Feeding Ecology of the Endangered Barbary Deer (Mammalia: Cervidae) in the Akfadou Forest Enclosure, North Algeria

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

General assessment:

 

This is an interesting manuscript which might be of interest to the journal's audience. This paper focuses on the field of feeding ecology. Its uniqueness lies in the geographical area, in North Africa, and the species studied, the red deer. It is the only deer species on the African continent. The authors analyze the composition of the diet throughout the year. However, the statistical analyses are insufficient. When comparing the components by season, no information is provided on whether the variations are significant or not. The interpretations are not well justified. Variations in dietary diversity (species richness, H’ and evenness) can probably be analyzed by factorial linear models or GLM. I also suggest grouping the data into broader functional categories to increase statistical robustness, e.g. Woody, Graminoids, Fabaceae herbs, and Other forbs and use an ANOVA o PERMANOVA analysis to detect significant variations between seasons of the year.

The paper should not be acceptable in the current form. Listed below are several minor aspects that need to be changed in order to improve the article.

 

Specific points:

 

• Study area: provide the number of animals.
• Lines 72-73: Is it possible that the average annual temperature is 22.3ºC and that the temperature of the hottest month (July) is only 18-19ºC?
• Lines 87-88: provide the total number of samples.
• Lines 121-122: The authors say: “The Shannon-Wiener diversity index (H′) of dietary plant species was higher in spring (2.66) than in summer (2.64)”. It's difficult to believe that the differences are significant without a statistical test to support it.
• Lines 128 and 137: CFA or FCA?
• Line 128 and Figure 2: Authors say: “The first axis of the CFA (11.82% of the variance)” but Figure 2 shows a variance of 58.12%. There is also a discrepancy with the second axis
• Lines 157-159: If availability data is not available, preferences cannot be discussed. The same in line 166.
• Lines 172-174: This sentence should go into the discussion.
• Lines 176-185: This paragraph should go into the M&M section.
• Lines 189-190: Provide the citation data to see the differences.
• Line 211: The presence or absence of other herbivores can also be important.
• Lines 215-218: This statement is speculative since no data is provided on secondary compounds.
• Line 228: This change is not confirmed by appropriate statistical analysis.
• Line 256: This greater diversity is not confirmed by statistical analysis.
• Line 297: Clarify if the sentence refers to Arbutus fruits or Quercus acorns?
• Lines 300-303: It doesn't make much sense to compare such different species.
• Lines 322-323: It also doesn't make much sense to compare sexes if that factor hasn't been taken into account in the analyses.

Author Response

This is an interesting manuscript which might be of interest to the journal's audience. This paper focuses on the field of feeding ecology. Its uniqueness lies in the geographical area, in North Africa, and the species studied, the red deer. It is the only deer species on the African continent. The authors analyze the composition of the diet throughout the year. However, the statistical analyses are insufficient. When comparing the components by season, no information is provided on whether the variations are significant or not. The interpretations are not well justified. Variations in dietary diversity (species richness, H’ and evenness) can probably be analyzed by factorial linear models or GLM. I also suggest grouping the data into broader functional categories to increase statistical robustness, e.g. Woody, Graminoids, Fabaceae herbs, and Other forbs and use an ANOVA o PERMANOVA analysis to detect significant variations between seasons of the year.

We tested the species richness: chi-square test = 0.33, p = 0.90.

We tested the seasonal variations of broad categories (shrubs and trees, grasses, forbs). The factor season was not significant (F = 0.01, DF = 3, p = 0.99). The factor “category” was highly significant (F = 30.24, DF = 2, p = 0.001). We also compared seasonal diet within each category, shrubs and trees (chi-square = 4.59, DF = 3, p > 0.05), grasses (chi-square = 15.81, DF = 3, p < 0.001), forbs (chi-square = 12.15, DF = 3, p < 0.01).

The paper should not be acceptable in the current form. Listed below are several minor aspects that need to be changed in order to improve the article.

Specific points:

• Study area: provide the number of animals.

Done

• Lines 72-73: Is it possible that the average annual temperature is 22.3ºC and that the temperature of the hottest month (July) is only 18-19ºC?

Corrected

• Lines 87-88: provide the total number of samples.

Done

• Lines 121-122: The authors say: “The Shannon-Wiener diversity index (H′) of dietary plant species was higher in spring (2.66) than in summer (2.64)”. It's difficult to believe that the differences are significant without a statistical test to support it.

Differences are not significant

• Lines 128 and 137: CFA or FCA?

Corrected in FCA

• Line 128 and Figure 2: Authors say: “The first axis of the CFA (11.82% of the variance)” but Figure 2 shows a variance of 58.12%. There is also a discrepancy with the second axis

Corrected. The previous values were for inertia

• Lines 157-159: If availability data is not available, preferences cannot be discussed. The same in line 166.

The first sentence were modified accordingly, the second was deleted

• Lines 172-174: This sentence should go into the discussion.

The sentence was deleted

• Lines 176-185: This paragraph should go into the M&M section.

Most papers discuss the accuracy of the surveyed method as part of the discussion (usually at the beginning)

• Lines 189-190: Provide the citation data to see the differences.

Data were added (see also supplementary tables)

• Line 211: The presence or absence of other herbivores can also be important.

We added this information in M&M section

• Lines 215-218: This statement is speculative since no data is provided on secondary compounds.

We do not consider this statement to be speculative as it is not issued from our data but provide information from the literature that can explain the consumption of woody plants

• Line 228: This change is not confirmed by appropriate statistical analysis.

This sentence does not refer to our results, but is a general statement from the literature. Our results are discussed below

• Line 256: This greater diversity is not confirmed by statistical analysis.

Differences among species richness are not significant, but we can write that diversity was the greatest during spring

• Line 297: Clarify if the sentence refers to Arbutus fruits or Quercus acorns?

We modified the sentence in order to clarify the major intake of Arbutus unedo of some foreign red deer

• Lines 300-303: It doesn't make much sense to compare such different species.

Sentences were deleted

• Lines 322-323: It also doesn't make much sense to compare sexes if that factor hasn't been taken into account in the analyses.

Precisely, as sexes can have quite different diets we suggest that further studies should investigate if it can exist within the enclosure

 

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Authors

Thank you for your manuscript «Feeding ecology of the Endangered Barbary deer in the Akfadou Forest enclosure, North Algeria» which gives important insight to the feeding egology of barbary deer. However, I have a few points to ameliorate this manuscript:

In general:

• Please describe the reasoning behind the forest enclosure. Why is it there? What is its purpose? Why did you conduct the study there? Is is a reserve or an enclosure?
• The discussion is very hard to read and not really a discussion, but a list of all the studies that were done so far. I would suggest du compare the different studies in a table and to shorten the discussion to the parts, where you really discuss your findings.
• The conclusion conlcudes yours paper and thus does not contain references

Specific:

L55: I am not sure why this small populations is susceptible to diseases and forest fires and not any other threats. Maybe you can describe it in more detail or delete the sentence, as it is not relevant tot he work.

L60: You write that is is necessary to publish this work. It is not necessary in the deeper meaning of the word. Please change to it was decided or something similar.

L66: delete patrimonal, as it is out of context

Figure 1: I cannot find drivable tracks nor the water points in your card. You also describe in text form several streams, but only one is on the map. Please adjust.

L100: I do not understand how you identified the first 300 epidermal fragmenst per season. Please explain the process of identification.

L122/123: This is not a complete sentence. Please adjust.

Table 1: I do not unterstand the order of the listesd plants. Please do them in alphabetical order or in the order of annual N(%)

L128: Please explain abbreviations when you first use them. Also, here you list CFA, in the figure you write about FCA. What is correct?

L214-219: animals eat to cover their nutritional requirements. Is it known, that shrubs and trees have lots of protein? Why do you only discuss protein? And the protein-tannin connection is proven, therefore remove «appears». What I do not understand is why you make this connection, when you do not know nutritional contents of your trees and shrubs. Are there any other nutritional reasons that the animals consume shrubs and trees

L223-226: Please mention, what high energy plants you have identified in the enclosure and if they were readily consumed

 L241: Not a complete sentence, also L242

Author Response

Thank you for your manuscript «Feeding ecology of the Endangered Barbary deer in the Akfadou Forest enclosure, North Algeria» which gives important insight to the feeding egology of barbary deer. However, I have a few points to ameliorate this manuscript:

In general:

    Please describe the reasoning behind the forest enclosure. Why is it there? What is its purpose? Why did you conduct the study there? Is is a reserve or an enclosure?

1. Please describe the reasoning behind the forest enclosure. Why is it there? What is its purpose?

This enclosure was created for breeding Barbary red deer for future reinforcements of the wild population. The site was chosen after the habitat which is quite similar to that of the wild population.

Some information was added to the text.

2. Why did you conduct the study there?

This enclosure is the unique captive breeding center for Barbary deer in Algeria, and this study is part of the monitoring of the population. This information was added.

3. Is it a reserve or an enclosure?

This is an enclosure which is an official reserve devoted to the conservation of a small population of Barbary red deer.

The discussion is very hard to read and not really a discussion, but a list of all the studies that were done so far. I would suggest du compare the different studies in a table and to shorten the discussion to the parts, where you really discuss your findings.

We tried to discuss each of our results, first the annual diet, then each seasonal diet by comparison with diet studies, focusing on the Mediterranean area, and we even provide unpublished data as supplementary material.

The conclusion concludes yours paper and thus does not contain references.

In order to ground our perspectives, we think useful to bring in some papers that support some of them. Should we entitle this part: conclusion and perspectives?

Specific:

L55: I am not sure why this small populations is susceptible to diseases and forest fires and not any other threats. Maybe you can describe it in more detail or delete the sentence, as it is not relevant to the work.

We now refer to the conservation concern of this small population of an Algerian-Tunisian endemic subspecies. This work has been conducted in order to provide more information for the conservation of the species. The Akfadou forest enclosure is part of the program of conservation and protected from the main events which threaten the wild population.

L60: You write that is is necessary to publish this work. It is not necessary in the deeper meaning of the word. Please change to it was decided or something similar.

The sentence was modified accordingly.

L66: delete patrimonal, as it is out of context

The word was deleted.

Figure 1: I cannot find drivable tracks nor the water points in your card. You also describe in text form several streams, but only one is on the map. Please adjust.

The map was revised.

L100: I do not understand how you identified the first 300 epidermal fragments per season. Please explain the process of identification.

We collected ten fresh pellets from different latrine sites each month, then pooled pellets within seasons and identified the first 300 fragments of the 30 pellets pooled together. We added the information concerning the use of 30 pellets per season.

L122/123: This is not a complete sentence. Please adjust.

The sentence was modified.

Table 1: I do not understand the order of the listed plants. Please do them in alphabetical order or in the order of annual N (%)

Plant species in table 1 are now in alphabetic order within categories (in shrubs and trees we formerly listed trees befor shrubs).

L128: Please explain abbreviations when you first use them. Also, here you list CFA, in the figure you write about FCA. What is correct?

This was corrected and explained (FCA: Factorial correspondence analysis, line 111).

L214-219: animals eat to cover their nutritional requirements. Is it known, that shrubs and trees have lots of protein? Why do you only discuss protein? And the protein-tannin connection is proven, therefore remove «appears». What I do not understand is why you make this connection, when you do not know nutritional contents of your trees and shrubs. Are there any other nutritional reasons that the animals consume shrubs and trees

We thank the reviewer for this insightful comment. We agree that animals select their diet to meet a range of nutritional requirements, not only protein. Shrubs and trees are known to provide variable but sometimes substantial levels of crude protein, particularly in young leaves and during certain phenological stages. In addition to protein, these plant categories may also supply other important nutrients such as fiber, minerals, and secondary metabolites that can influence feeding behavior.

We acknowledge that focusing exclusively on protein in the previous version was an oversimplification. In the revised manuscript, we have broadened the discussion to include other potential nutritional drivers of shrub and tree consumption, including fiber content, mineral supply, and the role of secondary compounds.

Regarding the protein–tannin relationship, we agree that this interaction is well established in the literature and have therefore removed the term “appears” to reflect this consensus more accurately.

Among other nutritional reasons that the animals consume shrubs and trees, the nutrient values of plants is also discussed. Moreover, we clearly write that plant availability should me measured in the enclosure, and we add their nutrient value.

L223-226: Please mention, what high energy plants you have identified in the enclosure and if they were readily consumed

We identified four consumed plants with high energy content: Avena sterilis, Mentha sp, Stellaria sp and Trifolium sp. However, they account for only 22.44% of the diet.

L241: Not a complete sentence, also L242

Sentences were modified.

 

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

The dataset presented in this manuscript is valuable and provides relevant insight into the feeding ecology of a poorly studied and threatened subspecies. The collection of year-round dietary data represents a useful contribution to the understanding of Barbary deer ecology, particularly in a semi-controlled environment. However, in order to fully support the conclusions drawn, it is essential that the data are presented and interpreted strictly within the limits imposed by the study design and methodology.

The literature review is largely descriptive and lacks a clear conceptual framework. While the authors correctly point out the limited availability of species-specific studies on Barbary deer, this does not justify the absence of a broader theoretical context. Key ecological concepts relevant to herbivore feeding ecology are missing, including optimal foraging theory, resource selection in relation to food availability, niche theory, trophic plasticity, and functional responses of herbivores.

Although Barbary deer is geographically restricted, these frameworks have been extensively developed for other cervid species and large herbivores in Mediterranean and temperate systems, and could be used to strengthen the interpretation of the results. At present, the introduction reads more as a compilation of existing knowledge than as a structured argument leading to clearly defined ecological questions or hypotheses. Integrating these concepts would not only improve the scientific depth of the manuscript but also allow the authors to better position their findings within a broader ecological and comparative context.

The sampling design raises concerns regarding the independence and representativeness of the data. The collection of 10 pellet samples per month from latrine sites does not guarantee that samples originate from different individuals, and repeated sampling of the same individuals is highly likely. This issue is particularly important because the manuscript does not provide any information on population size or density within the enclosure. Without such context, it is impossible to assess whether the analysed samples adequately represent the dietary habits of the population, or rather reflect the feeding behaviour of a limited number of individuals. The authors should clarify how they minimized the risk of resampling the same individuals and provide at least a rough estimate of population size. This information is essential to evaluate the robustness and ecological relevance of the results.

The description of the microhistological procedure is incomplete, particularly regarding the preparation of faecal samples prior to microscopic analysis. Key steps such as rehydration, homogenization, subsampling, and slide preparation are not described. This lack of detail limits the reproducibility of the study and raises concerns about potential sampling bias. In addition, it is unclear how the 300 fragments per season were obtained (e.g. number of slides, fields of view, or samples analysed), which further complicates the interpretation of the quantitative results. Furthermore, the manuscript does not specify the size range of the analysed plant fragments or whether any size-based selection criteria were applied. Fragment size can strongly influence detectability and identification success, and without standardization (e.g. through sieving or defined microscopic fields), there is a risk that certain plant taxa are over- or under-represented. Clarification of these aspects is essential to assess the robustness of the dietary estimates.

The results provide a coherent and valuable overview of the diet composition of Barbary deer and its seasonal variation within the study area. The authors successfully document the consumption of a relatively broad spectrum of plant taxa (22 species) and demonstrate a consistent pattern in which shrubs and trees dominate the diet, complemented by grasses and forbs. The seasonal trends are ecologically plausible and align with expectations for Mediterranean environments, with increased grass consumption in spring and a shift towards browsing during the drier summer and autumn periods. The identification of key plant species contributing to the diet further strengthens the descriptive value of the dataset. Overall, the results suggest a certain degree of trophic flexibility, supporting the classification of Barbary deer as an intermediate feeder capable of adjusting its diet according to seasonal resource availability.

However, the results are presented primarily in a descriptive manner and are not fully exploited analytically. The statistical analysis is generally appropriate for exploratory purposes but remains limited in scope. The study relies mainly on descriptive statistics and multivariate ordination (FCA), without applying inferential tests to assess seasonal differences in diet composition. As a result, it is unclear whether the observed patterns are statistically meaningful or reflect sampling variability. In addition, the explanatory power of the FCA is relatively low and should be more explicitly acknowledged. The absence of measures of variability (e.g. standard deviations or confidence intervals) further limits the interpretability of the results. Furthermore, the interpretation occasionally extends beyond the data, particularly when inferring dietary preferences without information on resource availability.

Overall, the dataset itself is informative and represents a useful contribution to the knowledge of this poorly studied subspecies. However, the statistical treatment does not fully exploit its potential. The inclusion of more robust analytical approaches (e.g. PERMANOVA, GLMs, or methods suitable for compositional data), together with a clearer distinction between descriptive patterns and ecological inference, would substantially strengthen the study and enhance its impact.

The discussion would benefit from a stronger integration of established ecological frameworks. In its current form, it remains largely descriptive and comparison-based, without fully embedding the results within broader concepts such as optimal foraging theory, niche theory, or trophic plasticity. While the observed seasonal shifts in diet composition are ecologically plausible, they are not explicitly interpreted in terms of adaptive foraging strategies or niche dynamics. For instance, the apparent switching between grazing and browsing across seasons could be discussed within the framework of optimal foraging theory as a response to changing resource availability and quality. Similarly, the relatively broad spectrum of consumed plant taxa suggests a degree of trophic plasticity that could be more explicitly addressed in relation to niche breadth and flexibility. Even though Barbary deer is a geographically restricted subspecies, these concepts have been widely applied to other cervids and large herbivores, and their inclusion would considerably strengthen the ecological interpretation and general relevance of the study.

Finally, the discussion would benefit from a more explicit and comprehensive consideration of the study’s limitations. While some methodological constraints of microhistological analysis are mentioned, their implications for data interpretation are not fully explored. In particular, the potential bias associated with differential digestibility of plant tissues and the fact that fragment frequency does not reflect biomass intake should be clearly acknowledged. In addition, the representativeness of the dataset is not discussed. Issues related to sample size, potential pseudoreplication, and the lack of information on population size should be considered, as they may influence the generality of the results. A major limitation is the absence of data on food availability, which prevents interpretation in terms of resource selection. However, even a qualitative discussion of vegetation structure within the enclosure (e.g. relative abundance of woody versus herbaceous species) could provide useful ecological context and strengthen the interpretation of the findings. Addressing these points would lead to a more balanced and transparent discussion and improve the ecological relevance of the study.

 

Author Response

The dataset presented in this manuscript is valuable and provides relevant insight into the feeding ecology of a poorly studied and threatened subspecies. The collection of year-round dietary data represents a useful contribution to the understanding of Barbary deer ecology, particularly in a semi-controlled environment. However, in order to fully support the conclusions drawn, it is essential that the data are presented and interpreted strictly within the limits imposed by the study design and methodology.

The literature review is largely descriptive and lacks a clear conceptual framework. While the authors correctly point out the limited availability of species-specific studies on Barbary deer, this does not justify the absence of a broader theoretical context. Key ecological concepts relevant to herbivore feeding ecology are missing, including optimal foraging theory, resource selection in relation to food availability, niche theory, trophic plasticity, and functional responses of herbivores.  Although Barbary deer is geographically restricted, these frameworks have been extensively developed for other cervid species and large herbivores in Mediterranean and temperate systems, and could be used to strengthen the interpretation of the results.

At present, the introduction reads more as a compilation of existing knowledge than as a structured argument leading to clearly defined ecological questions or hypotheses. Integrating these concepts would not only improve the scientific depth of the manuscript but also allow the authors to better position their findings within a broader ecological and comparative context.

The aim of the study is to describe the seasonal diet of a poorly studied subspecies of red deer endemic of North Africa for conservation purpose, and not to test the foraging strategy hypothesis, which has already been done on the species in Europe. If absolutely necessary we can add the paragraph:

Dietary flexibility in ruminants is constrained by digestive morphology, particularly stomach structure, which nonetheless permits the exploitation of plant resources with contrasting physical and chemical properties. Feeding strategies are shaped by key ecological processes, including optimal foraging, resource selection in relation to food availability, niche differentiation, trophic plasticity, and functional responses. Large herbivores (Ungulate) feeding types have been widely classified along a continuum from grazers to browsers, based on diet composition (Hofmann & Stewart 1972; Gordon & Illius 1994, 1996; Gagnon & Chew 2000). Within this framework, species may specialize on grasses, woody vegetation, or adopt mixed feeding strategies, reflecting differences in trophic niche breadth and flexibility (Grzimek 1990). Intermediate feeders, in particular, exhibit greater trophic plasticity, enabling adaptive responses to environmental variability. However, diet selection in mammalian herbivores remains complex (Shipley 1999), as it results from the interaction between intrinsic constraints and fluctuating resource availability.

Gagnon M. & Chew A.E. 2000: Dietary preferences in extant African Bovidae. J. Mammal. 81: 490–511.

García-Gonzalez R. 1983: Epidermis foliares de algunas especies de Festuca, Poa y Bellardiochloa en el Pirineo Occidental. An. Jardin Bot. Madrid 39: 389–404.

Gill R.B., Carpenter L.H., Bartmann R.M. et al. 1983: Faecal analysis to estimate mule deer diets. J. Wildl. Manage. 47: 902–915.

Gordon I.J. & Illius A.W. 1994: The functional significance of the browser-grazer dichotomy in African ruminants. Oecologia 98: 167–175.

Gordon I.J. & Illius A.W. 1996: The nutritional ecology of African ruminants: a reinterpretation. J. Anim. Ecol. 65: 18–28. Grzimek B. 1990: Grzimek’s encyclopedia of mammals. McGraw-Hill, New York, U.S.A.

Hofmann R.R. & Stewart D.R.M. 1972: Grazers and browsers: a classification based on the stomach structure and feeding habits of East African ruminants. Mammalia 36: 226–240.

Shipley L.A. 1999: Grazers and browsers: how digestive morphology affects diet selection. In: Launchbaugh K.L., Sanders K.D. & Mosley J.C. (eds.), Grazing behavior of livestock and wildllife. Idaho Forest, University of Idaho, Wildlife and Range Experimental Station, bulletin no. 70, Moscow: 20–27.

The sampling design raises concerns regarding the independence and representativeness of the data. The collection of 10 pellet samples per month from latrine sites does not guarantee that samples originate from different individuals, and repeated sampling of the same individuals is highly likely. This issue is particularly important because the manuscript does not provide any information on population size or density within the enclosure. Without such context, it is impossible to assess whether the analysed samples adequately represent the dietary habits of the population, or rather reflect the feeding behaviour of a limited number of individuals. The authors should clarify how they minimized the risk of resampling the same individuals and provide at least a rough estimate of population size. This information is essential to evaluate the robustness and ecological relevance of the results.

We added information on the population size within the enclosure (26-32 individuals during the study. Barbary red deer most often occurred in groups of 5 to 15 individuals, so faecal samples were taken from unidentified individuals. Each month pellets were collected from latrines located apart in different zones of the enclosure. Following the protocol of Stewart & Stewart (1971), who studied 10 faecal pellets per area and season, we randomly selected 10 dried pellets for 300 epidermis identifications.

The description of the microhistological procedure is incomplete, particularly regarding the preparation of faecal samples prior to microscopic analysis. Key steps such as rehydration, homogenization, subsampling, and slide preparation are not described. This lack of detail limits the reproducibility of the study and raises concerns about potential sampling bias.

The microhistological procedure is described in our previous papers (Benamor et al., 2019, 2021, and Bounaceur et al., 2023). Is it necessary to recall it?

Fresh Pellets were air-dried and stored in paper bags in the field. Based on the work of García-Gonzalez (1983), we prepared a reference collection of the epidermis of the main plant species (families) collected within the enclosure. The method used to obtain the epidermis of the plant fragment involved carefully scraping it with a scalpel to separate tissues above the epidermis from different parts of the plant (adaxial and abaxial surfaces of leaves, stems, flowers). After rinsing in sodium hypochlorite solution (12 %), the epidermis was washed in distilled water, placed in a glycerine solution (50 %) and sealed on a slide with nail varnish (García-Gonzalez 1983). The diagnostic features of the plant epidermis, such as cells, fibres, trichomes, pores, stomata, vessels, intercellular structures and cell walls from each reference slide were photographed using a camera (Sony Cyber-Shot DSC-W380) fitted to an optical microscope at (10 × 0) and (10 × 40) magnification.

Preparation of sample slides. According to Chapuis (1980) we:

• dissociated the faecal material in sodium hypochlorite solution;
• let macerated for at least 4 hours;
• filtered to separate the fragments from the liquid part;
• rinsed several times with water.

In addition, it is unclear how the 300 fragments per season were obtained (e.g. number of slides, fields of view, or samples analysed), which further complicates the interpretation of the quantitative results.

The 300 fragments were obtained after pooling the content of 10 pellets collected each season. This is an unavoidable bias that is weakly relevant in a comparative study. The number of identifications per sample varies according to authors, between 100 (Martin 1955, Sparks & Malechek 1968, Lulich & Hansen 1981) and 400 (Hansen & Dearden 1975, Hansen & Clark 1977). Chapuis (1980) for rabbit and Delaunay (1982) for chamois did not find statistical differences in the diet composition with 200 and 300 epidermis.

Chapuis J.L. 1980. Méthodes d‘étude du régime alimentaire de Lapin de garenne Oryctolagus cuniculus (L.) par l‘analyse micrographique des fèces. Rev. Ecol.-Terre Vie 34, 159-198.

Delaunay, G. 1982. Contribution a la misse au point de méthodes de suivi des populations d'ongulés de haute montagne en milieu protégé: étude sur le chamois dans le Parc National des Écrins. n.º 748, Thèse Univ. Rennes.

García- Gonzalez, R. (1983). Epidermis foliares de algunas especies de Festuca, Poa y Bellardiochloa en el Pirineo occidental. Anales Jardín Botánico de Madrid, 39(2), 389-404.

Hansen, R.M., & Clark, R.C. 1977. Foods of elk and other ungulates at low elevations in northwestern Colorado. J. Wildl. Manage., 41(1), 76-80.

Hansen, R.M., & Dearden, B.L. 1975. Winter foods of mule deer in Piceance Basin, Colorado. J. Range Manage., 28(4), 298-300.

Lucich, G.C., & Hansen, R.M. 1981. Autumn mule deer foods on heavily graced cattle ranges in Northwestern Colorado. J. Range Manage., 34(1), 72-73.

Martin, D.J. 1955. Features of plant cuticles. An aid to the analysis of the natural diet of grazing animals. Trans. Bot. Soc. Edinburgh, 36, 278-288.

Sparks D. R. & Malechek J. C. 1968. Estimating percentage dry-weight in diets using a microscope technique. J. Range Manage. 21: 264-265

Stewart D.R.M. & Stewart J. 1971: Comparative food preferences of five East African ungulates at different seasons. In: Watt E. & Duffey A.S. (eds.), The scientific management of animal and plant communities for conservation. Blackwell Science Publishing, Oxford: 351–366.

Furthermore, the manuscript does not specify the size range of the analysed plant fragments or whether any size-based selection criteria were applied. Fragment size can strongly influence detectability and identification success, and without standardization (e.g. through sieving or defined microscopic fields), there is a risk that certain plant taxa are over- or under-represented. Clarification of these aspects is essential to assess the robustness of the dietary estimates.

We recorded the first 300 fragments identified with certainty, discarding unidentifiable ones. This is an unavoidable bias that is weakly relevant in a comparative study.

The results provide a coherent and valuable overview of the diet composition of Barbary deer and its seasonal variation within the study area. The authors successfully document the consumption of a relatively broad spectrum of plant taxa (22 species) and demonstrate a consistent pattern in which shrubs and trees dominate the diet, complemented by grasses and forbs. The seasonal trends are ecologically plausible and align with expectations for Mediterranean environments, with increased grass consumption in spring and a shift towards browsing during the drier summer and autumn periods. The identification of key plant species contributing to the diet further strengthens the descriptive value of the dataset. Overall, the results suggest a certain degree of trophic flexibility, supporting the classification of Barbary deer as an intermediate feeder capable of adjusting its diet according to seasonal resource availability.

However, the results are presented primarily in a descriptive manner and are not fully exploited analytically. The statistical analysis is generally appropriate for exploratory purposes but remains limited in scope. The study relies mainly on descriptive statistics and multivariate ordination (FCA), without applying inferential tests to assess seasonal differences in diet composition. As a result, it is unclear whether the observed patterns are statistically meaningful or reflect sampling variability. In addition, the explanatory power of the FCA is relatively low and should be more explicitly acknowledged.

Some inferential statistics were added in the revised version, particularly for testing the seasonal variations of the main plant categories. The explanatory power of the FCA was erroneously low as we reported inertia and not percentages of variance which are 58.12% for axe 1 and 29.01% for axe 2. Nevertheless, the main seasonal differences identified by the FCA are linked with poorly consumed plant species, which does not support long exposition.

The absence of measures of variability (e.g. standard deviations or confidence intervals) further limits the interpretability of the results. Furthermore, the interpretation occasionally extends beyond the data, particularly when inferring dietary preferences without information on resource availability.

Most studies of ungulate diet only provide monthly or seasonally grouped data in order to avoid biases due to sampling protocol, which cannot permit to obtain measures of variability.

Overall, the dataset itself is informative and represents a useful contribution to the knowledge of this poorly studied subspecies. However, the statistical treatment does not fully exploit its potential. The inclusion of more robust analytical approaches (e.g. PERMANOVA, GLMs, or methods suitable for compositional data), together with a clearer distinction between descriptive patterns and ecological inference, would substantially strengthen the study and enhance its impact.

We performed a PERMANOVA analysis but the results were identical to the results that the more basic statistics that we added. More, due to the pooled data, GLMs were not relevant.

The discussion would benefit from a stronger integration of established ecological frameworks. In its current form, it remains largely descriptive and comparison-based, without fully embedding the results within broader concepts such as optimal foraging theory, niche theory, or trophic plasticity. While the observed seasonal shifts in diet composition are ecologically plausible, they are not explicitly interpreted in terms of adaptive foraging strategies or niche dynamics. For instance, the apparent switching between grazing and browsing across seasons could be discussed within the framework of optimal foraging theory as a response to changing resource availability and quality. Similarly, the relatively broad spectrum of consumed plant taxa suggests a degree of trophic plasticity that could be more explicitly addressed in relation to niche breadth and flexibility. Even though Barbary deer is a geographically restricted subspecies, these concepts have been widely applied to other cervids and large herbivores, and their inclusion would considerably strengthen the ecological interpretation and general relevance of the study.

On the whole, the Barbary deer's diet varied across seasons in North Algeria. Such behavioural plasticity in feeding niche is usually associated with habitat use and was reported in introduced populations (Ismaili et al 2018 ).Our results illustrate the feeding plasticity of this subspecies of  cervides  which can live from the sea coast in Morocco  ( Cuzin et al., 2017 ; Desmet 1989) to the border limit of Western Tunisiats in  North Africa ( Salez, 1959 ; Kowalski, Rzebik-Kowalska ,1991 ; Burthey et al., 1992 ; Burthey-Mandret,  and Burthey 1997). Feeding mostly on  Shrubs and trees (woody browse)  and secondarily on grasses and forbs all through the year, the North Africain Barbary deer should be considered as an “intermediate feeder” according to Hofmann and Stewart (1972), or more precisely as a “generalist mixed- feeder” according to Gagnon and Chew (2000). Selecting among a few available food items, it can also be classified as an “obligatory generalist” (Shipley et al., 2009), having a wide realised niche within a relatively narrow fundamental niche.

Finally, the discussion would benefit from a more explicit and comprehensive consideration of the study’s limitations. While some methodological constraints of microhistological analysis are mentioned, their implications for data interpretation are not fully explored. In particular, the potential bias associated with differential digestibility of plant tissues and the fact that fragment frequency does not reflect biomass intake should be clearly acknowledged.

This comment supports keeping the first paragraph of the discussion (contrary to a comment of reviewer 1). In this paragraph we list the various biases of the microhistological analysis; technician error, differential digestibility of plant species, low discernibility of epidermal fragments, sample preparations. We added that fragment frequency does not reflect biomass intake.

Indeed, the microhistological method tends to underestimate highly digestible plant species (Brand, 1978) as well as those rarely consumed, while overestimating species that readily fragment into small particles (Johnson et al., 1983). However, its reliability in evaluating herbivore diet composition has long been established (Holechek et al., 1982), and its limitations remain acceptable, especially in comparative studies and in environments with low plant diversity.

In addition, the representativeness of the dataset is not discussed. Issues related to sample size, potential pseudoreplication, and the lack of information on population size should be considered, as they may influence the generality of the results.

Some information on population size has been added which partly answers this comment. Pellet collecting followed previous sampling protocols (see references) to achieve a reasonable compromise between accuracy and cost in a comparative perspective. Burthey-Mandret and Burthey (1997) reported that the diet did not change beyond 10 pellets, a number which reflects the population size avoiding a strong pseudoreplication that cannot be avoided in the absence of genotyping the pellets for identifying the producing individual.

A major limitation is the absence of data on food availability, which prevents interpretation in terms of resource selection. However, even a qualitative discussion of vegetation structure within the enclosure (e.g. relative abundance of woody versus herbaceous species) could provide useful ecological context and strengthen the interpretation of the findings. Addressing these points would lead to a more balanced and transparent discussion and improve the ecological relevance of the study.

We develop the need of evaluating the food availability as a perspective in the conclusion. Unfortunately, the vegetation structure in the enclosure has not been recorded at the time of the pellet collecting, and it is too late to do it. Nevertheless, the main plant species in the enclosure at the time of the study are listed in the material and methods section.

 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I believe the manuscript requires a discussion of dietary diversity. No significant differences in diversity indices were detected between seasons, and this should be highlighted. Furthermore, the discussion focuses on seasonal differences among the main dietary species (e.g., Avena sterilis) or the major groups (woody plants, grasses, forbs). However, no statistical analysis has been performed to demonstrate that these differences are significant. It is important to conduct some statistical test, such as analysis of variance or Dirichlet regression for positional data, to be able to discuss the differences between seasons.

Author Response

I believe the manuscript requires a discussion of dietary diversity. No significant differences in diversity indices were detected between seasons, and this should be highlighted.

A statistical test has been added.

Furthermore, the discussion focuses on seasonal differences among the main dietary species (e.g., Avena sterilis) or the major groups (woody plants, grasses, forbs). However, no statistical analysis has been performed to demonstrate that these differences are significant. It is important to conduct some statistical test, such as analysis of variance or Dirichlet regression for positional data, to be able to discuss the differences between seasons.

For seasonal differnces among the main dietary species, namely Avena sterilis and Carex sp, we added statistical tests which showed a significant lower consumption in summer for the first one, and a significant higher consumption in spring for the second. For differences among the major groups we already added: “the seasonal consumption of the three plant categories revealed some significant differences. If the consumption of shrubs and trees was quite similar among seasons (chi-square = 4.59, DF = 3, p > 0.05), variations were significant for forbs (chi-square = 12.15, DF = 3, p < 0.01), and highly significant for grasses (chi-square = 15.81, DF = 3, p < 0.001)”.

Author Response File: Author Response.doc

Reviewer 3 Report

Comments and Suggestions for Authors

I appreciate the substantial improvements made in the revised version of the manuscript. In particular, the clarification of population size, the expanded description of the microhistological procedure, and the inclusion of basic inferential statistics have significantly strengthened the methodological transparency of the study.

However, some concerns remain only partially addressed. The use of pooled samples at the seasonal level prevents any assessment of within-season variability and limits the strength of statistical inference. While this approach may be acceptable for descriptive purposes, it should be more explicitly acknowledged as a limitation when interpreting the results.

In addition, the procedure for fragment selection (i.e. identification of the first 300 fragments) is now clearer, but the absence of any standardization with respect to fragment size or sampling effort across slides still raises concerns about potential bias in taxon representation.

Most importantly, the lack of data on food availability continues to constrain ecological interpretation. In its current form, the manuscript occasionally implies dietary preferences or resource selection, which cannot be robustly inferred without information on resource availability. I recommend that the authors consistently restrict their interpretation to diet composition and avoid overinterpretation in terms of selection.

Finally, although some elements of ecological theory have been introduced, the discussion would still benefit from a more explicit integration of established frameworks (e.g. optimal foraging, niche theory, trophic plasticity), which would strengthen the broader relevance of the study.

Author Response

I appreciate the substantial improvements made in the revised version of the manuscript. In particular, the clarification of population size, the expanded description of the microhistological procedure, and the inclusion of basic inferential statistics have significantly strengthened the methodological transparency of the study.

However, some concerns remain only partially addressed. The use of pooled samples at the seasonal level prevents any assessment of within-season variability and limits the strength of statistical inference. While this approach may be acceptable for descriptive purposes, it should be more explicitly acknowledged as a limitation when interpreting the results.

We fully agree that it should have been better to be able to investigate the within season variability. This is acknowledged in the revised version of the manuscript.

In addition, the procedure for fragment selection (i.e. identification of the first 300 fragments) is now clearer, but the absence of any standardization with respect to fragment size or sampling effort across slides still raises concerns about potential bias in taxon representation.

In the text we write “we identified the first 300 epidermal fragments from 30 pooled pellets per season”. There is no potential bias due to fragment size or sampling effort across slides. Such method has been published previously for Aoudad (Bounaceur et al. 2023 - Feeding ecology of the vulnerable aoudad (Ammotragus lervia) in north‐western Sahara. Afr. J. Ecol., 61(1): 28-36.

Most importantly, the lack of data on food availability continues to constrain ecological interpretation. In its current form, the manuscript occasionally implies dietary preferences or resource selection, which cannot be robustly inferred without information on resource availability. I recommend that the authors consistently restrict their interpretation to diet composition and avoid overinterpretation in terms of selection.

We carefully checked the discussion in order to avoid any overinterpretation in terms of selection, and to express our results in terms of consumption.

Finally, although some elements of ecological theory have been introduced, the discussion would still benefit from a more explicit integration of established frameworks (e.g. optimal foraging, niche theory, trophic plasticity), which would strengthen the broader relevance of the study.

The framework of this study was identifying / confirming the place of Barbary red deer within the gradient browser / grazer of foraging ungulates in an unusual semi-arid habitat. Our data cannot support optimal foraging as we did not record food availability and quality. For testing trophic plasticity, we should investigate several populations in quite similar habitats as we did for gundis (Lasgaa et al. 2021, 2023).

Author Response File: Author Response.doc

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

Lines 126 -128: In my opinion, the diversity index varies little between seasons. Furthermore, species richness is also similar. Therefore, the discussion should point out that the animals' diet maintains diversity and richness throughout the year, even though the botanical composition varies. This is a unique and noteworthy finding of the study.

Author Response

Lines 126 -128: In my opinion, the diversity index varies little between seasons. Furthermore, species richness is also similar. Therefore, the discussion should point out that the animals' diet maintains diversity and richness throughout the year, even though the botanical composition varies. This is a unique and noteworthy finding of the study.

We didn’t realize that this result was noteworthy. In the bibliography we didn’t find any study giving the list of plant species consumed by red deer during the four seasons. This is a benefit of using a microhistological analysis. So, we added a new paragraph before discussing the diet of each season:

In Akfadou Forest enclosure the diet of Barbary deer exhibited a similar diversity throughout the year, even though the botanical composition varied. Such result was previously reported in Algeria [20], and Tunisia [16]. Unfortunately, none of the published studies from the Mediterranean region [47-51] provided either a list of consumed species or even species richness in diet for comparison. More widely in Europe, as most analyses rely on rumens from hunted animals, such information is only available for autumn and winter times [52].

Author Response File: Author Response.docx