Native and Invasive Small Mammals in Urban Habitats along the Commercial Axis Connecting Benin and Niger, West Africa
Julien Claude
Round 1
Reviewer 1 Report
I would like to thanks the authors in having taken my remarks into consideration and modified their ms accordingly.
In the first set of comments, I am now reinsured that inter-trap spacing at the site scale was small enough (i.e. – at the room scale meaning only a few m²) to not bias capture probability according to species trait.
In the second set of comments, I agree that year and spatial gradient of sampling were correlated – which reinforces the authors results and conclusion.
Author Response
Please see the attachment.
Reviewer 2 Report
The authors have followed all my suggestions. The only one that made me confused was Comment 4 Sampling completeness. I do not think that line 215-217 answered the question (or the version I read have different line numbers), but since Simpson and Shannon indices are now provided elsewere, I consider it is fine (ideally having a confidence interval on S10 for each locality would probably have been the best, but maybe would have had too much information to the reader). I have therefore no problem with the version, and hope to see that paper published soon.
best regards,
Julien CLAUDE
Institut des Sciences de l'Evolution de Montpellier.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 3 Report
The present manuscript of Hima et al. is a comprehensive analysis of community structure of commensal small mammals in urban habitats in western Africa. It reports the summary of intensive field work (27 000 trap nights) performed during 13 years at 66 (!) localities in two West-African countries, which makes it one of the most valuable sources of this kind of information. The data are properly analysed by the approaches of community ecology, paper is clearly written and the results are very interesting (e.g. the south-north gradient in the proportion of invasive species, significant associations among some species, etc.). I think the manuscript is worth of publishing in the Diversity journal, as it is very actual and the results can have important consequences e.g. in epidemiology of emerging rodent-borne zoonoses. I have only rather minor comments that could be used for the improvement of presentation of results.
More important comments:
(1) The taxonomy (species identification) can play an important role in all community ecology analyses. Even if the authors probably did their best, I think that this could be still improved. Especially important it can be in the case of Crocidura as this genus was very abundant (in lesser extent it can be also the case of Arvicanhis). I believe that most of specimens are C. olivieri, but the authors already mention that part of material in Niamey are C. viaria and there were probably even more species in the samples (as suggested at r. 354). Individual species can have different ecological requirements and it really make difference (e.g. in analyses of species richness in species-poor communities) if there was a single species or three species. I think authors should somehow argue that the results are not significantly biased by incorrect/poor species identification.
(2) One of the most obvious results is higher proportion of invasive species in the South compared to the North. I think this result should be better visualized, e.g. in the form of pie-charts on the map (for example as a separate map forming a part of Figure 1).
(3) I do not see any reason to show Fig. 2. What is its aim? Just to show that the trapping was intensive enough to detect most common commensal species (r. 147-149)? The rarefaction method is usually used to compare species richness in samples with unequal size, but this is not the case. Also, it is questionable to merge "northern" and "southern" localities, and there is also a problem with identification of "rare" species (see above). Anyway, even if you (and the editor) decide to keep Fig. 2, you should at least describe the axes.
(4) I do not understand the right part of Fig. 3 ("Cartography of species and their abundance distribution"). The legend is not satisfactory, the description of axes is missing ... On the left part of Fig. 3 (factorial analysis) - what are the lines connecting the points? This should be described in the legend.
(5) Some parts of Discussion are too much disputable. Is it really reasonable to compare species richness of commensal small mammals in West Africa with randomly selected studies in Pakistan and Borneo? There are so many confounding factors, e.g. completely different ecological and historical conditions, reliability of species identification/delimitation (also in Benin/Niger - see above), etc. The same for r. 449-455. I think there are many other results properly discussed, so it is possible to remove these questionable parts.
(6) In general, the paper overlooks the temporal variation in community structure. The data were collected during 13 years - is there any obvious change e.g. in the proportion of invasive species? (like in Senegal, where some of the authors have worked on ongoing house mouse invasion in northern Senegal or black rat invasion in south-eastern Senegal)
Minor comments:
55 - "history", not "History" 86 and 92 - the same information is given twice, redundant 164-165 - a single sentence should not form a separate paragraph 186 - "were", not "where" 224-225 - ad identification of Arvicanthis - see above, it would be interesting to have more precise species identification for this genus. Based on the recent paper of Bryja et al. 2019 (J Zool Syst Evol Res, Fig. 7) there should be only a single species in the south, but the new data from Benin can be very valuable to solve this question). Can different species be sympatric? Also, it is not clear what exactly are "central sites". 229 - reference "Jacquet et al., comm. pers." can be replaced by "Jacquet et al. 2015, BMC Evol Biol" - at least I found that C. viaria were reported from Niamey in this paper. 281-282 - if the difference was not statistically tested, remove the word "significantly" 362 - are these estimates really similar (9 "species" in present study vs. 17 or 19 in other mentioned studies)? I suggest some rewording. 416 - I think you can replace "our own unpublished observations" by "this study", if I understood it correctly 438 - add reference to Table 1 440-448 - too redundant with Results 483-484 - should this really be the first sentence of "Conclusion"? The primary aim of the paper is not focused on shrews. 495-497 - the contribution of E.A. should be specified
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 4 Report
I only completed a partial review of this manuscript (i.e., first 4 pages) because the Editor contacted me that they had already received 3 reviews back from 3 other reviewers. For this reason, my assessment is incomplete, and what I have to offer is based on the first four pages. In general, the article appears to be of interest and significance to readers in West Africa as the paper details small mammals (mostly rodents) in urban habitats along the border between Benin and Niger. Commensal rodents were the most commonly found, which is not surprising. From what I could tell, it appeared that the paper was put together fairly well, but there needs significant editing of the grammar and the methodology/experimental design. Also, I was unsure if the statistical analysis was correct given the experimental design, but this may just need more clarification. There were some inferences in the Abstract that seemed to be fully speculation, and those should be fixed both in the Abstract and probably (I'm assuming appearing again) in the Discussion. I have sent along a scan of my comments on the first four pages of the manuscript, which includes corrections to grammar and content as well as general edits that should be considered in a revision.
Additional info after skimming the remaining parts of the manuscript are:
1-You need to add x-axis and y-axis labels to Figure 2
2-It makes sense that R. rattus is highlighted because by far there were the most numbers of individuals captured in your study, but the 2nd and 3rd most abundant were the two natives (Mastomys natalensis, and Crocidura spp, respectively)—these two native rodents deserve more attention and discussion (in the Discussion at least, maybe too in the Intro).
3-Table 1 is a lot of information, but I think it is fine to leave in. I suggest two things: 1) define “trap nights” in the Table caption (eg, 10 traps set for 1 night is 10 trap nights). It would be good to add one more column next to “Total captures” that is “Captures per trap-night”….this is the only (appropriate) way that your sites can be compared for total captures.
4-Discussion:
-Line 352: It says “In rural and urban areas” but in the title it only says urban areas; in the methods it sounded like only urban areas. Please clarify here and elsewhere.
-Typically Rattus rattus is found in areas where there are structures to climb—ie, natural structures like trees, and unnatural structures like buildings. Did your study allow you to make any such statement between incidence of vertical structure and Rattus rattus presence or abundance? If so, state this. There are a few good reviews about Rattus rattus, including:
Shiels et al (2104) Pacific Science 68: 145-184.
Banks and Hughes (2011) Wildlife Research 39: 78-88.
Line 375-378: Any ideas of why such a pattern was found? If so, explain.
Comments for author File: Comments.zip
Author Response
Please see the attachment.
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
Review of “Native and invasive small mammals in urban habitats along the commercial axis connecting Benin and Niger, West Africa” by Hima et al. - Diversity – ms n° 600069.
Based on a 12 years survey across a South-to-North gradient in latitude between Benin and Niger (West Africa), the present study describes the species richness and diversity of the small mammal community assemblage with a focus on native-to-invasive species association patterns. The main result of the study shows that native-to-invasive association segregates – specifically between “large bodied” rodent species (ie Mastomys versus Rattus).
The scope of the paper is very interesting, and very well written (fully understandable and fluent to read - at least for a non-native English speaker/writer).
I have two main remarks
1] Although all the analyses presented in the study are sophisticated and robust to the most important pitfalls the data could have contained – I found a lack of details involving the trapping methods used on the different sites investigated. More specifically – there is a lack of information on the spatial arrangement of the trapping design per site. Capture probability highly depends on space use behavior of each species – which are very different between very small mammals like shrews or mice compared to larger rodents i.e., rats. As a consequence – if inter trap spacing is not at a minimal scale to cope with the space use behavior of the species with the lowest movement amplitude – capture probability is unequal according to species – and proxy of species abundance based on the total number of individuals caught on a grid is biased toward those for which the trapping design is the most adapted.
More details on the trapping design at the site scale is necessary to fully encompass any variation in capture probability according to species specific space use behavior.
2] The data comprises trapping sessions on sites that were made over 12 years. One can suspect that species population abundances varied from year to year – therefore generating potential temporal heterogeneity in capture probabilities of the different species. Again – such heterogeneity can bias toward an increase or decrease in trapping subset of individuals of a given species compared to other according to sites – not due to spatial pattern of species distribution - but to temporal variation in species abundance. There is no mention of how the authors dealt with this situation in the manuscript.
In conclusion, while the analyses presented in the paper dealing with specific richness may be less sensitive (but only less – information on trapping effort is really needed – i.e., how many traps per three to five consecutive night) to trapping design, those involving a proxy of species abundance only based on the number of individuals trapped are – in my opinion – less trusty. Authors should provide convincing details about trapping design to keep their abundance-based diversity index analysis in their paper. Temporal variation should also be presented – and eventually discussed in front of the conclusion made about native-to-invasive species associations.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 2 Report
This paper describes rodent communities in urban habitats in West Africa. The paper is well written and it is an important synthesis of trapping campaigns in Benin and Niger. I have very few comments on it, the analyses are well done and most conclusions are based on factual data and/or done with caution. My remarks are all minor. I suggest to add some more info for the reader in tables and on figures, and to provide some details about the way that localities and sites are defined.
1. In sampling section, author should explain how they differentiate localities from sites (how many squared km for a locality, what is the surface sampled for a site).
2. On the map border between countries should be better differentiated from road.
3. If there is any approval notice from ethical comity, they should be provided in the sampling section.
4. Sampling completeness. I agree that the whole sampling may be representative of a community at large geographical scale, but the rarefaction analysis as it is done, does not say whether small localities were sufficiently sampled. In particular, many localities contained less than 50 individuals sampled. A zoom in the rarefaction curve seems needed to have an idea of the percentage of species that might be missed in these localities. l.194. you say that most taxa have been sampled in the community. I would rather say that most taxa have been sampled on the whole investigated area. Note that you selected to pool the species across all localities which may generate biases, indeed the rarefaction curve may differ from localities to localities since species diversity differ geographically (you show that in the paper). In addition, it should be noted that Arvicanthis is not identified to the species level while it may contain up to, at least 3 species (Crocidura spp is a similar problem), since the author have molecular data for some of them, it would be good to state somewhere the minimal potential number of species units they might represent.
Elsewhere, authors rarefied localities to 10 individuals to have an idea on how latitude shape diversity. It is a good idea. Simpson and Shannon indices might be a good complement to analyse since rarefaction curves may depend on surprise and evenness. I suggest to add the three indices in table 1.
Interpretation for rare species should be done with caution. For instance, mice are only reported in the two largest samples suggesting that they could have been unnoticed elsewhere because of insufficient sampling (although authors provide another explanation).
5. caption table 1. remove "species" in relative abundance p(species), it makes the sentence unclear.
6. I was wondering whether native (african) species could be finally considered as invasive. At first, it seems indeed that authors only consider exotic species as invasive in the intro, but the development of urban conditions can also generate good conditions for some african species to become invasive. Authors seem to modify their a priori in the discussion for Crocidura. Well... that's a semantic problem, but It might be interesting to insist on that because change in structure of local communities due to urbanisation (and not necessarily by the introduction of exotic species) could also generate risks of emergence (that could be also amplified (or buffered) by non native).
7. l.210-211 is not necessary (it is repeating just what is written before).
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
Please see the attachment.
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.docx
