Seasonal Variation in the Diversity, Abundance, and Spatial Distribution of Terrestrial Mammals in the Pénéssoulou Classified Forest

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors present the results of an interesting study of the diversity of a terrestrial mammal community in northwest Benin. They examine variation in diversity, relative abundance and spatial distribution in relation to a climate divided into distinct wet (rainy) and dry seasons. Their data derives from a large spatial array of camera traps within a state forest reserve. The deployment of the camera traps was in relation to forest trails and local knowledge of mammal use. This had an admitted limitation of under-sampling a key habitat, gallery forest, and advice on the proportional sampling of habitats would be welcome to clarify this constraint. The reserve is divided by a state road but there is little reference to any bias it may introduce such as roadkill and/or a barrier to dispersal with a seasonal effect. Interpretation of causality in the results is somewhat speculative because there was no contemporaneous sampling of resource drivers for any of the mammal species. Rather the authors rely on the literature and assumed logic.

The manuscript is well-written, and the results presented in informative figures and tables. The metrics chosen to represent diversity, equitability and dispersion were well-founded and employed constructively. The authors concluded that resource dynamics (not actually measured) and habitat constraints (were habitats equitably sampled?) drive seasonal distribution patterns in this mammal community. This is a useful and somewhat novel result that can improve management of this and other communities in at least seasonal wet-dry climates.

There are a few minor issues that should be corrected before acceptance as follows:

L40: into any environmental

L75: The study is about mammals not animals so ‘number of mammals’.

L88-9: Is 1100-1300 mm the total rainfall for the study site or the differential from neighbouring areas? This is unclear.

L96: What ‘creatures’? Is this a reference to mammals consistent with the focus of the study?

L146: account of rare

L182: Font size change from equations 5 to 6?

L206-322: Why is the text of the results in italics?

L234: 10 mammal species

L343: ‘author(s) name’ [24] explain

L427: shape mammal activity

L464: Add an additional appendix giving a more expansive description of habitat types used in the analysis.

Comments on the Quality of English Language

A few minor errors listed above

Author Response

Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted/in track changes in the re-submitted files. We colored our changes in yellow for this reviewer. 

Comments 1: The authors present the results of an interesting study of the diversity of a terrestrial mammal community in northwest Benin. They examine variation in diversity, relative abundance and spatial distribution in relation to a climate divided into distinct wet (rainy) and dry seasons. Their data derives from a large spatial array of camera traps within a state forest reserve. The deployment of the camera traps was in relation to forest trails and local knowledge of mammal use. This had an admitted limitation of under-sampling a key habitat, gallery forest, and advice on the proportional sampling of habitats would be welcome to clarify this constraint. The reserve is divided by a state road but there is little reference to any bias it may introduce such as roadkill and/or a barrier to dispersal with a seasonal effect. Interpretation of causality in the results is somewhat speculative because there was no contemporaneous sampling of resource drivers for any of the mammal species. Rather the authors rely on the literature and assumed logic

Response 1: Thank you for this important remark. In a recent study by N’Woueni et al. (2019), we showed that road networks within forests can influence biodiversity through various mediating factors. However, in our study area, awareness campaigns and road safety measures implemented by forestry authorities are actively enforced and monitored. Animal movements across the road are observed to be safe, and no cases of roadkill have been recorded to date. We agree that this variable could potentially influence mammal dispersal, but it was not included in the study because the situation is regulated and any impact on mammal diversity is likely to be marginal. Reference: M'Woueni, D., Gaoue, O. G., Balagueman, R. O., Biaou, H. S., & Natta, A. K. (2019). Road mediated spatio-temporal tree decline in traditional agroforests in an African biosphere reserve. Global Ecology and Conservation, 20, e00796.   Comment 2: L40: “into any environmental” Response 2: This has been corrected in the revised version (page 3)   Comment 3: L75: Use of “animals” instead of “mammals” Response 3: This has been corrected to “number of mammals” in the revised version (paged 3)   Comment 4: L88–89: Is 1100–1300 mm the total rainfall for the study site or the differential from neighbouring areas? This is unclear. Response 4: Localized rainfall data are difficult to obtain due to the lack of appropriate measuring equipment. We therefore rely on meteorological data from nearby stations that cover the study area. The value of 1100–1300 mm represents the differential from surrounding areas rather than an exact localized measurement. This was clarified in the reviewed version (page 4).   Comment 5: L96: Use of the word “creatures” Response 5: The term “creatures” has been replaced with “mammal species” in the revised version to maintain consistency with the study focus (page 4)   Comment 6: L146: “account of rare” Response 6: This issue has been addressed in the revised version (page 5)   Comment 7: L182: Change in font size between equations 5 and 6 Response 7: This was an unintentional formatting error and has been corrected (page 5 to 7)   Comment 8: L206–322: Results section in italics Response 8: We initially followed the journal’s template, but this appears to have been a formatting mistake. The italics have been removed in the revised version (page 7 to 14).   Comment 9: L234: “10 mammal species” Response 9: This has been corrected as suggested (page 9)   Comment 10: L343: “author(s) name” [24] explain Response 10: This has been corrected in the revised version (page   Comment 11: L427: “shape mammal activity” Response 11: This has been corrected in the revised version (page 17)   Comment 12: L464: Add an additional appendix giving a more expansive description of habitat types used in the analysis. Response 12: An additional appendix has been added with a detailed table describing all habitat types included in the analysis (see Appendix A3 on page 20).  Additional clarifications Regarding english and figures improvement, the manuscript was sent to the language editing services pending their feedback.

Reviewer 2 Report

Comments and Suggestions for Authors

Review

This article is a classic and well-executed faunistic study. Its strengths include the well-organized method of animal census, the careful placement of cameras, and the accurate mathematical analysis of the data, which allows for reliable ecological estimates of animal abundance and diversity. The simplicity and clarity of the illustrations are also noteworthy. Some captions may need to be changed for some of them. The description of the animal census methodology also requires some clarification. In particular, were the cameras moved during different observation seasons?

I have some comments regarding the terminology used. These are noted in the text. In some places, the authors use colloquial expressions. I don't believe the text requires a complete linguistic revision, but the ecological terminology and scientific style require more attention. The Acknowledgments section also requires a complete rewrite.

Generally, the study is valuable for the study and conservation of natural biological diversity and is worthy of publication.

Please also see the comments in the PDF file.

Keywords

Please add the keyword Biodiversity.

 

Figure 1 title.

“Pénéssoulou Classified Forest map with position of camera traps.”  

 

Results.

At the very beginning of this section, clearly and concisely provide information on how many observations were recorded overall, how many in each observation season, and how many on average per camera. Were there any cameras that recorded no observations or that recorded a significantly lower number of observations? This information is mandatory.

 

Line 212

Here and further in the text, also give the English trivial name for all animal species.

 

Figure 3b

This squad representation chart is completely pointless. Remove it.

 

Table 2 title

Must be “List of occurrence, mammalian species registered by seasons”

The table should contain the correct English names of animals.

Line 299 Body size distribution of mammal species according to seasons

The text does not explain how species were grouped by body size, nor which species were assigned to which groups. This should be clearly stated in the Materials and Methods section.

Line 313

In this section, as well as in Figure 6, it would be very helpful to indicate the number of animal observations per camera. This would show not only the ratio of species but also their total abundance in the area.

Figure 6b

Looking at the figure, it seems as if the cameras were moved seasonally, and therefore the survey area also changed. The Materials and Methods say nothing about the cameras being moved. If this is indeed the case, it would constitute a significant violation of the methodology and significantly alter the authors' assessment of all the data obtained. Camera movements should be clearly and thoroughly documented in the Materials and Methods.

Line 354-355

No, that's not true; you're misinterpreting these meanings. Remove this phrase.

Acknowledgments

Rewrite the Acknowledgments section completely. The text is grammatically incorrect.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Rewrite the Acknowledgments section completely. The text is grammatically incorrect.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

I recommend that you review Burton et al (Burton, A.C., Neilson, E., Moreira, D., Ladle, A., Steenweg, R., Fisher, J.T., Bayne, E. and Boutin, S., 2015. Wildlife camera trapping: a review and recommendations for linking surveys to ecological processes. Journal of applied ecology, 52(3), pp.675-685) as you revise your manuscript, especially in regard to camera trap methods and reporting results (e.g., species richness, relative abundance).

I also recommend that you consult Hall et al 1987. (The habitat concept and a plea for standard terminology) regarding the various terms you use when referring to “habitat” in this manuscript.  You lack consistency in this terminology.

Your study did not address the potential impacts of climate change on mammal diversity and distribution, which could alter habitat suitability and resource availability in the future.

Land use type and land cover type are used interchangeably in the manuscript.  Land cover type is a preferable term to use because the categories do not all reflect land use.

Terminology is not consistent throughout the manuscript for several terms.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The article structure is quite confusing. Please reorganize it more neatly based on the following points.

• You should include important results data in the abstract, it should attract the reader. Condense abstract to 150–200 words per MDPI style.
• Line 75: Please provide official meteorological data on rainfall mm for the months of application.
• In the Study Area section, mention the seasonal and climatic characteristics of the study area (which climate zone is it in, etc.).
• Line 104: It's not clear in Figure 1 how many phototraps there are between C0007 and C0011. Please indicate this either in the figure or in the text. Why is the distance between the devices here so close?
• Some figures (e.g., Figures 4, 5, and 6) are not descriptive enough, forcing the reader to rely on the text to understand the visual. Figure 4 should also include a genre-based explanation in the text. Figure 5 should be more descriptive. Was the distribution in Figure 6 based on camera images?
• Figure 1 caption begins with verbs (“Showing…”), others with nouns. Standardize to noun form, e.g., “Figure 1. Map of camera trap locations…”.
• Line 104: The term “rigorously monitored” in line 104 is vague; specify whether it refers to daily, weekly, or monthly checks.
• Line 107: The routine with which camera data was analyzed should be clearly stated. How many people were involved, and with what method were the recordings dataified? In other words, were they analyzed solely for presence or absence? The reasons for choosing GardePro A3S cameras (e.g., low-heat detection, night vision range, trigger speed) are not explained. Such technical information is necessary for scientific reporting.
• Using only 39 cameras across a 5,470-hectare area is too low a density to represent habitat diversity. Furthermore, having only two cameras in the gallery forest severely limits the results. This is noted in the discussion section but not justified in the methodology section.
• The cameras are said to be placed "1.5 km apart," but they have also been described as "100–200 m apart." These two pieces of information are contradictory and make the sampling design unclear. It would have been possible to conduct statistical analysis by grouping the distances. This would have made the inferences clearer.
• Include habitat type, water distance, and season as fixed effects in GLM models. Use a "GLMM" (mixed model) by defining the camera ID or location variable as a random effect.
• If there are any interesting or different images among the camera trap images, you can add them in the equipment features display mode (optional).
• To avoid a lack of randomness, explain why you decided to place cameras on "trails frequently used by mammals" (previous institutional inventory studies, remote sensing drones, etc.).
• Specify the number of camera traps/day for each season and make the analyses "effort-corrected." The total number of days each camera was active (trap days) was not specified. Without this information, the relative abundance index (RAI) values ​​of species cannot be compared.
• Whether human presence affected animal behavior during camera placement and monthly inspections was not discussed.
• Line 111: Briefly describe the brand and prominent features of the camera traps.
• Discussion: Assess the relationship between the increase in mammal numbers and camera distances and their relevance to water resources.Line 338-340: If the rainy season is winter and these mammals are hibernating species, indicate which ones they are and re-evaluate the difference in numbers.
• Lines 340-357 Discussion: These sources are confidential. Please structure the sentences as follows: “Compared to the bla bla findings in 10 and 27, this study includes the following different findings…”.
•  
• Some references are given collectively in the text, such as [24, 25, 26], but it is not clear which one supports which finding. More specific citations are preferred.
•  
• Line 382: For those unfamiliar, please add this definition: “A gallery forest is a forest formed as a corridor along rivers or wetlands, projecting only into sparsely forested landscapes such as savannas, grasslands, or deserts.”
•  
• Lines 387-388: Mention small, medium, or large mammals, classification as body size or species (as number or name) you found.
• Lines 390-391: What do you mean by that? Please expand the meaning.
• Line 421: The phrase "the greatest diversity" is too bold; rephrase it more gently.
• Lines 429-430: Explain the meaning of the sentence below, giving an example.
• The introduction doesn't sufficiently emphasize the study's difference and innovation from similar studies in the literature. This overshadows the article's academic merit. The gap in the literature should be more clearly highlighted. Similar studies, particularly those from West Africa (e.g., studies in neighboring countries), should be cited to clarify the contribution of this study.
• One of the initial hypotheses (that small mammals would cluster and large mammals would scatter) contradicts the results. This contradiction needs to be addressed more strongly in the discussion. Numerous statistics were applied, but no clear inferences were drawn regarding the hypotheses. Present the results more concretely and evaluate them more clearly in terms of the two hypotheses.
• Limitations such as the number of camera traps and the low representation in the gallery forest (only two cameras) were noted, but the impact of these limitations on the results should be discussed in more detail. They should be emphasized more clearly. For example, the placement of camera traps on specific paths limits the randomness of the sampling.
• GLM results are presented, but effect sizes and confidence intervals are not reported. Furthermore, Moran's I values ​​were found to be statistically insignificant; however, they are still extensively discussed in the discussion.
• The importance of the results for ecological and conservation management should be emphasized more strongly in the discussion section.
• Species behavior in different seasons (e.g., herbivore-carnivore differences, habitat use, hibernation rate) should be explained in more detail.
• Reasons for discrepancies between the hypothesis and the result should be more convincingly addressed. The conclusion section is currently too repetitive. It should be concise and conclude with more concrete management/conservation recommendations.
• Numerous statistics were applied, but not all of them clearly draw conclusions based on the hypotheses. Present the results more concretely and evaluate them more clearly in terms of the two hypotheses.
• The manuscript addresses an important ecological and conservation topic, focusing on the seasonal dynamics of terrestrial mammal communities in the Pénéssoulou Classified Forest, Benin. The use of camera trapping provides valuable baseline data, and the findings have relevance for biodiversity monitoring and forest management.
• However, the paper has several weaknesses that need to be addressed before it can be considered for publication. The originality of the study is not clearly articulated, the link between hypotheses and results requires stronger discussion, and methodological limitations are underdeveloped. Furthermore, citation practices and reference formatting need revision for consistency and clarity.
• Line 197: Model simplification based on AIC was performed, but the results (AIC values, justification for model selection) are not provided.
• GLM results may violate the independence assumption because random effects or spatial autocorrelation effects are not accounted for. This should be checked.
• The study area is small (5,470 ha) and the sample size is low (N=39). In this case, Moran's I value is not expected to be statistically significant; this analysis has low power. Furthermore, the Moran's I analysis uses a grid cell of 1 m²; this resolution is incompatible with the actual data scale (camera trap range) and is meaninglessly small. (Getis-Ord Gi* or Ripley’s K analyses may be more meaningful instead of Moran’s I).
• The measurement method for environmental variables such as "distance from water points" and "vegetation type" is not sufficiently explained (for example, do water points change seasonally?).
• The date on which Landsat imagery was used to extract land use classes (savanna, dense forest, etc.) is not specified. This classification may produce inaccurate results if there are differences in vegetation between seasons.
• The national or institutional permit number for wildlife observation is not specified. This is an ethical requirement, especially in studies conducted with camera traps. Furthermore, the participation of local communities (hunters, forest rangers) is described, but the ethical framework for informed consent or the use of local knowledge is not specified.
• When citing R packages (vegan, camtrapR, ape), include their full citations in the references section with version numbers (as partially done).
• “1 m² grid cell” in Moran’s I: Scientifically implausible given 1.5 km spacing. Correct scale or clarify it refers to analytical grid resolution.
• What is the relevance of the citations to articles that apply SPSS and study snail, such as the first two articles in the reference list?
• Define all abbreviations at first mention (e.g., GF, DF, CFM). Also give important ones under abbreviations title.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

Thank you for the comprehensive response to my review comments.  I believe the manuscript is substantially improved.  However, I do have some follow-up comments:

Wild-3880824-Seasonal variation in the diversity, abundance and spatial distribution of terrestrial mammals in the Pénéssoulou Classified Forest – post review comments

Original Comment 21: Page 4 – line 110: It is not clear why a 50-m radius sampling unit was established. Trail cameras generally provide a Field Of View from 45° to 70° so a 50-m radius sampling seems meaningless. Also, the general detection zone depth for a trail camera is typically between 12 and 30 m (the GardePro A3S has a 30-m detection range).

Response 21: The 50-m radius represents an effective sampling area centered on each camera; although detection depth is around 30 m, the area accounts for movement and detection probabilities within the immediate environment around the camera.

Follow-up Comment:  I continue to question the utility of using a 50-m radius sampling unit when the camera only sampled a limited field of view.

Original Comment 22: Page 4 – lines 110-111: How did you determine the probability of detection? And, how did this placement ensure a standardized probability of detection?

Response 22: Probability of detection was standardized by uniform camera deployment height, orientation, and spacing. Detectability was estimated indirectly through control checks and spatial sampling consistency rather than formal statistical models.

Follow-up Comment:  A reader will need to know specifically how probability of detection was “estimated” using control checks and spatial sampling consistency.  Also, why were formal statistical models not used and how do the results of your method compare to results from more commonly used statistical methods?

Author Response

Original Comment 21: Page 4 – line 110: It is not clear why a 50-m radius sampling unit was established. Trail cameras generally provide a Field Of View from 45° to 70° so a 50-m radius sampling seems meaningless. Also, the general detection zone depth for a trail camera is typically between 12 and 30 m (the GardePro A3S has a 30-m detection range).

Follow-up Comment:  I continue to question the utility of using a 50-m radius sampling unit when the camera only sampled a limited field of view.

Response to Comment 21 and Follow-up:

In order to represent the larger environment surrounding the camera where animal movement and detections could take place, the 50-meter radius sampling unit was selected as an efficient sampling area conceptually centered on each camera trap. The sampling radius takes into consideration possible mammal movement close to the trap and imperfect detection probabilities. This method aims to approximate an area where mammals that use the area can be reasonably detected by the camera, including those that pass or enter near the camera's effective range but might not be directly captured in the narrow field of view.

 

We recognize that this radius is a useful spatial unit used for data aggregation and analysis to standardize sampling effort rather than a literal, fixed detection boundary defined by camera optics. Given trail use and movement patterns, the 50-meter radius represents a spatial scale that corresponds to open habitat patches surrounding the camera where animals are likely to be found. It is in line with methods used in comparable ecological research to allow for cross-site and cross-camera comparison.

 

 

Original Comment 22: Page 4 – lines 110-111: How did you determine the probability of detection? And, how did this placement ensure a standardized probability of detection?

Response 22: Probability of detection was standardized by uniform camera deployment height, orientation, and spacing. Detectability was estimated indirectly through control checks and spatial sampling consistency rather than formal statistical models.

Follow-up Comment:  A reader will need to know specifically how probability of detection was “estimated” using control checks and spatial sampling consistency.  Also, why were formal statistical models not used and how do the results of your method compare to results from more commonly used statistical methods?

Response to Comment 22 and Follow-up:

Due to logistical limitations, the probability of detection was not directly modeled statistically; instead, it was standardized through uniform camera deployment protocols, which included uniform mounting height, orientation parallel to the ground and trails, and systematic spatial spacing (approximately 1.5 km). The goal of these actions was to lessen camera-to-camera variation in detection probability.

Field control techniques, such as the following, were used to indirectly estimate detectability.

           

Due to the constraints of the study design (e.g., number and spacing of cameras, available data) and the lack of resources, formal statistical models like occupancy or spatial capture-recapture were not used. Nonetheless, a practical method that approximates a standardized detection probability appropriate for the objectives of evaluating seasonal variation at the landscape scale is provided by the field-validation controls and uniformity in camera setup.

Although we acknowledge that formal statistical models typically provide more reliable detection probability estimates and correction for imperfect detectability, our method struck a balance between rigor and practicality given the study's context and logistical limitations. When sample sizes and data permit, such models may be incorporated into future work.

 

Additional clarifications

Regarding the English improvements, we have thoroughly edited the language throughout the manuscript, and the quality of the figures has been substantially enhanced in the revised version.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

Thank you for adding that habitat management should be effective in your wildlife conservation efforts, even if the research question gets stuck at certain points. You should position this research on more robust research questions in the future work. I wish you continued success in your new and innovative work.

Author Response

Original comment :

Thank you for adding that habitat management should be effective in your wildlife efforts, even if the research question gets stuck at certain points. You should position this research on more robust research questions in the future work. I wish you continued success in your new and innovative work.

Response to the comment :

Thank you very much for your thoughtful feedback and encouraging comments. We appreciate your recognition of the importance of effective habitat management in wildlife conservation, even though some aspects of the research question remain challenging.

We fully agree with your suggestion to position future studies around more robust and targeted research questions. This will allow us to deepen our understanding of the ecological mechanisms at play and enhance the scientific and practical contributions of our work.

Thank you again for your constructive insight and for your supportive words regarding our ongoing research efforts.

 

Additional clarifications

Regarding the English improvements, we have thoroughly edited the language throughout the manuscript, and the quality of the figures has been substantially enhanced in the revised version.