1 of 22

Population study of the Golden Jackal in Bulgaria, what factors influence the increased abundance of the species

Ilya Acosta, PhD

3NJC & 3IJC

Cluj-Napoca 09.04.2025

2 of 22

INTRODUCTION

Apparently, its expansion into Europe began in Bulgaria (Atanasov, 1953; Corbet, 1978; Spassov, 1989; Demeter & Spassov, 1993; Spassov, 2007).

Bulgaria has the largest population of the species in Europe (Kryštufek, et al., 1997; Markov G. , 2012).

The need to study its population characteristics

The golden jackal (Canis aureus Linnaeus, 1758) - is of great interest for the European fauna.

The expansion of the species in Europe is a relatively recent phenomenon.

3 of 22

OBJECTIVE

The main objective of this research was to establish the presence of Canis aureus in Bulgaria, to determine its distribution and relative density in the country.

4 of 22

TASKS

To collect new field data on the presence of the species in Bulgaria.

To collect data on the environmental factors for the study area.

To analyse and determine the influence of environmental factors on the distribution of the species.

To establish the relative density of Canis aureus populations in Bulgaria by recording territorial groups responding to acoustic stimulation.

To model the modern distribution of Canis aureus in Bulgaria.

5 of 22

MATERIALS AND METHODS (1)

Field research methods

Stimulated voice communication - Bioacoustic method

Playing a recording of a jackal howl over a megaphone at designated calling stations (CS) (Giannatos, et al., 2005).

621 calling stations (CS): 189 CS (high density-red) 271 CS (medium d.-orange) 161 CS (low d.-blue). (Spassov, 2007).

6 of 22

MATERIALS AND METHODS (2)

Field research methods (2)

Stimulated voice communication - Bioacoustic method (2)

It has been assumed that only territorial groups of jackals respond to stimulation with the reproduced recording (Van Lawick & Van Lawick-Goodall, 1970; Golani & Mendelssohn, 1971; Golani & Keller, 1975; Nikolskij & Pojarkov, 1981; Demeter & Spassov, 1993; Giannatos, et al., 2005; Comazzi, et al., 2016).

Positive response, the direction (angle of the direction of response-azimuth) is determined - the distance to the source was estimated using the categories close (<500 m), medium (500 - 1000 m), far (>1000)

Approximate locations of responses were mapped using QGIS 3.4.4 software (QGIS Development Team, 2019).

7 of 22

MATERIALS AND METHODS (3)

Field research methods (3)

Stimulated voice communication - Bioacoustic method (3)

8 of 22

MATERIALS AND METHODS (4)

Field research methods (4)

Data collection on environmental factors

9 of 22

MATERIALS AND METHODS (5)

Field data analysis and statistical methods

Exploratory analysis (univariate analysis) - data on environmental factors collected in the field.

Multivariate analysis in R 4.0.3 (R Core Team, 2020) using logistic regression - natural factors data collected in the field.

With the estimated area of a calling stations (7.0 km² ) and the registered responses (territorial groups), the density of registered responses at a point was established.

Kernel Function (Kernel Estimation Density) based map of recorded responses - QGIS 3.4.4 (QGIS Development Team, 2019). Radius influence on Kernel Function (50.00 km)

10 of 22

MATERIALS AND METHODS (7)

Statistical methods - maximum entropy model (MaxEnt) (Phillips, et al., 2006)

Maximum entropy model (MaxEnt) (Phillips, et al., 2006) - MaxEnt - version 3.4.4 (Phillips, Dudík, & Schapire, 2020).

Variables:
19 climate variables (BIOx) of WorldClim v. 1.4 (Hijmans, et al., 2005).
4 geographical variables (aspect, dembg35n, rivers, satel) (variables generated by Assoc. Popgeorgiev (NPM-BAS) using Spatial Analyst ArcGIS 10.3).
habitat type (clc) (CORINE Land Cover, 2012).
interspecific competition with Canis lupus (canis_lupus) (model developed by Prof. Dr. Vasil Popov (IBEI-BAS).
drought index (ai_yr) (Trabucco & Zomer, 2009).

The model was developed following the approach of Kornilev et al. (2017)

removed all highly and very highly correlated variables with a correlation coefficient above 0.70 (Spearman correlation test).
100 reps for each variable.
25 % of the points randomly selected for verification in each repetition cycle (random test percentage 25).
100 000 background points.
Logistic model output format (Output Format - Logistic).
Jackknife test in the model

11 of 22

RESULTS AND DISCUSSION

Stimulated voice communication - Bioacoustic method

For Bulgaria, 328 territorial groups

12 of 22

RESULTS AND DISCUSSION (2)

Environmental factors - 328 positive answers

13 of 22

RESULTS AND DISCUSSION (3)

Environmental factors (2) - 328 positive answers

14 of 22

RESULTS AND DISCUSSION (4)

Environmental factors (3) - 328 positive answers

15 of 22

RESULTS AND DISCUSSION (9)

Results of the logistic regression model

16 of 22

RESULTS AND DISCUSSION (11)

Population density of Canis aureus

Categorizing relative density. ACS: Registered answers at a single calling station. TG: Territorial groups at a single calling station. SA: Survey area per calling station. RD: Relative density. RDC: Relative density category.

17 of 22

RESULTS AND DISCUSSION (12)

Canis aureus relative population densities (2)

Band: Maximum = 24.57 • Mean = 5.13 • Minimum = 7.42e-14 • Stddev = 5.03 • Valid percent = 71.35

18 of 22

RESULTS AND DISCUSSION (13)

Species distribution model - MaxEnt

19 of 22

RESULTS AND DISCUSSION (14)

Species Distribution Model - MaxEnt (2)

(Average maximum test sensitivity plus specificity logistic threshold) е 0.44

20 of 22

CONCLUSIONS

Based on the number of responses recorded, it was found that Canis aureus is present throughout the country except in high mountain areas with steep terrain and large forest stands.

The factors that most strongly influence the distribution of the species in Bulgaria are altitude (the combination in our conditions of the factors of steep topography and large forests), habitats (land cover) and population of settlements.

Data indicate that the most favourable areas in the country are lowlands (< 500 m) with moderate slopes, habitats such as arable land and areas with mixed landscapes close to settlements but with populations no greater than 1000. Unfavourable areas for the species are high mountain areas with steep terrain and dense forest vegetation.

21 of 22

The locations with the highest densities in the country are the Southeast, Northeast, Northwest and Eastern Rhodopes with more than 4 territorial groups per 7.00 km .2

The species distribution model (MaxEnt) shows that the South East, North East, North West and Eastern Rhodopes are the areas of highest habitat suitability.

The results obtained from the bioacoustic survey in Bulgaria show that it is an effective method and easily applicable year-round for detecting the presence and distribution of Canis aureus, as well as for establishing the density of its populations.

CONCLUSIONS (2)