Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1. Introduction

• The introduction provides a solid background and justification for the study. To strengthen it further, consider briefly expanding on the ecological importance of prey diversity for Lesser Kestrel reintroduction success. Linking diet to potential fitness or breeding outcomes could enhance the reader’s appreciation of the study’s applied significance.

• Although references are appropriate and current, a concise synthesis of diet studies from adjacent regions (e.g., Balkans, southern Europe) could provide a broader comparative framework, emphasizing what is unique to the Bulgarian context.

2. Materials and Methods

• The description of pellet collection and identification protocols is clear and detailed. Adding a short note on how ambiguous or unidentifiable prey remains were treated (excluded, grouped as “Insecta indet.”, etc.) would improve methodological transparency.

• Consider briefly discussing any potential limitations of the pellet analysis method, such as biases in prey detectability (e.g., soft-bodied prey) or temporal sampling constraints, to contextualize the results.

• Providing GPS coordinates or more detailed habitat descriptions for both colonies would facilitate reproducibility and comparisons by future researchers.

• Clarify if any measures were taken to avoid repeated sampling of pellets from the same individuals, which could affect the independence of samples.

3. Results and Discussion

• The results are clearly presented and adequately supported by figures. However, integrating some basic statistical comparisons (e.g., tests of proportional differences between colonies) could strengthen claims of differences or similarities in prey composition rather than relying solely on descriptive indices.

• The discussion of plastic ingestion at the Stara Zagora site is important. Expanding on potential ecological or physiological impacts on Lesser Kestrels, possibly referencing comparable studies of plastic ingestion in raptors or farmland birds, would highlight the conservation implications.

• The presence of millipedes and annelids is interesting. Could the authors speculate on whether these represent accidental ingestion, alternative food sources during scarcity, or behavioral novelty? This could stimulate further research questions.

• Some parts of the discussion could better integrate the quantitative dietary niche and overlap indices with ecological interpretation. For example, what does a high Pianka’s overlap imply about resource competition or habitat use between these colonies?

4. Figures and Tables

• The figures, as described, are informative and well-labeled. To improve clarity, consider adding error bars or measures of variability where relevant (e.g., prey proportion confidence intervals if possible).

• If figure captions currently only describe content, expanding them to briefly state the main message or finding illustrated would enhance reader comprehension.

• Consider including a summary table comparing key ecological and habitat parameters of the two colonies (location, habitat types, sample sizes, niche breadth values, main prey categories) for quick reference.

5. Language and Style

• The manuscript is generally well written in clear academic English. Minor editorial polishing to improve flow and consistency—especially avoiding redundancy in describing insect groups and prey categories—would enhance readability.

• A careful proofread for typographical errors, punctuation, and formatting (especially in-gaps between sections) is recommended to meet publishing standards.

6. Additional Suggestions

• If available, consider including a brief note on future research directions, such as long-term dietary monitoring, assessment of prey availability in habitats, or analysis of breeding success relative to diet composition.

• Clarify data accessibility (raw pellet data, photographs) for transparency and potential reuse.

• A short statement confirming compliance with any ethical guidelines or permits relevant to fieldwork (even if non-invasive) would be appropriate for completeness.

Overall, this manuscript presents valuable original data on Lesser Kestrel trophic ecology that substantially contributes to regional conservation knowledge. Implementing these refinements would improve methodological clarity, ecological interpretation, and presentation quality, facilitating acceptance and impact.

Author Response

Thank you for the thorough revision, we have addressed all comments and made revisions accordingly, to the best of our abilities.

Please find attached the point-by-point reply.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

Studies of kestrel diets are common because the species is popular under management and recovery plans throughout Europe, and because pellets can be collected near nest boxes. The abundance of such studies does not undermine the validity of the present work; rather, it illustrates the collective effort to characterize the geographic and temporal variability of this species. As the authors are undoubtedly aware, studies of the actual diet are significant when compared with the potential diet: the former is found in pellets, while the latter is available in the vicinity of the nest boxes. Given the current state of our knowledge of kestrel diets, it is essential to measure both in order to advance our understanding of this species and, above all, to ensure the viability and success of recovery plans. However, the present study only shows the frequency of solid remains. Without measuring the availability of the prey ingested, it is impossible to answer the question of the extent to which the results can impact the management of the species. The management of kestrels is one of the main aims of LIFE projects.

The study also does not show the biomass of solid remains, which is relevant information for balancing those groups with high frequencies of remains but small sizes. This absence should be explained and, if necessary, corrected. The biomass of the solid remains prior to ingestion should also be discussed. Equivalency tables can be used to calculate the size (weight and/or energy) of prey ingested based on the length of the remains. Despite the degree of speculation involved in these calculations, it is possible to conclude that a group containing a large number of remains does not necessarily represent a main food type.

It is important to ensure that the results are presented correctly, both in the text and in the figures. The graphical representation chosen to show the proportions of the remains must allow for visual comparison between the study sites. Pie charts that are distorted by perspective and use different colors are inadequate. A bar chart would be a more suitable option. It also allows the two zones of each type of remains to be paired. Unlike a slice of a pie, each bar should also include a confidence interval for the proportion/percentage. The uncertainty of a proportion/percentage depends on the number of cases and the value of that proportion/percentage. Note that the sizes of the upper confidence limit (UCL 95%) and bottom confidence limit (BCL 95%) are different, so approximating a normal distribution is not helpful. The BCL is always smaller than the UCL if the proportion or percentage is small (less than 0.1 or 10%) and always larger if the proportion or percentage is large (more than 0.9 or 90%). Additionally, the upper and lower intervals never include zero or one (100%), respectively. These intervals can be calculated using many statistical calculators (e.g. http://www.vassarstats.net/).

Adding the asymmetric BLC and UCL boundaries to the columns makes it possible to visually identify whether the proportion of each type of solid debris differed between the two study areas without the need to calculate a statistical test. The bar representation can be changed to points, one per each remains type and study area. The scale of the vertical axis should be set to [0,1] or [0/100%], and the horizontal axis should include all solid debris types that occurred in at least one zone. The axes can be interchanged to facilitate the reading of the combined figure, so that the vertical bars become horizontal bars, as well as the BCL and UCL. In the case of using points instead of bars, it is the UCLs and BCLs that keep changing to a horizontal orientation. The new figure is simple to construct, but even simpler is to change Figures 2 and 3 to a Table 1. This Table 1 contains the number of items of each type of solid debris in both zones, the proportions with respect to the total number of solid debris, as well as the BCLs and UCLs. The usefulness of the new Table 1 is maximal in enabling future revisions of kestrel diet studies. In fact, it is standard practice to include a supplementary table in diet studies, showing the frequency and biomass of the identified remains, categorized by type and study area. It is unusual for this manuscript not to follow this standard practice. The solution is obvious: you must add a descriptive table showing the frequencies of the remains (and, if possible, the biomasses). Include the proportions/percentages that allow comparison between zones, and therefore add the UCL and BCL of each type of solid debris. Admittedly, these limits can be calculated from the sample sizes. If they are included in the table, it is not necessary to add them separately. However, if they plot them rather than tabulating them, it is unavoidable to include UCL/BCL in the combined new Figure 1. The reason for not including the table of frequencies must be explained.

The discussion should comment on the results in relation to studies of kestrel diets in other countries. While some references to previous studies are cited in the introduction, these are not included in the discussion. This is another area for improvement. The discussion should also comment on the effect of the difference in sample collection dates in each study area, which were carried out years apart.

Minor comments

Page 2, line 58. Delete the words 'totally different contemporary'. Without food availability data, it is questionable whether they were really that different. Regarding 'contemporary', this word should be placed at the beginning of the sentence: '... were simultaneously collected...', '... were contemporaneously collected...', etc. However, lines 78–81 state that the sample collection was not simultaneous. In one study area, fieldwork was carried out for almost three weeks in late May 2025, and in the other area, it was carried out for almost three months (February to April) in 2022 and 2023. Therefore, the term 'contemporary' is confusing in the context of the fieldwork.

P. 2, l. 72. Typo: country

P. 2, l. 78. The collection of pellets was completed on 3 June 2025. There has been little time to carry out the laboratory work and prepare the manuscript. Check the dates. Perhaps the collection was completed earlier.

P. 3, l. 97. The percentages are shown to two decimal places, which is excessive when there are only a hundred cases or fewer. I recommend using a single decimal place.

P. 3, l. 93–102. These ten lines should be deleted and a new Table added (see main comments). The same comment applies to lines 105–110 and 131–144.

P. 3, l. 103–104. While there may not be any other studies in this country identifying birds as one of the most important prey groups, there could be in other countries. The manuscript should explain why these results cannot be compared with those obtained in other countries. If the journal had a local distribution area, it would be understandable for the preliminary results to refer only to that country. However, since this is a worldwide journal, the manuscript must explain why the results are not being compared with those obtained in studies performed in other countries.

P. 3, l. 107–110. In the new Table of results within the Insecta category, include as many rows as there are orders that you have been able to identify. Add another row to accommodate the remains of unidentified insects.

P. 3, lines 103–104. In the Methods section, please explain the meaning you assign to the expression 'most important component'. The manuscript reports the number of solid remains, despite the fact that birds are particularly fragile, which may have resulted in a greater quantity of remains. The biomass of the remains (mean and sum total) could either support or refute your assertion. Without the dry biomass estimate, I fear that mere frequency may not be enough. Bear in mind that the frequency of beetle thorax would represent number of individuals, whereas the remains of bird bones would likely represent one individual. This aspect should be explained in more detail in the Material and Methods section. Lines 86–89 do not explain this key aspect: 'All prey items were grouped into broad taxonomic categories to standardize comparison (Annelida, Insecta, Diplopoda, Aves, Mammalia), and the proportions were calculated based on the total number of identifiable items in the analyzed pellets for each colony.'

P. 4, l. 156–159. A statistical test should be included to compare the Levins' index between zones. Without statistical support, it is impossible to conclude that the index was 'relatively similar' in both zones.

P. 4, l. 162: What does the expression 'despite differences in prey abundance and availability' mean? The study did not measure prey availability. Perhaps it did, but those results are excluded from the manuscript. This phrase is key to the whole study, so the authors must explain whether prey availability was measured, even if the results are not included in this manuscript. If so, they must also explain the reasons for its exclusion.

P. 4, l. 163. Figure 1. The main text does not cite this figure. Add a citation for this image.

P. 5, l. 164–167: Figures 2 and 3 show pie charts illustrating the proportion of seven prey groups in pellets collected at Stara Zagora (Fig. 2) and Sakar (Fig. 3). Pie charts are generally discouraged, especially when they are distorted by perspective, as is the case with these two charts. As stated in the main comment, it is very difficult to make a visual comparison of the two graphs because the color of the same prey group differs between the two graphs (e.g. Coleoptera is green in Fig. 2 and reddish in Fig. 3). Do not change the color of groups appearing in both figures.

P. 5, l. 168. Figure 4. The main text does not cite this figure, but it should.

P. 6, l. 173: Information on access to the data needs to be added.

P. 6, l. 179. Review the editorial guidelines regarding the style of references and in-text citations. References are arranged in citation order with a number, but this number is absent from the main text. Another citation style (author, year) has been used instead. Use the same citation system in the main text and in the reference list.

P. 6, l. 184. There is a typographical error: reference 3 is not the continuation of reference 2. Join both lines and revise the numbering of the references in the manuscript text.

Author Response

Thank you for the thorough revision, we have addressed all comments and made revisions accordingly, to the best of our abilities.

P. 2, l. 78. The collection of pellets was completed on 3 June 2025. There has been little time to carry out the laboratory work and prepare the manuscript. Check the dates. Perhaps the collection was completed earlier. Response: The dates are correct, the team has enough experience to determine the pellets within these deadlines, and their number is also not large.

Please find attached the point-by-point reply.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

To enable comparison, all prey items were grouped into broad taxonomic categories. For example: Mammalia, Aves, Diplopoda, Insecta and Annelida (see Fig. 2). Unidentifiable fragments were excluded from the study or categorised at the lowest possible taxonomic level (e.g. 'Insecta indet.', line 185, page 5). Therefore, the total number of items included in this study is the sum of the identified and unidentified items. One notable example is the class Insecta in the Sakar study area. Thirty-three insect fragments were assigned to Coleoptera (n = 17), Orthoptera (n = 1), Dermaptera (n = 2), and unknown insects (n = 13). This accounts for 68.8% (56.0–81.5%, BCL95%-UCL95%) of the total Insecta fragments (n = 48, p. 5, l. 179). This percentage is not different from that obtained in Stara: 76.2% (67.1–83.5%, BCL95%-UCL95%). However, the results shown in Figure 2 for Insecta differ from these. This is because 13 insects whose order, family or genus could not be identified were excluded. I consider excluding them is a flaw, given that it is accepted that they belonged to the Class Insecta. When they are properly included in the calculations, the Insecta percentages in both study areas are close to each other, which makes the study's conclusion more reasonable. After all, insects are the main food source for kestrels in this study, as well as in the other cited studies. Figure 2 and other calculations in the manuscript should be updated.

 

A file is included in this review to show how the Figure 2 might appear once revised. In this example, some frequencies could not be included because they are not found in the main text. This illustrates the usefulness of including a table showing the distribution of the 101 and 48 remains (Stara and Sakar, respectively) by class, order, family, etc. Tabulating the frequencies is very important. This table also highlights the absence of frequencies in some cells. Ultimately, this table would render Figure 2 redundant.

 

Contrary to what is stated in the manuscript (p. 3, line 112), unidentifiable fragments must be included in the table. This is particularly important if they cannot be associated with any class.  A main category that includes any biological fragment not identified in a class must appear at the end of the table. It is crucial not to exclude this category from the study, as it is used to calculate the percentage of the diet that has been identified. A diet based on 5% identified fragments is not as credible as one based on 85%. To assess the credibility of the study, the total number of fragments must be reported, as well as the number that could not be identified taxonomically.

 

 

Comments for author File: Comments.pdf

Author Response

Comment 1: To enable comparison, all prey items were grouped into broad taxonomic categories. For example: Mammalia, Aves, Diplopoda, Insecta and Annelida (see Fig. 2). Unidentifiable fragments were excluded from the study or categorised at the lowest possible taxonomic level (e.g. 'Insecta indet.', line 185, page 5). Therefore, the total number of items included in this study is the sum of the identified and unidentified items. One notable example is the class Insecta in the Sakar study area. Thirty-three insect fragments were assigned to Coleoptera (n = 17), Orthoptera (n = 1), Dermaptera (n = 2), and unknown insects (n = 13). This accounts for 68.8% (56.0–81.5%, BCL95%-UCL95%) of the total Insecta fragments (n = 48, p. 5, l. 179). This percentage is not different from that obtained in Stara: 76.2% (67.1–83.5%, BCL95%-UCL95%). However, the results shown in Figure 2 for Insecta differ from these. This is because 13 insects whose order, family or genus could not be identified were excluded. I consider excluding them is a flaw, given that it is accepted that they belonged to the Class Insecta. When they are properly included in the calculations, the Insecta percentages in both study areas are close to each other, which makes the study's conclusion more reasonable. After all, insects are the main food source for kestrels in this study, as well as in the other cited studies. Figure 2 and other calculations in the manuscript should be updated.

Response 1: The corresponding summary data table has been added. Figure 2 has been replaced with Table 1. Only the nylon items have been excluded, as they are not related to food and are not comparable to the other items.

 

Comment 2: A file is included in this review to show how the Figure 2 might appear once revised. In this example, some frequencies could not be included because they are not found in the main text. This illustrates the usefulness of including a table showing the distribution of the 101 and 48 remains (Stara and Sakar, respectively) by class, order, family, etc. Tabulating the frequencies is very important. This table also highlights the absence of frequencies in some cells. Ultimately, this table would render Figure 2 redundant.

Response 2: Done. Fig. 2 is replaced with Table 1.

 

Comment 3: Contrary to what is stated in the manuscript (p. 3, line 112), unidentifiable fragments must be included in the table. This is particularly important if they cannot be associated with any class.  A main category that includes any biological fragment not identified in a class must appear at the end of the table. It is crucial not to exclude this category from the study, as it is used to calculate the percentage of the diet that has been identified. A diet based on 5% identified fragments is not as credible as one based on 85%. To assess the credibility of the study, the total number of fragments must be reported, as well as the number that could not be identified taxonomically.

Response 3: All fragments have been taxonomically identified to the appropriate level and included in the table. There are no biological fragments that remain unidentified.Only the nylon items have been excluded, as they are not related to food and are not comparable to the other items.

Round 3

Reviewer 2 Report

Comments and Suggestions for Authors

I have a couple of small comments on the third version of the manuscript. According to response 3, all fragments isolated in the analysed pellets have been taxonomically identified and included in the table. However, line 112 on page 3 states that fragments that could not be identified were excluded from the analysis. This contradiction needs to be addressed. As this appears to be a misunderstanding that arose during the review process, I recommend adding the following statement to the main text: 'No biological fragments from the pellets remained unidentified.' I commend the authors, as I have never been able to identify 100% of biological fragments.

Page 4 contains a paragraph distinguishing insects from hymenopterans (lines 145–146):
 “The identified prey items included representatives of six major groups: insects, annelids, mammals, birds, and hymenopterans, along with non-food synthetic material.”
As the Hymenoptera is just one of the orders within the class Insecta, it would be advisable not to list them as separate representative groups. The table satisfactorily presents the classification system: Hymenopteran fragments are included in the Insecta class.

Author Response

Comment 1: I have a couple of small comments on the third version of the manuscript. According to response 3, all fragments isolated in the analysed pellets have been taxonomically identified and included in the table. However, line 112 on page 3 states that fragments that could not be identified were excluded from the analysis. This contradiction needs to be addressed. As this appears to be a misunderstanding that arose during the review process, I recommend adding the following statement to the main text: 'No biological fragments from the pellets remained unidentified.' I commend the authors, as I have never been able to identify 100% of biological fragments.

Response 1: Thank you, we agree, the passage is revised accordingly, and highlighted.

Comment 2: Page 4 contains a paragraph distinguishing insects from hymenopterans (lines 145–146):
 “The identified prey items included representatives of six major groups: insects, annelids, mammals, birds, and hymenopterans, along with non-food synthetic material.”
As the Hymenoptera is just one of the orders within the class Insecta, it would be advisable not to list them as separate representative groups. The table satisfactorily presents the classification system: Hymenopteran fragments are included in the Insecta class.

Response 2: Thank you, we agree, the passage is revised accordingly, and highlighted.