Research Article |
|
Corresponding author: Georgi S. Popgeorgiev ( georgi.popgeorgiev@gmail.com ) Academic editor: Peter Mikulíček
© 2026 Georgi S. Popgeorgiev, Dimitar Plachiyski, Nedko Nedyalkov, Yurii V. Kornilev.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
Popgeorgiev GS, Plachiyski D, Nedyalkov N, Kornilev YV (2026) Reptiles and amphibians in Sakar Mountains and Dervent Heights, SE Bulgaria: distribution, diversity, and threats. Herpetozoa 39: 401-418. https://doi.org/10.3897/herpetozoa.39.e206655
|
The herpetofauna of the biodiverse Sakar Mountains and Dervent Heights (SE Bulgaria) remains understudied. The most recent review provided only limited distributional data for 8 amphibian and 20 reptile species (
batrachofauna, conservation, fires, grasslands, habitat loss, herpetofauna, mapping, road mortality, threats
— To Andro, Dobri, and Niki, true friends
Globally, multiple amphibian and reptile species are experiencing population declines, at least partly due to human activities (
The low-elevation Sakar Mountains and Dervent Heights in southeastern Bulgaria constitute an understudied region that supports high herpetofaunal diversity and abundances. The first data were published in 1912, whereas most subsequent publications have reported only a limited number of incidental observations (
The most recent comprehensive overview of the Bulgarian herpetofauna was provided by
Here, we compiled and synthesized available herpetofaunal records with precise locality information from the Bulgarian parts of the Sakar Mountains and Dervent Heights. We analyzed the distribution of recorded taxa, examined reptile–habitat associations, and identified the principal conservation challenges facing herpetofauna in the study area.
The study area is located in southeastern Bulgaria, near the borders with Turkey and Greece, and encompasses the Sakar Mountains, the Dervent Heights and the westernmost part of the Strandzha Mountains. To support nature management and conservation planning, we delineated the study area to include four adjacent Special Areas of Conservation (SACs) of the Natura 2000 ecological network: “Sakar” (BG0000212), “Zhdreloto na reka Tundzha” (BG0000217), “Derventski vazvishenia 1” (BG0000218), and “Derventski vazvishenia 2” (BG0000219). Together, these sites cover approximately 2,337 km2. Specifically, the dataset included observation records located within the 716 squares of the 2-km MGRS grid (Military Grid Reference System; UTM zone 35N, datum WGS 1984) (Fig.
The highest elevations within the study area are Vishegrad Peak (856 m a.s.l., Sakar Mountains), Gyurgen Bair Peak (555 m a.s.l., Dervent Heights), and Kervansaray Peak (632 m a.s.l., Strandzha Mountains) (
The study area is characterized by a heterogeneous landscape mosaic comprising fragmented forest stands, open herbaceous communities, shrub-encroached grasslands, and isolated or linear groups of trees embedded within a broader agricultural matrix. Within this anthropogenically modified landscape, these structural elements create numerous ecotones and may promote ecological connectivity. Vegetation is predominantly xerothermic. The remaining woodland patches consist primarily of mixed oak forests (Quercus frainetto, Q. cerris, Q. pubescens, Q. virgiliana) and Carpinus orientalis. Historically, native oak forests were replaced by C. orientalis and shrubs of Paliurus spina-christi (
We downloaded on 13 February 2026 the herpetofaunal records available for the study area from two digital databases: SmartBirds (https://SmartBirds.org) and GBIF (Global Biodiversity Information Facility; https://www.GBIF.org).
SmartBirds is an integrated system for the collection, centralized storage, management, and retrieval of biological field observations. Since its launch in 2015, it has been widely used in Bulgaria by professional and citizen scientists and has become the largest available repository of herpetofaunal records in the country. Records are entered into SmartBirds predominantly through the SmartBirds Pro Android application, which uses the GPS receiver integrated into mobile devices; less frequently, records are submitted through a web-based map interface that allows users to place observations on a high-resolution digital map. Observations made prior to 2015 onto paper datasheets have been imported into the SmartBirds database; the geographic coordinates for these records were obtained with hand-held GPS units. Accordingly, the geographic accuracy for SmartBirds records is generally ~10 m. The detailed data on SmartBirds, including exact locations, counts, and further biological information, are only visible to the contributors and are not publicly available.
SmartBirds includes all records collected by the authors, which constitute the majority of the dataset, as well as observations contributed by colleagues and citizen scientists. Data were typically collected during targeted herpetofaunal surveys for several projects, including mapping initiatives in protected areas, but also include incidental observations.
To supplement the taxon-distribution mapping, we queried GBIF for georeferenced amphibian and reptile occurrence records without reported geospatial issues, using a polygon encompassing the study area (
As noted in the Introduction, most published herpetofaunal observations from the study site lack available precise geographic coordinates and were therefore unsuitable for mapping at the spatial resolution used here. An exception is some publications co-authored by us and for which we contributed the majority of the data, available at the time in SmartBirds. However, these studies focused on nationwide taxon-distribution mapping and localities were presented only as centroids of 10-km grid cells (e.g.
All GIS-based analyses and map production were conducted in ArcGIS 10.4. We mapped occurrence records to the 10-km and the 2-km MGRS grids. For visualization purposes, classes of abundance and taxon richness per grid cell were defined using Jenks natural breaks optimization.
Frequency analyses were based on the number of records rather than on individuals’ counts. This minimizes bias arising from imprecise or effectively categorical abundance estimates, such as counts of amphibian larvae reported as “100” or “1,000” individuals. In addition, most records represented single individuals.
Although systematic sampling of every grid cell within the study area was not an objective of this study, surveys conducted over the sampling period encompassed most available terrestrial habitat types and most aquatic features, which are comparatively scarce within the study area. Based on a preliminary analysis, amphibians were excluded from habitat-diversity analyses because the sampling effort was insufficient to provide representative coverage, particularly given the limited extent and uneven availability of aquatic habitats.
For most SmartBirds records, the predominant habitat in which an individual was observed was recorded in the field. We grouped habitats into 12 categories: “Coniferous forests”, “Deciduous forests”, “Mixed forests”, “Ecotone”, “Meadows with shrubs”, “Micro-elements” (e.g. stone walls, isolated rock piles, clusters of bushes within a grassland), “Grasslands”, “Rocks” (predominantly rocky habitats), “Standing waters”, “Flowing waters”, “Farmlands”, and “Human habitations”. We used these data to assess the reptile diversity among habitat categories and identify key habitat types.
To identify habitat categories of particular importance for reptiles, we calculated coverage-standardized Hill numbers for each habitat category, that includes equalized richness (q = 0), and modified Shannon (q = 1) and Simpson (q = 2) indices (
To evaluate the effects of roads on the herpetofauna, we used a dedicated field in SmartBirds that denotes if an individual was found dead due to automotive traffic. We summarized the number of dead individuals for each taxon.
To assess the potential severity of fires on the study site, we downloaded data from the Fire events in the European Forest Fire Information System (v.2-3-1), for the period 2016–2025 (
Since the most recent comprehensive review of Bulgarian herpetofauna (
We compiled 10,649 observations (150 from GBIF and 10,499 from SmartBirds). The SmartBirds data has been published on GBIF (
We report the first record of Salamandra salamandra from the study area. We recorded 34 taxa within the study site (Table
Amphibian and reptile taxa recorded in the study area and their distribution among 2-km and 10-km MGRS grid cells. n indicates the number of grid cells in which a taxon was recorded. Percentages indicate the proportion of grid cells in which each taxon was recorded, relative to the total number of grid cells containing at least one herpetofaunal record (42 of the 10-km grid cells and 432 of the 2-km grid cells).
| Taxon | Records | 2×2 km | 10×10 km | ||
|---|---|---|---|---|---|
| n | % | n | % | ||
| Amphibia | |||||
| Pelophylax ridibundus | 414 | 154 | 35.6 | 34 | 81.0 |
| Rana dalmatina | 136 | 74 | 17.1 | 29 | 69.0 |
| Bufo bufo | 114 | 52 | 12.0 | 24 | 57.1 |
| Hyla orientalis | 114 | 55 | 12.7 | 24 | 57.1 |
| Triturus ivanbureschi | 101 | 43 | 10.0 | 23 | 54.8 |
| Lissotriton schmidtleri | 85 | 39 | 9.0 | 17 | 40.5 |
| Bufotes viridis | 63 | 39 | 9.0 | 16 | 38.1 |
| Bombina variegata | 34 | 15 | 3.5 | 10 | 23.8 |
| Pelobates balcanicus | 8 | 6 | 1.4 | 5 | 11.9 |
| Salamandra salamandra | 1 | 1 | 0.2 | 1 | 2.4 |
| Reptilia | |||||
| Podarcis tauricus | 1723 | 193 | 44.7 | 36 | 85.7 |
| Lacerta viridis | 1565 | 186 | 43.1 | 34 | 81.0 |
| Lacerta diplochondrodes | 1404 | 140 | 32.4 | 32 | 76.2 |
| Testudo hermanni | 1060 | 205 | 47.5 | 37 | 88.1 |
| Ablepharus kitaibelii | 1059 | 132 | 30.6 | 29 | 69.0 |
| Pseudopus apodus | 711 | 157 | 36.3 | 30 | 71.4 |
| Testudo graeca | 508 | 117 | 27.1 | 32 | 76.2 |
| Dolichophis caspius | 394 | 149 | 34.5 | 36 | 85.7 |
| Emys orbicularis | 342 | 107 | 24.8 | 31 | 73.8 |
| Vipera ammodytes | 186 | 82 | 19.0 | 26 | 61.9 |
| Mauremys rivulata | 143 | 33 | 7.6 | 19 | 45.2 |
| Natrix natrix | 80 | 57 | 13.2 | 25 | 59.5 |
| Malpolon insignitus | 67 | 44 | 10.2 | 16 | 38.1 |
| Elaphe sauromates | 59 | 43 | 10.0 | 20 | 47.6 |
| Natrix tessellata | 55 | 25 | 5.8 | 17 | 40.5 |
| Podarcis muralis | 50 | 29 | 6.7 | 12 | 28.6 |
| Xerotyphlops vermicularis | 48 | 7 | 1.6 | 6 | 14.3 |
| Zamenis longissimus | 41 | 37 | 8.6 | 18 | 42.9 |
| Mediodactylus kotschyi complex | 38 | 13 | 3.0 | 10 | 23.8 |
| Eryx jaculus | 23 | 13 | 3.0 | 6 | 14.3 |
| Anguis colchica | 11 | 10 | 2.3 | 8 | 19.0 |
| Platyceps najadum | 7 | 6 | 1.4 | 5 | 11.9 |
| Coronella austriaca | 4 | 4 | 0.9 | 4 | 9.5 |
| Darevskia praticola | 1 | 1 | 0.2 | 1 | 2.4 |
Records spanned the period 2003–2025 and were unevenly distributed through time, with three pronounced peaks corresponding to major national and local mapping initiatives, particularly those associated with the national obligations for reporting on the state of the Natura 2000 sites (Fig.
Records were located in 42 of 46 (91%) 10-km MGRS grid cells and in 432 of 716 (60%) 2-km grid cells within the study area (Fig.
We documented 34 taxa – 10 amphibians (orders Caudata, 3 taxa, and Anura, 7) and 24 reptiles (Testudines, 4, and Squamata, 20, of which 9 lizards and 11 snakes) (Table
The study area is among the most important regions for the herpetofauna in Bulgaria, supporting high taxon diversity within a relatively small area. Relative to the extant native Bulgarian herpetofauna, the recorded diversity represents approximately 37% of Caudata taxa, 46% of Anura, 100% of Testudines, 56% of lizards, and 64% of snakes. The 45 2-km grid cells containing more than 47 herpetofaunal records (6.3% of all grid cells) are particularly noteworthy and may indicate high local relative abundance. Based on the distribution of records and our field knowledge of the study area, key herpetological localities include the vicinity of Levka Village and the Levchenska River, the limestone hills near the town of Topolovgrad, the Tundzha River gorge near the Srem, Ustrem, and Lesovo villages, and the areas around Momina Tsarkva and Golyam Dervent villages.
As expected, taxa with broad distribution and high abundances were generally recorded more frequently and in more localities than rare taxa with limited distribution (Table
We report the first observation of S. salamandra from the study area: five larvae were found in a man-made stone-walled spring near Dripchevo Village on 17.05.2021 (41.9893°N, 26.2367°E). The taxon may be more widespread, but its activity and abundance are probably low under the region’s dry climatic conditions.
Podarcis muralis was not recorded by
Mediodactylus kotschyi complex has probably been under-sampled because it is nocturnal and closely associated with human settlements, which we have not visited extensively at night. It has been recorded only within the SAC “Sakar”; however, we expect it to occur in settlements throughout the study area.
Eight taxa were absent from the filtered GBIF data, collected mostly opportunistically and by citizen scientists, but were detected by our sampling: S. salamandra, B. variegata, P. balcanicus, D. praticola, A. colchica, C. austriaca, P. najadum, and Zamenis longissimus. These taxa were among the ten least frequently recorded taxa in the dataset (Table
Citizen-science data are invaluable in certain contexts, for example, by providing records from private properties that may otherwise be difficult or impossible to survey (
Twenty-four taxa, more than two-thirds of all recorded, were widely distributed across the four target SACs (Table
Amphibian and reptile taxa recorded in the four Special Areas of Conservation (SACs) comprising the study area. Values indicate the number of records per taxon within each SAC.
| Sakar | Zhdreloto na reka Tundzha | Derventski vazvishenia | Outside the target SACs | ||
|---|---|---|---|---|---|
| 1 | 2 | ||||
| Approximate area (km2) | 1321 | 79 | 387 | 550 | |
| Taxa per SAC | 33 | 27 | 28 | 29 | 28 |
| Taxon | |||||
| Ablepharus kitaibelii | 480 | 10 | 263 | 269 | 37 |
| Anguis colchica | 1 | 2 | 5 | 3 | |
| Bombina variegata | 32 | 1 | 1 | ||
| Bufo bufo | 67 | 9 | 11 | 17 | 10 |
| Bufotes viridis | 30 | 5 | 1 | 5 | 22 |
| Coronella austriaca | 1 | 1 | 2 | ||
| Darevskia praticola | 1 | ||||
| Dolichophis caspius | 211 | 14 | 37 | 78 | 54 |
| Elaphe sauromates | 33 | 1 | 4 | 15 | 6 |
| Emys orbicularis | 228 | 21 | 18 | 43 | 32 |
| Eryx jaculus | 18 | 3 | 1 | 1 | |
| Hyla orientalis | 60 | 20 | 5 | 17 | 12 |
| Lacerta diplochondrodes | 749 | 89 | 164 | 325 | 77 |
| Lacerta viridis | 709 | 247 | 197 | 323 | 89 |
| Lissotriton schmidtleri | 71 | 4 | 5 | 3 | 2 |
| Malpolon insignitus | 57 | 3 | 3 | 2 | 2 |
| Mauremys rivulata | 121 | 3 | 6 | 13 | |
| Mediodactylus kotschyi | 25 | 1 | 12 | ||
| Natrix natrix | 55 | 4 | 4 | 14 | 3 |
| Natrix tessellata | 36 | 3 | 2 | 4 | 10 |
| Pelobates balcanicus | 5 | 2 | 1 | ||
| Pelophylax ridibundus | 267 | 26 | 21 | 68 | 32 |
| Platyceps najadum | 5 | 2 | |||
| Podarcis muralis | 22 | 4 | 1 | 16 | 7 |
| Podarcis tauricus | 1033 | 67 | 156 | 395 | 72 |
| Pseudopus apodus | 421 | 69 | 68 | 127 | 26 |
| Rana dalmatina | 71 | 8 | 17 | 28 | 12 |
| Salamandra salamandra | 1 | ||||
| Testudo graeca | 301 | 65 | 53 | 74 | 15 |
| Testudo hermanni | 585 | 125 | 141 | 157 | 52 |
| Triturus ivanbureschi | 72 | 7 | 9 | 5 | 8 |
| Vipera ammodytes | 110 | 21 | 17 | 29 | 9 |
| Xerotyphlops vermicularis | 46 | 2 | |||
| Zamenis longissimus | 22 | 4 | 5 | 8 | 2 |
| Records | 5945 | 835 | 1214 | 2045 | 610 |
The red-eared slider, Trachemys scripta, is an invasive freshwater turtle with a nationwide distribution in Bulgaria, resulting from deliberate releases by pet owners, and it may reproduce locally, particularly in southern Bulgaria (
The heterogeneity of the data sources, including the data collection methods, and survey project objectives inevitably resulted in uneven sampling effort and geographical bias across the compiled dataset. In addition to the taxon-specific sampling biases discussed above, the temporal and spatial distribution of records and the observed taxon richness are likely strongly influenced by survey effort. For example, the four 2-km grid cells containing the highest numbers of records are those in which we conducted intensive targeted surveys consisting of multiple visits, often for the purposes of mapping and/or assessing the current state of the local herpetofauna. Because of time and financial constraints, such mapping initiatives typically require the prioritization of sites expected to support high biodiversity and abundance, potentially at the expense of less promising or less accessible sites. Furthermore, localities that are readily and rapidly accessible, including those close to the road network, are more likely to be surveyed than remote or difficult-to-access localities. Additional surveys are therefore needed in localities that have so far been under-sampled. Greater survey effort, and the resulting increase in occurrence records, will generally increase the probability of detecting additional taxa, including rare ones. Consequently, further surveys will likely refine—and may alter—the currently known distribution maps.
The analysis of reptile diversity revealed clear differences among habitat categories in both taxon richness and assemblage structure (Fig.
As the Hill-Shannon diversity (q = 1) increases, the effective number of taxa decreased in some categories, indicating uneven distributions of records among taxa. “Rocks” retained relatively high diversity values, suggesting a comparatively even assemblage structure. In contrast, “Grasslands” and several forest types showed more pronounced dominance by individual taxa. In “Grasslands”, this pattern was particularly associated with the predominance of Podarcis tauricus (431 records) and Lacerta diplochondrodes (89 records), whereas in forest habitats it was associated primarily with Testudo hermanni (181 records) and Lacerta viridis (92 records).
The Hill-Simpson diversity (q = 2), emphasizing dominant taxa, likewise showed substantial differences among habitat categories. “Rocks” and “Ecotones” retained relatively high values, consistent with comparatively stable assemblages without strong single-taxon dominance. In contrast, “Grasslands” and the aquatic habitats, generally had low values for most taxa, reflecting the dominance of habitat specialists. In particular, E. orbicularis accounted for 113 records from standing waters and 166 records from flowing waters, whereas Mauremys rivulata accounted for 25 and 89 records, respectively.
Habitat categories strongly affected by human activities, namely “Farmlands” and “Human habitations”, consistently exhibited low diversity values across all three measures. This pattern indicates their limited contribution to the maintenance of high reptile diversity within the study area.
Overall, the Hill diversity metrics indicate that structurally complex and transitional habitats — particularly “Ecotone”, “Meadows with shrubs”, and “Micro-elements” — are important for maintaining high reptile taxonomic diversity. In contrast, more homogeneous and anthropogenically modified environments are characterized by lower diversity and stronger dominance by individual taxa. This pattern is consistent with the well-established positive relationship between habitat heterogeneity and taxon diversity (
“Meadows with shrubs” appear to be particularly important habitats for reptiles in the study area. However, they have been subject to substantial anthropogenic pressure over the past two decades, including complete destruction resulting from poorly managed agri-environmental practices (
The NMDS ordination of reptile assemblages across habitat categories was broadly consistent with our a priori ecological expectations (Fig.
Roads exert multiple widely documented negative effects on herpetofauna (
Road-mortality records of amphibians and reptiles in the study area, based on 10 499 SmartBirds observations. % is based on all records for the taxon.
| Taxon | Count | % |
|---|---|---|
| Pseudopus apodus | 154 | 22.4 |
| Dolichophis caspius | 102 | 27.4 |
| Malpolon insignitus | 30 | 46.2 |
| Vipera ammodytes | 15 | 8.3 |
| Zamenis longissimus | 14 | 34.1 |
| Elaphe sauromates | 14 | 24.6 |
| Bufo bufo | 11 | 9.9 |
| Testudo hermanni | 10 | 0.9 |
| Lacerta viridis | 9 | 0.6 |
| Lacerta diplochondrodes | 7 | 0.5 |
| Natrix natrix | 7 | 8.9 |
| Bufotes viridis | 6 | 9.7 |
| Natrix tessellata | 4 | 8.3 |
| Emys orbicularis | 2 | 0.6 |
| Podarcis tauricus | 2 | 0.1 |
| Eryx jaculus | 1 | 4.8 |
| Pelophylax ridibundus | 1 | 0.2 |
| Ablepharus kitaibelii | 1 | 0.1 |
| Mauremys rivulata | 1 | 0.7 |
| Total | 391 | 3.7 |
High numbers of road-mortality records, both in absolute terms and as a proportion of the records for each taxon, were observed among large, non-fossorial snakes. The proportion of taxon-specific records classified as road mortality ranged from approximately 8% for Vipera ammodytes, Natrix natrix, and N. tessellata to 46.2% for Malpolon insignitus. Notably, we documented 102 road-killed Dolichophis caspius individuals, representing 27.4% of all records of this taxon. These snakes generally attain large body sizes, actively forage during the day, frequently occupy open habitats, and may utilize road surfaces for thermoregulation – traits that increase their susceptibility to being struck by a vehicle. Similar patterns have been documented elsewhere in Bulgaria (
Overall, the highest number of road-mortality records was documented for Pseudopus apodus (154 records; 22.4% of the 711 records for this taxon). Several characteristics may account for this result. It is Bulgaria’s largest lizard, reaching a total length of up to 1.4 m (
Amphibian road-mortality records were comparatively few, likely because relatively few amphibian breeding water bodies occur adjacent to roads within the study area. Approximately 10% of records of both Bufo bufo and Bufotes viridis involved road mortality. In contrast, only a single road-mortality record was available for Pelophylax ridibundus, despite this being a widespread taxon and the most frequently recorded amphibian in the study area (414 records). Nevertheless, in particular local contexts — especially where roads intersect breeding sites or seasonal migration routes — P. ridibundus may be among the taxa most frequently affected by vehicle traffic (e.g.
Our analysis indicates that fire is likely a major threat to biodiversity in the Sakar Mountains and Dervent Heights. During the nine-year assessment period, 73 fire events were recorded, affecting areas equivalent to up to 17.2% of the study area (Table
Annual fire occurrence and burned area in the study area, 2016–2025. Burned area is reported as the total annual area affected by fire, together with the median and minimum–maximum burned area per fire event (km2). Percentages represent the total annual burned area relative to the area of the 2-km grid cells of the study area.
| Year | Number | Area | % | Median | Minimum | Maximum |
|---|---|---|---|---|---|---|
| 2016 | 6 | 40.0 | 1.4 | 5.4 | 2.2 | 12.6 |
| 2017 | 3 | 7.2 | 0.3 | 1.5 | 0.4 | 5.4 |
| 2019 | 1 | 0.8 | 0.0 | 0.8 | 0.8 | 0.8 |
| 2020 | 8 | 43.0 | 1.5 | 1.8 | 0.1 | 22.0 |
| 2021 | 3 | 4.0 | 0.1 | 0.5 | 0.3 | 3.2 |
| 2022 | 10 | 25.7 | 0.9 | 2.3 | 0.2 | 5.2 |
| 2023 | 9 | 37.6 | 1.3 | 0.3 | 0.1 | 19.7 |
| 2024 | 20 | 206.3 | 7.2 | 0.4 | 0.0 | 81.2 |
| 2025 | 13 | 127.4 | 4.4 | 0.6 | 0.1 | 78.5 |
A key driver for the occurrence of fires in the study area was intentional burning of arboreal and shrub vegetation as a management practice to maintain open grasslands. Additional harmful land management practices include large-scale clearing of arboreal-shrub vegetation in grasslands, often carried out with heavy equipment such as shredders and mulchers, and the conversion of permanent grasslands into arable lands. Such land management practices have detrimental effects for the herpetofauna (
Both agricultural activities and fires need to be addressed to guarantee that the SACs will maintain their purpose of protecting the target biodiversity. Our results confirm the importance of shrub vegetation and micro-elements for reptile diversity and abundance Thus, conservation-oriented management should prioritize the retention of shrub patches, scattered trees, stone structures, and other microhabitat elements within grassland landscapes, while avoiding broad-scale mechanical clearance and the burning of suitable habitats.
Our greatest gratitude goes to Borislav Naumov, Mihail Iliev, Svetoslav Spasov, Dimitar Demerdziev, Vladimir Dobrev, Vladimir Mladenov, Dobromir Dobrev and Stoycho Stoychev for their invaluable help in species distribution research. We thank the editor Peter Mikulíček, Ljiljana Tomović and an anonymous reviewer for comments that improved the manuscript.
The authors have declared that no competing interests exist.
No ethical statement was reported.
An AI tool was used for language improvements of the final version of the manuscript. The authors reviewed all suggestions and take full responsibility for the content of the manuscript.
This study was supported by the projects “LIFE14 NAT/BG/001119 Restoration and sustainable management of Imperial Eagle’s foraging habitats in key Natura 2000 sites in Bulgaria”; “BG16M1OP002-3.028-0001-C02 Habitat Sakar” and “LIFE18 NAT/BG/001051 Conservation measures for the Lesser Spotted Eagle and its habitats in Bulgaria”.
Georgi Popgeorgiev: conceptualization, data curation, formal analysis, investigation, methodology, software, supervision, visualization, writing—original draft, writing—review and editing; Dimitar Plachiyski: conceptualization, data curation, investigation, writing—original draft, writing—review and editing; Nedko Nedyalkov: investigation, writing—original draft, writing—review and editing; Yurii V. Kornilev: conceptualization, data curation, formal analysis, investigation, methodology, software, visualization, writing—original draft, writing—review and editing.
G. Popgeorgiev https://orcid.org/0000-0002-7625-8898
D. Plachiyski https://orcid.org/0000-0002-3579-2517
N. Nedyalkov https://orcid.org/0000-0003-4586-3224
Y.V. Kornilev https://orcid.org/0000-0002-8596-8728
All of the data that support the findings of this study are available in the main text.