Wintering Red Kites in Central Spain: Macrohabitat Selection and Population Density Estimate

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Summary:

In this well-written manuscript the authors aim to determine the size of the wintering population of Red Kites in central Spain where the species has not previously been considered to be abundant.  Using an established survey protocol over a two-year period, the authors firstly determine the importance of this region for wintering Red Kites and, secondly, identify a preference in habitat use more commonly found in lowland agricultural landscapes and mosaics. This latter point is a useful emphasis, especially when considering the extent and importance of land used for agricultural production across Europe and the potential implications for this species.

 

General comments:

I have only two general comments relating to the ms.

1. Although the manuscript title includes the phrase '...Role of Roads and Tracks in Detection' this part of the manuscript is fairly limited in extent and is perhaps slightly overstated by phrasing it in the title in this way.  In its current form the title reads as though there could be an assessment of how detections from roads differ from detections across the landscape from randomised routes (i.e. away from roads). 

While the authors do explore the differences in detections between paved and unpaved roads, this surely relates to detections from each of these road/track types and not on them.  Any difference in detectability might be more meaningfully assessed by examining the distribution (or means) of detection distances of birds from each of the road types, since birds seen at a perpendicular distance of many hundreds of metres could reasonably be assumed not to have been attracted to the road. Smaller mean detection distances to paved roads may indicate a level of attraction. Inclusion of this information would be useful for the reader.

2. The ms would benefit greatly from the inclusion of a map. Most importantly it provides the reader with the context of location if they are unfamiliar with the geography of Spain. As an example, I was unfamiliar with the MHBR and was confused by the point raised re. abundance of overwintering Red Kites in Castilla León and Extremadura  - the text can be interpreted as meaning that the MHBR is very close to, or within, one of these areas. A map would immediately remove such confusion.

 

Specific Comments:

Keywords - perhaps include 'distance sampling'

 

Line 38 - and throughout.  I'm not sure of the need to include the word 'itineraries' throughout the ms. Given that it is defined as being a line transect count (at line 104), why not continue to use the term 'transect'? This would remove the need for any definition (which appears after already being used in the abstract).  If use of this word is retained, then the location of Table 1. should be moved to a point after line 103 (where 'itineraries' is defined).

Lines 104-105. and Discussion. Perhaps Buckland et al. 2007 and Marques 2007 should also be referenced when discussing the implications of violating this important assumption of distance sampling i.e. using non-randomly located transects (along linear landscape features). Further treatment is given in Marques et al. 2010.

Lines 109-113.  While the habitat composition along the transects is given, it is used primarily for analysis of dissimilarity between the two general regions. Is it possible to use this data to explore the impact on detectability? There is no indication of the vegetation lining the roads along the transect routes (in the UK, and other countries, most roads would be bounded by a hedge).  Although the description of the study area gives the vague suggestion that this is likely to be fairly open, this is not explicitly stated. Given the nature of the process (and the assumptions within distance sampling), any restriction of view as a result of dense roadside vegetation (especially if it stretches across a large area) might have implications re. detectability. A brief indication of whether or not this was a factor would remove this question. 

Line 124. 'Telemeter' is more frequently referred to as a 'laser rangefinder' in English.

Line 162. It may be useful for someone wishing to replicate this work if the authors included the average time taken per km across all of the itineraries.  Was there a difference in speed of the transects undertaken on paved versus unpaved roads? 

Line 228. If the data are available, some information relating to the behaviours or locations of individuals detected during the study would be useful - both of which may impact detectability. It would be informative to know if the majority of observations were of perched birds or those in flight.

 

Tables

As a general point, using words rather than abbreviations for each column in the tables (where space allows) would improve the readability. This would also reduce the space taken up by table captions.

 

References

Buckland, S. T., Borchers, D. L., Johnston, A., Henrys, P. A., and Marques, T. A. (2007). Line transect methods for plant surveys. Biometrics 63, 989-998.

Marques, T. A. (2007). Incorporating measurement error and density gradients in distance
sampling surveys. PhD thesis, University of St. Andrews.

Marques, T. A., Buckland, S. T., Borchers, D. L., Tosh, D., & McDonald, R. A. (2010). Point transect sampling along linear features. Biometrics, 66(4), 1247-1255.

Comments on the Quality of English Language

Generally, the English is very good and does not require any improvement except for one sentence at lines 183-184: Presumably the meaning here is something along the lines of 'no significant differences in presence/absence within localities were identified between winters.' 

Author Response

General comments:

1. Although the manuscript title includes the phrase '...Role of Roads and Tracks in Detection' this part of the manuscript is fairly limited in extent and is perhaps slightly overstated by phrasing it in the title in this way.  In its current form the title reads as though there could be an assessment of how detections from roads differ from detections across the landscape from randomised routes (i.e. away from roads). 

We have deleted the mention of the role of roads from the title as suggested. The current title, Wintering Red Kites in Central Spain: Macrohabitat Selection and Population Density Estimate, more accurately reflects the content of the paper.

While the authors do explore the differences in detections between paved and unpaved roads, this surely relates to detections from each of these road/track types and not on them.  Any difference in detectability might be more meaningfully assessed by examining the distribution (or means) of detection distances of birds from each of the road types, since birds seen at a perpendicular distance of many hundreds of metres could reasonably be assumed not to have been attracted to the road. Smaller mean detection distances to paved roads may indicate a level of attraction. Inclusion of this information would be useful for the reader.

We compared detection distances between paved roads and tracks and found no significant differences. This information has been included in both the Methods and Results sections.

“Detection distance between roads (n=52) and tracks (n=72) did not differ significantly (Road = 97.3±2.4 m, Track = 109.0±17.7, Mann-Whitney U = 0.74, p = 0.460).”

2. The ms would benefit greatly from the inclusion of a map. Most importantly it provides the reader with the context of location if they are unfamiliar with the geography of Spain. As an example, I was unfamiliar with the MHBR and was confused by the point raised re. abundance of overwintering Red Kites in Castilla León and Extremadura  - the text can be interpreted as meaning that the MHBR is very close to, or within, one of these areas. A map would immediately remove such confusion.

We agree with your comment and have added a map (see Figure 1), along with the corresponding caption, to help the reader locate the study area in Spain. This addition should enhance the clarity and contextualization of the study.

Specific Comments:

Keywords - perhaps include 'distance sampling'

We have added “distance sampling” to the list of keywords as suggested to clarity the methodological approach used in our study. 

Line 38 - and throughout.  I'm not sure of the need to include the word 'itineraries' throughout the ms. Given that it is defined as being a line transect count (at line 104), why not continue to use the term 'transect'? This would remove the need for any definition (which appears after already being used in the abstract).  If use of this word is retained, then the location of Table 1. should be moved to a point after line 103 (where 'itineraries' is defined).

We have replaced every instance of the term “itinerary” with “transect,” as suggested, to better align with standard ecological terminology.

Lines 104-105. and Discussion. Perhaps Buckland et al. 2007 and Marques 2007 should also be referenced when discussing the implications of violating this important assumption of distance sampling i.e. using non-randomly located transects (along linear landscape features). Further treatment is given in Marques et al. 2010.

We have stated the assumptions of distance sampling in the text based on Buckland et al. (2015), which is the most recent and comprehensive reference available. Furthermore, Marques, a co-author of the reference book, contributes to the methodological foundation we applied. We have also acknowledged the limitation of sampling along roads and tracks both in the Methods section and in the Limitations part of the Discussion, as recommended (see Editor response to a similar comment).

Lines 109-113.  While the habitat composition along the transects is given, it is used primarily for analysis of dissimilarity between the two general regions. Is it possible to use this data to explore the impact on detectability? There is no indication of the vegetation lining the roads along the transect routes (in the UK, and other countries, most roads would be bounded by a hedge).  Although the description of the study area gives the vague suggestion that this is likely to be fairly open, this is not explicitly stated. Given the nature of the process (and the assumptions within distance sampling), any restriction of view as a result of dense roadside vegetation (especially if it stretches across a large area) might have implications re. detectability. A brief indication of whether or not this was a factor would remove this question.

Thank you for the insightful comment. The vegetation along the transect routes in our study areas is predominantly open habitat, with no dense roadside vegetation such as hedges commonly found in other regions like the UK. This open habitat condition around all transects ensured good visibility, which did not limit detectability during surveys. We have now explicitly stated this in the manuscript to clarify that roadside vegetation did not pose a restriction to animal detection, thus supporting the assumptions of distance sampling in our context.

It was included into the methods section (2.2. Red Kite sampling): Vegetation along transect routes was predominantly open, without dense roadside hedges, ensuring good visibility and no limitation to detectability during surveys.

Line 124. 'Telemeter' is more frequently referred to as a 'laser rangefinder' in English.

Thank you for the suggestion; we have modified the term accordingly in the manuscript.

Line 162. It may be useful for someone wishing to replicate this work if the authors included the average time taken per km across all of the itineraries.  Was there a difference in speed of the transects undertaken on paved versus unpaved roads?

Please refer to our response to the Editor, as the same notation was raised, and we have addressed it comprehensively there.

Line 228. If the data are available, some information relating to the behaviours or locations of individuals detected during the study would be useful - both of which may impact detectability. It would be informative to know if the majority of observations were of perched birds or those in flight.

Almost 50% of the detected kites were perched (60 out of 124 individuals). We have included this information in the text (first line of results) and discussion.

Tables

As a general point, using words rather than abbreviations for each column in the tables (where space allows) would improve the readability. This would also reduce the space taken up by table captions.

We changed it accordingly as suggested by the Editor. 

References

Buckland, S. T., Borchers, D. L., Johnston, A., Henrys, P. A., and Marques, T. A. (2007). Line transect methods for plant surveys. Biometrics 63, 989-998.

Marques, T. A. (2007). Incorporating measurement error and density gradients in distance sampling surveys. PhD thesis, University of St. Andrews.

Marques, T. A., Buckland, S. T., Borchers, D. L., Tosh, D., & McDonald, R. A. (2010). Point transect sampling along linear features. Biometrics, 66(4), 1247-1255.

Thank you for suggesting these references. However, we prefer to use Buckland et al. (2015) as it comprehensively contains all relevant information consolidated in a single reference.

Comments on the Quality of English Language

Generally, the English is very good and does not require any improvement except for one sentence at lines 183-184: Presumably the meaning here is something along the lines of 'no significant differences in presence/absence within localities were identified between winters.' 

Thank you very much for your kind words; as we are not native speakers. We have replaced the sentence with your suggestion, which is clearer.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear authors and editors,

The manuscript entitled ‘Wintering Red Kites in Central Spain: Habitat Selection and the Role of Roads and Tracks in Detection’ is a small but valuable contribution to the knowledge of an endangered species within Iberia. I had the opportunity to read the entire manuscript and I consider that it falls within the scope and aims of Birds since the results of the manuscript are interesting for Red Kite specialists or local researchers. However, I have both major and minor comments to improve the manuscript. I hope they help authours to revise their manuscript.

 

MAJOR COMMENTS

 

• Simple summary. I would include a statement about the importance of Iberia (or specifically La Mancha Húmeda) as wintering grounds of the Red Kites. Does the Iberian Peninsula regions host an important part of the European population during winter?

 

• Simple summary. Similarly, after you stated that “the reserve hosts a wintering population of around 1,400 red kites”, could you provide a simple estimation of which proportion of Iberian wintering red kites are those 1,400 based on estimated Iberian wintering population? (see for example https://seo.org/ave/milano-real/#:~:text=El%20área%20de%20distribución%20de,la%20invernada%20del%202013-2014.)

 

• Simple summary. Lines 26-27 and Abstract (L43-45). Are all agricultural landscapes equally key to supporting wintering Red Kites? Including super-intensive agricultural lands? Is it enough only to “protect” those areas or habitat improvements are also necessary to enable the colonization and expansion of raptors with preference for open and heterogeneous landscapes which incorporates agricultural lands within?

 

• Section 2.2. I think the manuscript would greatly improve if the authors provided some geographical information on the sampling protocol. For example, I think a map of the study area is necessary; for instance a map from Europe/Iberia to locate the study area and a smaller scale map of La Mancha Húmeda showing the diversity of habitats of the area. Could you also provide a figure with the location of the line transects? I think it would be useful to understand the distribution of the sampling effort in order to assess the degree of coverage of the study area.

 

• Section 2.2. With your approach, densities were estimated at the level of locality or transect? Could you estimate kites density for each specific transect? If so, why did you compare non-corrected abundance for some of the analysis if you could use non-biased density? Please, clarify.

 

• Section 2.2. I have concerns about the methodological approach to analyse these data. The authors used a wide variety of parametric and non-parametric test —most of them essentially specific cases of linear models— which mostly indicated if there were differences among levels of some factors (localities, type of road, winters, itinerary length…). The problem I found is that some of the variables may be interacting with other variables or at least modulating them. If all variables were incorporated into the same model, it would be possible to assess the influence of explanatory variables in response variables controlling the variability explained by all of them, as well as to assess the effect size and its confidence intervals. In this regard, I think that authours could use simple Generalized Linear (Mixed) Models to get a more parsimonious approach. For example:

 

Presence/absence data // Density ~ Type of road + Locality + Itinerary length + Winter + Type of road*Locality…

 

...or similar models.

 

• Section 2.3. Why did you only consider 100-meters intervals to characterize habitat types? You surveyed kites up to 1000 perpendicular meters, so why did you not quantified habitat types until that distance?

 

• Section 2.3. Did you consider to use not only the surface area of different habitat types but a Heterogeneity Index of landscape composition? Red Kites may select one habitat or another at small scale, but they may also select heterogeneous landscapes at broader scales, for instance composed by flat agricultural lands, shrubs areas, wood croops, semiurban areas, etc.

 

 

• I detected a lack of references of studies of the Red Kite within Spain. The Spanish population of the species (including breeding habitat selection, wintering movement strategies, migratory patterns, juvenile dispersal, post-reproductive strategies selection, etc.) have been studied just a few years ago, greatly increasing the knowledge of the species. The authours cited some papers from Germany, UK and a few from Spain, but I think they should incorporate the most recent findings on the spatial ecology of the species (when it was necessary). For example:

 

• García-Macía, J., De La Puente, J., Bermejo, A., Raab, R., Urios, V. (2022). High Variability and Dual Strategy in the Wintering Red Kites (Milvus milvus). Diversity, 14(2): 117. https://doi.org/10.3390/d14020117
• García-Macía, J., López-Poveda, G., De la Puente, J., Bermejo, A., Galán, M., Álvarez, E., Morollón, S., Urios, V. (2022). The variability of juvenile dispersal in an opportunistic raptor. Current Zoology, zoac039. https://doi.org/10.1093/cz/zoac039
• García-Macía, J., Pomares, A., De la Puente, J., Bermejo, A., Martínez, J., Álvarez, E., Morollón, S., Urios, V. (2022). Striking Variability in the Post-Reproductive Movements of Spanish Red Kites (Milvus milvus): Three Strategies, Sex Differences, and Changes over Time. 2022. Animals, 12, 2930. https://doi.org/10.3390/ani12212930
• García-Macía, J., Vidal-Mateo, J., de la Puente, J., Bermejo, A., Urios, V. (2022). Spatial ecology of the Red Kite (Milvus milvus) during the breeding period in Spain. Ornis Fennica, 00–00. https://doi.org/10.51812/of.124714

 

I also recommend to check out these complete Spanish monographies on the species biology and spatial ecology:

 

• Urios, V. y García-Macía, J. (2022). Migración y ecología espacial de la población española de milano real. Monografía nº8 del programa Migra. SEO/Birdlife: Madrid, España. https://seo.org/boletin/seguimiento/migracion/08_milano_real/08_milano_real.pdf
• Sanz-Zuasti, J.; Velasco T.; Arroyo, B.; Rico, M.; Bermejo, A. & De la Puente, J. (2023). The red Kite. Biology and conservation. Fundación del Patrimonio Natural de Castilla y León. Valladolid

 

I will recommend specific citations throughout the manuscript.

 

• Despite its specific importance, I think that the role of roads and tracks in detection probability occupies too prominent a place throughout the manuscript, to the detriment of other issues that are not strictly methodological. For example, in my opinion, I think that it is not necessary to include this issue in the title. I would also move the first paragraph of the Discussion to somewhere in the end of that section, etc.

 

 

MINOR COMMENTS

• L52-L53. Where?
• L49-L54. Please provide more information on the ecology of the Red Kite and its population status throughout Spain. How many pairs are? Please, describe its basic migratory strategies and spatial ecology, i.e., providing overall information on the origin of the red kites wintering in Spain.
• L56-57. The spatial breeding ecology and habitat selection during this period has also been specifically studied in Spain by using GPS remote tracking: García-Macía, J., Vidal-Mateo, J., de la Puente, J., Bermejo, A., Urios, V. (2022). Spatial ecology of the Red Kite (Milvus milvus) during the breeding period in Spain. Ornis Fennica, 00–00. https://doi.org/10.51812/of.124714
• You can also cite the above reference ehre.
• You stated before that “the aim of the study was to determine the wintering population size of the Red Kite in the MHBR”, so the hypothesis of “no red kites during the breeding period” is out of the objectives of the paper. Please, revise.
• L107-108. Why did you select that time interval? Please, explain.
• L116-118. Why did you use a Fisher’s test instead of Chi-squared test in Motilla del Palancar? Please, explain.
• Please, provide the number of birds per km² too since it is easier to interpret.
• Presence of carcasses of which animals? Please, specify.
• Is “mountain areas” the proper term for Motilla del Palancar area? Those areas are slightly elevated and rough compared to the surrounding flat areas, but this term can be somewhat confusing. Please, clarify.
• L247-248. This was also studied within Spain with GPS-tagged Red Kites during the breeding season: García-Macía, J., Vidal-Mateo, J., de la Puente, J., Bermejo, A., Urios, V. (2022). Spatial ecology of the Red Kite (Milvus milvus) during the breeding period in Spain. Ornis Fennica, 00–00. https://doi.org/10.51812/of.124714
• This statement can be supported by these works: García-Macía, J., Pomares, A., De la Puente, J., Bermejo, A., Martínez, J., Álvarez, E., Morollón, S., Urios, V. (2022). Striking Variability in the Post-Reproductive Movements of Spanish Red Kites (Milvus milvus): Three Strategies, Sex Differences, and Changes over Time. 2022. Animals, 12, 2930. https://doi.org/10.3390/ani12212930

 

García-Macía, J., López-Poveda, G., De la Puente, J., Bermejo, A., Galán, M., Álvarez, E., Morollón, S., Urios, V. (2022). The variability of juvenile dispersal in an opportunistic raptor. Current Zoology, zoac039. https://doi.org/10.1093/cz/zoac039

 

These papers suggested that long-range post-breeding (= wintering) movements (since wandering dispersal to short-distance migrations) can be performed by both juveniles and adult females, so those birds can be mixed with the European Red Kites during the winter in Iberia, i.e. a given wintering roost in Iberia might include both European migratory birds and dispersal juveniles and adult females.

 

• Did you locate wintering roosts within the study area? Or, did you know about the presence of roosts before carrying out the surveys?
• May you provide specific targeted conservation measures within the MHBR to improve habitat suitability and quality for the Red Kite or other related species?
• I do not see the relationship between the increase of the England breeding population and the wintering population of Red Kites in Spain. As far as I know, red kites in UK seldom abandon the British Islands and did not perform long migrations to the European continent. It could happen (some ringed birds from England have appeared in Spain), but this relationship may be low. A stronger relationship may be found between the increasing continental European populations (from Germany for example) and the wintering birds in Spain.
• Table 1. Should it be “Motilla del Palancar” instead of “MoNtilla del Palancar”? Please, revise throughout the manuscript.

Author Response

MAJOR COMMENTS 

Simple summary. I would include a statement about the importance of Iberia (or specifically La Mancha Húmeda) as wintering grounds of the Red Kites. Does the Iberian Peninsula regions host an important part of the European population during winter?

Simple summary. Similarly, after you stated that “the reserve hosts a wintering population of around 1,400 red kites”, could you provide a simple estimation of which proportion of Iberian wintering red kites are those 1,400 based on estimated Iberian wintering population? (see for example https://seo.org/ave/milano-real/#:~:text=El%20área%20de%20distribución%20de,la%20invernada%20del%202013-2014.)

We combined the two suggestions above into the simple summary: “Spain hosts approximately 50,000 wintering Red Kites, which represents nearly three-quarters of the European total population, and the study area contains nearly 5% of this total.”

Simple summary. Lines 26-27 and Abstract (L43-45). Are all agricultural landscapes equally key to supporting wintering Red Kites? Including super-intensive agricultural lands? Is it enough only to “protect” those areas or habitat improvements are also necessary to enable the colonization and expansion of raptors with preference for open and heterogeneous landscapes which incorporates agricultural lands within?

We agree with the Reviewer; therefore, we have modified both the simple summary and the abstract to incorporate the suggested idea. However, we could not enlarge as much as we would like to given the words limitation for these sections.

Section 2.2. I think the manuscript would greatly improve if the authors provided some geographical information on the sampling protocol. For example, I think a map of the study area is necessary; for instance a map from Europe/Iberia to locate the study area and a smaller scale map of La Mancha Húmeda showing the diversity of habitats of the area. Could you also provide a figure with the location of the line transects? I think it would be useful to understand the distribution of the sampling effort in order to assess the degree of coverage of the study area.

We agree with the Reviewer and have added a map of the study area as suggested by the other reviewers and the Editor (see Figure 1). The distribution of the sampling effort is detailed in Table 1, which shows the number of transects per area and season.

Section 2.2. With your approach, densities were estimated at the level of locality or transect? Could you estimate kites density for each specific transect? If so, why did you compare non-corrected abundance for some of the analysis if you could use non-biased density? Please, clarify.

Density estimates were made at the locality level using Distance Sampling. Although it is possible to estimate density for each specific transect, this was not done because it would produce many zero values, complicating analysis. Therefore, we chose to provide estimates aggregated by zone for more robust results.

Section 2.2. I have concerns about the methodological approach to analyse these data. The authors used a wide variety of parametric and non-parametric test —most of them essentially specific cases of linear models— which mostly indicated if there were differences among levels of some factors (localities, type of road, winters, itinerary length…). The problem I found is that some of the variables may be interacting with other variables or at least modulating them. If all variables were incorporated into the same model, it would be possible to assess the influence of explanatory variables in response variables controlling the variability explained by all of them, as well as to assess the effect size and its confidence intervals. In this regard, I think that authours could use simple Generalized Linear (Mixed) Models to get a more parsimonious approach. For example:

Presence/absence data // Density ~ Type of road + Locality + Itinerary length + Winter + Type of road*Locality…

 ...or similar models.

We appreciate the comment but prefer to maintain the analyses using Distance Sampling, as itinerary length is already incorporated in the method itself, and only the winter season was considered. The relevance of road type was tested as well. We understand that generalized linear (mixed) models could provide complementary information; however, for the study's objectives, the current approach using Distance Sampling is sufficient. The only significant differences detected were among localities, which is why results are presented at the locality level.

Section 2.3. Why did you only consider 100-meters intervals to characterize habitat types? You surveyed kites up to 1000 perpendicular meters, so why did you not quantified habitat types until that distance?

Initially, we calculated habitat surface up to 1000 meters and compared it with the 100-meter interval method (frequency based on presence/absence) used in the paper. Please, take into account that we didn’t take only 100-metres. We took the most relevant habitat (on approximately 1000 metres using the rangefinder). We clarified it into the methods. Since no significant differences were found between the two methods, we chose the 100-meter interval approach as it simplifies and facilitates fieldwork. A similar approach has been used to determine habitat availability in other studies in the same environment, such as: Britto, V. O., Gil–Delgado, J. A., Gosálvez, R. U., López–Iborra, G. M., Velasco, A., 2018. Foraging habitat selection by gull–billed tern (Gelochelidon nilotica) in Central Spain (Castilla–La Mancha). Animal Biodiversity and Conservation, 41.2: 301–310.

Section 2.3. Did you consider to use not only the surface area of different habitat types but a Heterogeneity Index of landscape composition? Red Kites may select one habitat or another at small scale, but they may also select heterogeneous landscapes at broader scales, for instance composed by flat agricultural lands, shrubs areas, wood croops, semiurban areas, etc.

Unfortunately, we did not take into account a Heterogeneity Index. Besides, we did not consider microhabitat selection because the main aim of the paper was focused on the macrohabitat level. We agree with the reviewer that Red Kites might select habitats at smaller scales. However, given the Distance Sampling method we used, based on the first detection and considering both perched and flying birds, the microhabitat noted at first contact might differ from the habitat actually used e.g. for foraging, as the bird may simply be flying to another habitat. We acknowledge this as a limitation of our study, which could be addressed in future research using GPS tracking, as other researchers have done in very interesting studies about the species.

I detected a lack of references of studies of the Red Kite within Spain. The Spanish population of the species (including breeding habitat selection, wintering movement strategies, migratory patterns, juvenile dispersal, post-reproductive strategies selection, etc.) have been studied just a few years ago, greatly increasing the knowledge of the species. The authours cited some papers from Germany, UK and a few from Spain, but I think they should incorporate the most recent findings on the spatial ecology of the species (when it was necessary). For example:

García-Macía, J., De La Puente, J., Bermejo, A., Raab, R., Urios, V. (2022). High Variability and Dual Strategy in the Wintering Red Kites (Milvus milvus). Diversity, 14(2): 117. https://doi.org/10.3390/d14020117

García-Macía, J., López-Poveda, G., De la Puente, J., Bermejo, A., Galán, M., Álvarez, E., Morollón, S., Urios, V. (2023). The variability of juvenile dispersal in an opportunistic raptor. Current Zoology, zoac039. https://doi.org/10.1093/cz/zoac039

García-Macía, J., Pomares, A., De la Puente, J., Bermejo, A., Martínez, J., Álvarez, E., Morollón, S., Urios, V. (2022). Striking Variability in the Post-Reproductive Movements of Spanish Red Kites (Milvus milvus): Three Strategies, Sex Differences, and Changes over Time. 2022. Animals, 12, 2930. https://doi.org/10.3390/ani12212930

García-Macía, J., Vidal-Mateo, J., de la Puente, J., Bermejo, A., Urios, V. (2022). Spatial ecology of the Red Kite (Milvus milvus) during the breeding period in Spain. Ornis Fennica, 00–00. https://doi.org/10.51812/of.124714

 I also recommend to check out these complete Spanish monographies on the species biology and spatial ecology:

Urios, V. y García-Macía, J. (2022). Migración y ecología espacial de la población española de milano real. Monografía nº8 del programa Migra. SEO/Birdlife: Madrid, España. https://seo.org/boletin/seguimiento/migracion/08_milano_real/08_milano_real.pdf

Sanz-Zuasti, J.; Velasco T.; Arroyo, B.; Rico, M.; Bermejo, A. & De la Puente, J. (2023). The red Kite. Biology and conservation. Fundación del Patrimonio Natural de Castilla y León. Valladolid

 I will recommend specific citations throughout the manuscript.

Thank you very much for these interesting and relevant citations. We have cited them throughout the manuscript where appropriate and have also used them to expand the introduction and discussion sections, as suggested by the Editor.

Despite its specific importance, I think that the role of roads and tracks in detection probability occupies too prominent a place throughout the manuscript, to the detriment of other issues that are not strictly methodological. For example, in my opinion, I think that it is not necessary to include this issue in the title. I would also move the first paragraph of the Discussion to somewhere in the end of that section, etc.

We removed the reference to roads from the title as suggested also by the Editor. The current title, "Wintering Red Kites in Central Spain: Macrohabitat Selection and Population Density Estimate," more accurately reflects the content of the paper.

We agree that with the current modification of the title, it makes more sense to move the first paragraph. We have done so accordingly, as suggested. 

MINOR COMMENTS

L52-L53. Where?

We added: in Spain.

L49-L54. Please provide more information on the ecology of the Red Kite and its population status throughout Spain. How many pairs are? Please, describe its basic migratory strategies and spatial ecology, i.e., providing overall information on the origin of the red kites wintering in Spain.

As suggested by the Editor, we have expanded the introduction using the insightful references provided.

L56-57. The spatial breeding ecology and habitat selection during this period has also been specifically studied in Spain by using GPS remote tracking: García-Macía, J., Vidal-Mateo, J., de la Puente, J., Bermejo, A., Urios, V. (2022). Spatial ecology of the Red Kite (Milvus milvus) during the breeding period in Spain. Ornis Fennica, 00–00. https://doi.org/10.51812/of.124714

We have modified the introduction by incorporating several of the suggested references, as also recommended by the Editor. This reference was specifically included where suggested.

You stated before that “the aim of the study was to determine the wintering population size of the Red Kite in the MHBR”, so the hypothesis of “no red kites during the breeding period” is out of the objectives of the paper. Please, revise.

We have revised and more clearly defined the aim and the hypothesis for our study as suggested also by the Editor.

L107-108. Why did you select that time interval? Please, explain.

Between sunrise and midday is the peak of their activity, while from the afternoon onwards it declines. We added the following reference: Pfeiffer, T., & Meyburg, B. (2022). Flight altitudes and flight activities of adult Red Kites (Milvus milvus) in the breeding area as determined by GPS telemetry. Journal of Ornithology, 163, 867 - 879. https://doi.org/10.1007/s10336-022-01994-1.

L116-118. Why did you use a Fisher’s test instead of Chi-squared test in Motilla del Palancar? Please, explain.

We applied the Fisher exact test when expected frequencies were low (Zar 1984). We clarified it in the text.

Please, provide the number of birds per km² too since it is easier to interpret.

We provided it, as suggested.

Presence of carcasses of which animals? Please, specify.

We included a sentence to clarify that according to the suggestion: Cats and rabbit carcasses are frequently found on both paved and unpaved roads, while dog carcasses are also common along paved roads (pers. obs.).

Is “mountain areas” the proper term for Motilla del Palancar area? Those areas are slightly elevated and rough compared to the surrounding flat areas, but this term can be somewhat confusing. Please, clarify.

We agree that the previous terminology might have been confusing, so for those transects (areas) that were not inside the MHBR, we have changed the name to “non-reserve areas”. Thank you for this suggestion. This has been clarified in the study area section and modified throughout the paper.

L247-248. This was also studied within Spain with GPS-tagged Red Kites during the breeding season: García-Macía, J., Vidal-Mateo, J., de la Puente, J., Bermejo, A., Urios, V. (2022). Spatial ecology of the Red Kite (Milvus milvus) during the breeding period in Spain. Ornis Fennica, 00–00. https://doi.org/10.51812/of.124714

This statement can be supported by these works: García-Macía, J., Pomares, A., De la Puente, J., Bermejo, A., Martínez, J., Álvarez, E., Morollón, S., Urios, V. (2022). Striking Variability in the Post-Reproductive Movements of Spanish Red Kites (Milvus milvus): Three Strategies, Sex Differences, and Changes over Time. 2022. Animals, 12, 2930. https://doi.org/10.3390/ani12212930

 García-Macía, J., López-Poveda, G., De la Puente, J., Bermejo, A., Galán, M., Álvarez, E., Morollón, S., Urios, V. (2022). The variability of juvenile dispersal in an opportunistic raptor. Current Zoology, zoac039. https://doi.org/10.1093/cz/zoac039

 These papers suggested that long-range post-breeding (= wintering) movements (since wandering dispersal to short-distance migrations) can be performed by both juveniles and adult females, so those birds can be mixed with the European Red Kites during the winter in Iberia, i.e. a given wintering roost in Iberia might include both European migratory birds and dispersal juveniles and adult females.

We added the information and the suggested references in several sections of the paper (Introduction and Discussion, mainly).

Did you locate wintering roosts within the study area? Or, did you know about the presence of roosts before carrying out the surveys?

No, we were unaware of any roosts prior to the surveys, and only one group of three individuals was found during the surveys. Most were observed solitarily.

May you provide specific targeted conservation measures within the MHBR to improve habitat suitability and quality for the Red Kite or other related species?

We have added a paragraph outlining measures to protect the Red Kite both within the MHBR and throughout its native range. We cited: Mattsson, B. J., Mateo-Tomás, P., Aebischer, A., Rösner, S., Kunz, F., Schöll, E. M., Åkesson, S., De Rosa, D., Orr-Ewing, D., Bodega, D., Ferrer, M., Gelpke, C., Katzenberger, J., Maciorowski, G., Mammen, U., Kolbe, M., Millon, A., Mionnet, A., Puente, J., Raab, R., … Viñuela, J. (2022). Enhancing monitoring and transboundary collaboration for conserving migratory species under global change: The priority case of the red kite. Journal of environmental management, 317, 115345. https://doi.org/10.1016/j.jenvman.2022.115345

I do not see the relationship between the increase of the England breeding population and the wintering population of Red Kites in Spain. As far as I know, red kites in UK seldom abandon the British Islands and did not perform long migrations to the European continent. It could happen (some ringed birds from England have appeared in Spain), but this relationship may be low. A stronger relationship may be found between the increasing continental European populations (from Germany for example) and the wintering birds in Spain.

We agree with the Reviewer’s comments and have accordingly revised the discussion by adding a new paragraph. Furthermore, we have removed the unrelated section from the conclusions as suggested.

Table 1. Should it be “Motilla del Palancar” instead of “MoNtilla del Palancar”? Please, revise throughout the manuscript.

Yes, it should. Thank you for identifying this error. The correction has been made throughout the entire manuscript.

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript uses survey in order to estimate wintering population size and habitat selection of Red Kites in a Biosphere Reserve (MHBR) in central Spain. The authors estimate the wintering population size and report that the kites use the steppe-like agricultural inter-lagoon areas of the MHBR as important wintering grounds. I believe that the data can provide a nice contribution for the conservation of this species. I found the manuscript interesting, concise and well-written. I do have a few comments for the authors that I hope will help to clarify it, especially the methods used.

 

General comment:

Some of the analyses and statistical methods are not clearly described. For example, which type of habitats were recorded? How was road length controlled for? Habitat characterization was made by road/area – it is not clear how it is related to kits’ observations.

 

Other comments:

Lines 52-53: please provide more information on these species.

Lines 73-75: Is there a total estimation of the wintering population of Red Kites in Spain?

Lines 116-122: It is not clear to me how the length of the sampling road was controlled for? Why was Fisher’s exact test used?

Table 2: change “road tipe” to “road type”

Lines 149-153: Why didn’t you use remote sensing to calculate continuously the habitats?

Line 183: no significant.

Lines 187-192: Population size estimates have several assumptions such as even distribution of the kites. Please mention model assumptions and limitations.

Lines 220-221, 224-225: It was not clear to me the context of these sentences here. Please clarify this paragraph to explain the argument.

Lines 222-223: Scientific names should be in Italic.

Lines 228-232: Was there any difference between paved and unpaved roads in the surrounding habitats?

Lines 234-237: Are there older records for Red Kites breeding at MHBR? Please provide the available data on dates and population size. Can you suggest why the kites winter but do not breed there (although in Germany they breed in similar habitats – lines 247-248)?

Lines 254-259: Do you have data on food availability in these habitats? Is the Kite density comparable to ‘small home range’? I am not sure I was able to follow this argument.

Lines 271-273: I did not understand why lower density than expected in MHBR suggest that it is a particular favorable wintering area for the Red Kite.

Author Response

General comment:

Some of the analyses and statistical methods are not clearly described. For example, which type of habitats were recorded?

Every distinct available habitat was taken into account. The first paragraph of section 2.3, Habitat Characterization and Analysis, has been modified accordingly.

How was road length controlled for?

Given that censuses were carried out by car. It has an option to control for distance covered called “Trip”. It was used for controlling road length. Censuses were conducted on local, comarcal, and agricultural roads and tracks that allowed maintaining a cruising speed of approximately 30 km/h. See answer to Reviewer 2.

Habitat characterization was made by road/area – it is not clear how it is related to kits’ observations.

Yes it was carried out by road/area. As the main aim of the paper was to estimate density and to obtain macrohabitat preferences, and no observations of Red Kites were made out of the MHBR, this approach can be considered enough, given that microhabitat was not studied in the present study.

We confirm that habitat characterization was carried out by road or area as mentioned. The primary objective of this study was to estimate density and assess macrohabitat preferences of Red Kites within the MHBR. Importantly, all observations of Red Kites were restricted to the MHBR, with no sightings recorded outside of this area. So, our habitat characterization was focused on the macrohabitat scale relevant to where the species was observed. Although microhabitat features were not investigated in this study with relation to Red Kites, the road/area-based approach provides sufficient resolution for understanding broader habitat associations at the landscape level. Therefore, while microhabitat characteristics are important, our methods are appropriate and adequate for the stated aims of density estimation and macrohabitat preference analysis within this protected region. However, we state that as a limitation of our study.

Other comments:

Lines 52-53: please provide more information on these species.

Thank you for the suggestion. As noted by the Editor and Reviewer 2, we expanded the Introduction section to include several aspects of populations in Europe and Spain, as well as their migratory biology.

Lines 73-75: Is there a total estimation of the wintering population of Red Kites in Spain?

Yes, this has been stated in the paper, as also suggested by Reviewer 2.

Lines 116-122: It is not clear to me how the length of the sampling road was controlled for? Why was Fisher’s exact test used?

Thank you for the comment. As per the Editor's and your previous suggestion, we have updated the Methods section accordingly.

We applied the Fisher exact test when expected frequencies were low (Zar 1984). We clarified it in the text. It was also suggested by Reviewer 2.

Table 2: change “road tipe” to “road type”

Thanks for detecting the typo. We corrected it.

Lines 149-153: Why didn’t you use remote sensing to calculate continuously the habitats?

Habitat characterization was carried out in situ, following the approach used by Britto et al. (2018) in the same area, which we now cite in the revised manuscript. While remote sensing can provide continuous habitat data, our method has a key advantage in this heterogeneous agro-environment: habitats in the MHBR can vary considerably throughout the year. In situ observations therefore offer more reliable and current habitat information than remote sensing, which may not capture fine-scale temporal variability in habitat composition.

Line 183: no significant.

It has been already corrected, as suggested also by reviewer 1.

Lines 187-192: Population size estimates have several assumptions such as even distribution of the kites. Please mention model assumptions and limitations.

In the 2.2. Red Kite sampling we have included the assumptions, as it was also suggested by Reviewer 1.

Lines 220-221, 224-225: It was not clear to me the context of these sentences here. Please clarify this paragraph to explain the argument.

We have modified the discussion according to other Reviewers and the Editor.

Lines 222-223: Scientific names should be in Italic.

Thank you, we corrected both names.

Lines 228-232: Was there any difference between paved and unpaved roads in the surrounding habitats?

Given that no difference was found in density estimates between the two transects, we did not compare the surrounding habitats.

Lines 234-237: Are there older records for Red Kites breeding at MHBR? Please provide the available data on dates and population size. Can you suggest why the kites winter but do not breed there (although in Germany they breed in similar habitats – lines 247-248)?

Before the decline, the species was present in the area, but we were unable to determine the population size.

We incorporated your suggestion in the first paragraph of the discussion.

Lines 254-259: Do you have data on food availability in these habitats? Is the Kite density comparable to ‘small home range’? I am not sure I was able to follow this argument.

Yes, we have partial data on the availability of rabbit carcasses in the study area (also suggested by Reviewer 2). This information has been included in the Methods, Results, and Discussion sections. What we meant is not that the home range is small, but rather that, in comparison, those home ranges are smaller than others reported.

Lines 271-273: I did not understand why lower density than expected in MHBR suggest that it is a particular favorable wintering area for the Red Kite.

We have revised the entire paragraph in the Discussion section. Additionally, we have incorporated suggested measures to enhance the suitability of the MHBR for wintering Red Kites.

We also cited: Mattsson, B. J., Mateo-Tomás, P., Aebischer, A., Rösner, S., Kunz, F., Schöll, E. M., Åkesson, S., De Rosa, D., Orr-Ewing, D., Bodega, D., Ferrer, M., Gelpke, C., Katzenberger, J., Maciorowski, G., Mammen, U., Kolbe, M., Millon, A., Mionnet, A., Puente, J., Raab, R., … Viñuela, J. (2022). Enhancing monitoring and transboundary collaboration for conserving migratory species under global change: The priority case of the red kite. Journal of environmental management, 317, 115345. https://doi.org/10.1016/j.jenvman.2022.115345

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Although very readable in it's original submission form, the changes to the ms made by the authors have improved it. I have only very minor comments and suggestions for potential amendments and these should take very little time to address.

Line 175.  I was unfamiliar with the term 'comarcal', learning that it is an administrative division in Spain and equivalent to 'regional' or 'district' in English.  I don't think there is a need to change this, it won't harm anyone who is unfamiliar with the term to look it up for themselves.

Line 197. (Figure 2). Misspelling of 'Length' in x-axis label.

Line 245. (Figure 3).  Standard CDS methodology generally leads to some truncating of the dataset to improve fit. Did you consider (or try) truncating the data at 500/600m? Fit (indicated in this Figure) appears to be fairly good so perhaps you did not deem it to be necessary. Some indication that you considered (or tried) truncation would remove the potential question from the reader's mind but it is not an essential amendment. 

Lines 259 and 267 (twice). 'itineraries' needs swapping for 'transects'.

 

Author Response

Although very readable in it's original submission form, the changes to the ms made by the authors have improved it. I have only very minor comments and suggestions for potential amendments and these should take very little time to address.

 Line 175.  I was unfamiliar with the term 'comarcal', learning that it is an administrative division in Spain and equivalent to 'regional' or 'district' in English.  I don't think there is a need to change this, it won't harm anyone who is unfamiliar with the term to look it up for themselves.

In Spain, the autonomous communities are organized into provinces and municipalities. Additionally, some autonomous communities have comarcas, an administrative unit above municipalities. “Comarcas” are regions that group several municipalities sharing similar characteristics such as culture, economy, and history. However, to facilitate understanding in English (given the unfamiliarity of the reviewer and future readers), we preferred to change and refer to them as regions, as the meaning is equivalent.

Line 197. (Figure 2). Misspelling of 'Length' in x-axis label.

We corrected the word and modified the Font.

Line 245. (Figure 3). Standard CDS methodology generally leads to some truncating of the dataset to improve fit. Did you consider (or try) truncating the data at 500/600m? Fit (indicated in this Figure) appears to be fairly good so perhaps you did not deem it to be necessary. Some indication that you considered (or tried) truncation would remove the potential question from the reader's mind but it is not an essential amendment.

Yes, we attempted to truncate the dataset at 400 meters; however, this did not improve model fit. Therefore, we opted to use 125-meter intervals, as these provided better model fits and improved the overall performance.

We added this biref information to this part of the manuscript.

Lines 259 and 267 (twice). 'itineraries' needs swapping for 'transects'.

We changed itineraries for transects, as previously suggested.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear editor and authours, 

 

the authours adressed or discussed all the issues suggested, so considering the direct objetive of the manuscript, I believe that is acceptable in its current form and falls within the scope of Birds.

Author Response

The authours adressed or discussed all the issues suggested, so considering the direct objetive of the manuscript, I believe that is acceptable in its current form and falls within the scope of Birds.

We are very grateful for your valuable early comments, which greatly helped us improve the quality of our manuscript.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have done wonderful job in addressing the reviewers’ comments. Congratulations!

I do not have further comments on the manuscript.

Author Response

The authors have done wonderful job in addressing the reviewers' comments. Congratulations!

I do not have further comments on the manuscript.

We are very grateful for your valuable early comments, which greatly helped us improve the quality of our manuscript.