Advertisement Call Variation of Two Frog Species along an Urban–Rural Gradient in Shanghai, China

Round 1

Reviewer 1 Report

The manuscript presents some interesting data on the vocalisation of two frog species. However, I think that there are some serious methodological flaws that prevent the results and the respective conclusions from being published in their current form. The first major issue I have is that the atuhors have not tested for any other possible reasons for the registered differences (e.g., temperature differences during the recordings, the effects of syntopy/allotopy) - they just assume that whatever differences they find, should be due to the "urbanisation index" they have calculated. The index itself is strangely calculated using 10-year-old satellite images instead of just mapping the current situation on the locations they have visited while recording the forgs (it is not clearly explained, but even if the images are more recent, the on-site mapping of land use is still the better option for such type of study). Another major issue is the high number of sites and the low number of frogs recorded from each site, which increases the effects of individual variations. While it is not clearly stated in the text, from the supplementary materials it becomes evident that the actual urban locations are only two and three for Microhyla fissipes and Fejervarya multistriata, respectively. This is simply not enough. In view of the above, I recommend rejection of the manuscript - even though in theory revision is possible, it should be so thorough that it would present a new submission. Below are my comments from the text:

Line 111: Overlapping = syntopic?

Figure 1: Urban, not Urabn

Line 135: Why did you use 10-year-old satellite images instead of mapping current land-use types when you were in the field, recording the frogs?

Line 138: Generally "impervious area" means an area which either prevents or retards the entry of water into the soil - do you mean this, or something else? How did you define the different land-use types? How many types did you use? Describe in greater details.

Line 149: So, you captured the frogs in order to record their vocalization? Why was that necessary? Describe in greater details.

Line 155: DF is actually the only spectral parameter you use.

Line 157: Good that you mention Ziegler et al. - as they (and many other studies) point out, temperature has a pronounced effect on temporal call parameters. How have you controlled for that?

Line 166: "calculate" instead of "calculated".

Line 175: So, on average, 4-5 males from a site? This number is too low for accurate analyses.

Line 188: DF varies with size, and you have only recorded a few males from any given site. How have you controlled for this?

Line 204: No, you did not. You virtually have not tested for any other possible reason for the observed differences, your "urban" sites are too few, and on top of that you have decided that a 10-year-old satellite data could be reliably used as surrogate for current land-use types. 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

The manuscript “The changes of the advertisement calls of two frog species during the breeding season along an urban-rural gradient in Shanghai, China” aims to study the variation in the advertisement calls of two sympatric anurans along an urban–rural gradient. The subject addressed in the study is very important concerning the high degree of urbanization throughout observed worldwide, which causes loss and fragmentation of natural habitats, affecting populations of many species. Thus, I appreciate studies that seek to answer interesting ecological questions on this subject.

However, I found important theoretical and practical issues that must be considered in the study. First, urbanization itself does not affect frog calls; the increasing anthropogenic noise caused by urbanization influences the change of calls (as pointed out in some places in the introduction). Urbanization can affect the quality of the habitat and reduce the fitness of individuals, causing physiological changes in amphibians, which may reflect changes in call parameters. Many studies have investigated how anthropogenic noise affects the call of animals (see Zaffaroni-Caorsi et al. 2022, for a review on anurans; https://doi.org/10.1080/09524622.2022.2070543).

In this sense, I understood that the aim of the study was to assess whether the advertisement calls of two frog species are influenced by the degree of noise caused by urbanization, not by the level of urbanization as indicated in lines 88-89. It is assumed (by urbanization index) that more urbanized areas are noisier. This conceptual inference should be better stated in the introduction and methods of the study. However, if the authors wish to build an idea that frog calls are affected by urbanization due to urban constructions that offer barriers to signal transmission, I suggest that the authors look for the acoustic adaptation hypothesis, which postulates that animals change its acoustic signals to improve its transmission through the environment. Overall, Introduction should be improved.

Another important point to be considered is in relation to the analytical methods. Call parameters are influenced by body size, social factors and environmental variables such as temperature. Before testing the influence of the urbanization index on the call parameters I think it is necessary to verify at least the influence of temperature and body size on the call parameters. As the two species have a great overlap in the parameters of the calls, especially the duration of the call and the frequency, it is possible that when they occur in sympatry, they can exhibit changes in the calls. In this case, the best way to assess the influence of urbanization on species calls would be to control the factors that influence calls by building a mixed model = call parameter (response variable) ~ urbanization index + body size + temperature + situation (only one species / two species together).

The authors stated that they did not measure (body size) the males, nor did they take the temperature during recording, but sometimes it is possible to obtain at least the temperature on the day of recording from available online data sources. I think it is important to control these factors in the analyses to better access the influence of “urbanization” on call parameters. In addition, the discussion must be improved.

I find interesting that samples of the calls are deposited in a phonoteca (for example: Fonozoo – http://www.fonozoo.com/; FNJV – https://www2.ib.unicamp.br/fnjv/; Macaulay Library – https: //www.macaulaylibrary.org/). Additionally, although I am not a native English speaker, I strongly recommend that the study undergo a language review.

Some suggestions and comments can be found below. Based on my comments, I recommend that the manuscript needs a major review.

Comments

Title change suggestion: Advertisement call variation of two frog species along an urban-rural gradient in Xnagai, China".

Line 23: Delete “globally”

Lines 24-25: Suggestion: "Advertisement calls of anurans are used for communication and play an important role in reproduction; …"

Lines 25-26: It is not urbanization itself that affects frog calls, but the increasing anthropogenic noise caused by urbanization.

Line 17: Delete “dominant”. What does dominant species mean? Are species abundant in the city?

Lines 48-50: Additionally, anthropogenic substrate-borne vibrations (in urban environments) can also affect anuran calls (Caorsi et al. 2019; https://doi.org/10.1038/s41598-019-55639-0).

Lines 61-64: Another interesting study on this subject is that of Lourenço-de-Morais et al. (2028) (https://doi.org/10.1007/s11252-018-0767-5). They found that in the urban matrix there is lower species richness and that anuran communities in the urban matrix are nested subsets of non-urban matrix sites (rural areas).

Lines 96-97: Urbanization puts significant pressure on frog calls because of unnatural anthropogenic noise. Thus, changes in call parameters must be caused by noise, not urbanization itself. This should be made clear.

Line 138: Appendix S1; Add the coordinates (lat and long) of the sampling sites.

Lines 148-149: Is there the (number) collection permission or of the ethic committee?

Line 151: Delete “from the field recorder”

Line 170-171: Has the homoscedasticity of the data been assessed (it is also a necessary assumption of the regression)?

Line 175: “males of M. fissipes; males of F. multistriata”

Line 196: Figure 3. Show only the graphs of the parameters that were significant (CD, DF).

Line 200: Figure 4. Show only the graphs of the parameters that were significant (PR, DF). Figures 3 and 4 can be joined.

Line 204: To ensure that the variation in call parameters is due to urbanization, it is necessary to test at least the effects of body size and temperature on calls.

Lines 207-208: Hypothesis is a proposition that is assumed, regardless of whether it is true or false, as a principle from which a certain set of consequences can be deduced that explains a certain phenomenon. Broadly assuming that a variable will change is not an ecological hypothesis. A true hypothesis would be: "We hypothesized that the calls of the two frog species will present higher temporal and spectral parameters in more urbanized areas due to urban noise. By increasing the temporal parameters, they increase the detectability by conspecifics; and by increasing the spectral parameters, they reduce frequency overlap between calls and urban noise (anthropic noise frequencies are often concentrated at lower frequencies)." See the acoustic adaptation hypothesis, which proposes that species adjust call parameters to optimize their propagation through the environment (Morton 1975).

Line 210: Because of this, it is necessary to test the influence of body size on the dominant frequency. It is possible that frogs in urban areas are larger than those in rural areas. Urban areas may have fewer predators, for example. Larger males have lower frequencies, so the explanation for the lower dominant frequency in urbanized areas can be explained more by size than by anthropic noise (or “urbanization index”). It is necessary to remove the factors that cause confusion.

Lines 216-2017: Thus, males from urban areas probably have lower frequencies because they are larger.

Lines 224-226: Ok, but it is not possible to declare this in this study. As stated in the previous paragraph, the FD seems to be more influenced by body size.

Lines 27-228: Frog species increase the frequency to avoid overlapping and signal masking caused by anthropic noise.

Lines 231-232: How does this statement relate to the study?

Lines 237-239: I found this statement speculative. Even considering urban parks with high vegetation cover, there is probably a temperature gradient from rural to urban areas. Lower temperatures would be expected at sites in rural areas (higher pulse rate of F. multistriata) compared to sites in urban areas (lower pulse rate of F. multistriata).

Line 247: rural-urban gradient

Lines 247-249: CD of F. multistriata was not related to urbanization index.

Lines 249-250: It is not possible to infer anything about the choice or selection of females in this study.

Lines 256-257: Again, it is not possible to infer anything about the choice or selection of females in this study.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

This is a nicely structured and well written manuscript (which nevertheless might profit from careful language editing). Quantity and quality of the data, however, are limited. The only environmental variable considered is "Urbanization Index"; for frogs calling in a pond, local environmental conditions may be more relevant. Most importantly, temporal call parameters often change with temperature in frogs, but no temperature data are reported. Hence, the observed correlations (based on means of very small local samples) may be spurious.

As to interpretation of the results, the authors observe (line 219) that decreased Dominant Frequency may result from larger average body size in both species; the same relation may also hold for Call Duration. These patterns could be examined in more detail if morphometric data from the calling males were available (which does not seem to be the case).

The only other clear result is the lower Pulse Rate of F. multistriata in urban environments. This may be a side effect of lower temperatures at these sites (line 239).

In conclusion, I suggest to tone down the somewhat naive discussion on "adaptive mechanism of breeding behaviors" such as "the ability of males of either species to attract females might increase with increasing UI"; such adaptive stories are not supported by the data presented here (and arguing from first principles, one might suggest that intrasexual competition should be higher in denser populations in rural environments, and thus expect that in more stressful urban habitats selection on these traits might be relaxed, or constrained by other trade-offs).

Detailed comments:

line 1: suggest to shorten the title to

"Advertisement calls of two frog species along an urban-rural gradient in Shanghai, China"

lines 33ff: Abbreviations should not be used in an abstract, and never at the beginning of a sentence; I strongly recommend to write out the words here

line 95: Can you make more precise predictions (which parameters are expected to change in a certain direction)?

line 98: Not all observed changes have to be adaptive

line 107: What do the percentage values refer to (built-up area)?

line 149: Why "released"? There is no mention that the frogs were captured; please make clear whether there were only field recordings made at the breeding sites, or whether the frogs were captured to take some measurements?

Did you record environmental parameters, such as air and water temperature, during recording?

Please report (either here, or in results) whether frogs were calling in water or on land.

line 164: correct "to calculate"

lines 178ff: These values might be better presented in a small table

lines 187, 210, 234: Never start a sentence with an abbreviation

line 219: Did you measure body size of the frogs you recorded?

line 238: Did you measure air temperature, or temperature of the frogs?

line 242f: It is rather bold to make such a statement, and support it with a reference from 1995

lines 249f, 256f: Why would you expect increasing attractivity in urban environments? Your data would be in line with an interpretation based solely on body size and direct environmental effects.

lines 253f: The wording of this sentence implies a process that perhaps has not happened at all ("frog species changed ...; populations increasing ... decreasing ..."); you found a pattern along the gradient that may be explained by correlation with body size and unstudied local environmental variation, but does not provide strong evidence for changes of the calls as such.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The manuscript has been considerably improved, alhtough some points still remain problematic. The authors say they have conducted "ground surveys to distinguish the land-cover types if they were hard to define" - but have only used two land-use types for the UI (buildings and roads). Land cover is different to land use, but it seems that the authors use them as synonyms - additional details are needed in this section. Results regarding the UI are difficult to interpret and I suggest to include "distance to nearest road/building" parameter, as well as other land-use types such as "park", "arable land", etc. 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

I was happy to read the new version of the study. The authors did a good job and made substantial changes to the manuscript as suggested by the reviewers, improving its quality. I have just a few more suggestions before it is accepted for publication.

Line 30: Weren't 34 sites studied? 14 for M. fissipes, 12 for F. multistriata and eight for both species.

Line 118: Add the dates when recordings were obtained: May to July 2020 and 2021.

Lines 213-214: Delete "in 22 and 20 study sites, respectively, from May to July, 2020 and 2021".

Line 117: during the recordings, not experiment.

Line 218: In Appendix S2 or S3 (Table) it would be informative to include the number of males recorded at each study site.

Line 256: Is the number in parentheses the number of recorded males or sites? I suggest identifying the number of males recorded. The same for Appendix S2(Table).

Line 267: The

Lines 277-27: Some studies (Lukanov et al. 2014, Caorsi et al. 2017) also found that frog species decreased DF when exposed to anthropogenic noise. 

Lukanov S, Simenovska-Nikolova D, Tzankov N. 2014. Effects of traffic noise on the locomotion activity and vocalization of the marsh frog, Pelophylax ridibundus. North West J Zool. 10 (2):359–364.

Caorsi VZ, Both C, Cechin S, Antunes R, Borges-Martins M. 2017. Effects of traffic noise on the calling behavior of two neotropical hylid frogs. PLoS One. 12(5):14. doi:10.1371/journal. pone.0183342.

Lines 281-283: Discussion could be improved. Physical barriers (for example, the presence of dense vegetation or maybe buildings) in the environment may influence the propagation of acoustic signals, contributing to a greater attenuation and degradation of the signal. Therefore, the environmental effects on sound propagation will depend on the structural complexity of the environment. Lower frequency sounds propagate more efficiently in the environment (Kime et al. 2000, Gerhardt and Huber 2002). Animals that live in more heterogeneous habitats (forest environments) may present songs with a lower dominant frequency to improve signal transmission (Acoustic Adaptation Hypothesis). 

We know that the authors used a landscape measure to try to understand how urbanization affects the calls, but I think it's important that the limitations of the measure used be discussed. Other measures such as microhabitat characteristics and the sound pressure level (anthropic noise) of urban sites would be important variables to be included in future analyses.

Gerhardt, H. C. and F. Huber. 2002. Acoustic communication in insects and anurans: common problems and diverse solutions. 1st edn. University Of Chicago Press, Chicago.

Kime, N. M., W. R. Turner, and M. J. Ryan. 2000. The transmission of advertisement calls in Central American frogs. Behav. Ecol. 11: 71–83. 

Lines 290-291: ok, but how can PR be influenced by UI? Improve the discussion.

Author Response

Please see the attachment.

Author Response File: Author Response.docx