand correctly, something like “a higher compositional heterogeneity in landscapes by diversifying crops can support higher levels of biodiversity” might be more clear.
L 56: “a complex spatial pattern”
L 60: Suggest rephrasing to “lower impact on biodiversity”
L 67: Remove “the”
L 74: “landscapes”
L 75-76: Suggest rephrasing to “relate to amphibian diversity”
L 81: “and some ecologists”
L 84: Suggest rephrase to “from 500 to 2000 m radius”
L 87: remove “the”
L 104-105: Better just to state something like “the final survey sites were selected based on two criteria”
L 116: Some spaces missing between words
L 124-125: Suggest rephrasing to “Survey methods followed Zhang et al. (2016)”
L 127: Replace “chosen” to “performed”
L 141: Some spaces missing between words
Table 1. Please make the cut-off between the ‘Area Metrics (configurational heterogeneity)’ and ‘Diversity Metrics (compositional heterogeneity)’ slightly more clear. I would suggest displaying these in italics, and better aligned to the left and not centred.
L 201: Better to refer back to table 1 here for an overview of all used predictor variables.
L 202-203: Suggest rephrasing to “…to obtain a pair-wise correlation matrix for all predictors”.
L 209-210: Better list the abbreviation (AICc) here already for reference throughout the further text.
L 264: Full stop better replaced by comma in this sentence.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
General comments
Thank you for your responses to the first review and for clarifying the elements of your analyses. I think the revised version is sufficiently improved, including the presentation and visualization of the results, and I think the manuscript is acceptable for publication after some remaining minor and in-text clarifications and edits.
Although a data availability statement is now provided, it is up to the discretion of the editor to assess whether withholding the underlying data for the present paper (because the funded project is not yet completed) is in line with journal policies/ethics. I would argue that only the data that pertain to the present paper and are minimally needed for its replication could already be made publicly available, while additional project data can be made available with further research outputs.
Minor comments
Concerning the newly included figures outlining the model coefficients: these have identical captions and variables except for the different spatial extents. Alternatively, these could combined into a composite figure (with A, B, C and D corresponding to 500m, 1000m etc.) and captured under a single caption. I think adding the distances as titles above the various figures would generally allow for an easier interpretation.
Regarding my previous comment on L 231-232 about the detection probability of M. fissipes not aligning with the sampling effort: it is still stated in L 351-355 that “…indicating that the detection probability in the anuran surveys was sufficiently high, i.e., as long as an amphibian species existed, it would be detected without omission.” If I understand correctly however, this statement still does not apply to M. fissipes, and is at odds with the four surveys required for the species (versus the three conducted). A slight nuance might therefore be in order for this single species, although this would of course have minimal effects on the overall results and interpretation.
L 17-20. This central sentence in the abstract is a bit of a run-on sentence, and remains somewhat confusing. I would suggest rephrasing “nor have we identified” (L 19), and “the management/estimation of effects of the urban ecosystem” is rather unclear and could be simplified. The whole sentence might alternatively be split up in two shorter sentences.
L 21: “anuran abundance/richness”
L 31: I think this should read something like “… because not all anurans respond to the same spatial scale”
Author Response
Please see the attachment
Author Response File: 
Author Response.docx
