1 Introduction

Golden jackals (Canis aureus; Figure 1) have been surprising researchers over the past decade with a range expansion originating from the Balkans north- and westwards throughout Europe (Hatlauf, Bayer et al. 2021; Stefanović et al. 2024; Trouwborst, Krofel, and Linnell 2015). Currently, established populations can be found throughout Southeast and Central Europe (Hatlauf, Bayer et al. 2021). However, records have occurred as far west as Spain and France (Bouchet 2017; Buruaga et al. 2023; Royo-Vicente and García 2024) and as far north as Finland and the Subarctic and Arctic regions in Norway and Russia (Sørensen and Lindsø 2021, Rykov, Kuznetsova, and Tirronen 2022). This natural range expansion has raised many questions on why golden jackals have suddenly become such a successful species in Europe (Hatlauf, Bayer et al. 2021; Hatlauf, Böcker et al. 2021; Krofel et al. 2017).

It is commonly agreed upon that this range expansion illustrates the species' behavioural flexibility (Ćirović, Penezić, and Krofel 2016; Hatlauf, Bayer et al. 2021; Kebede 2017; Kojola et al. 2024; Moehlman and Hayssen 2018). Yet, current knowledge of the behavioural ecology of wild golden jackals in Europe is limited, mostly due to the elusive nature of the species (Csányi et al. 2023; Pecorella et al. 2023). However, this knowledge is crucial for developing (cost-)effective monitoring methods, identifying potential future human–animal conflicts and supporting coexistence (Berger-Tal and Saltz 2016; Bro-Jørgensen, Franks, and Meise 2019). One population of golden jackals holding great research potential is the insular population inhabiting Samos, a Greek island located in the eastern Aegean and separated by only a narrow sea strait (±2 km) from the Anatolian mainland (Giannatos 2004; Giannatos et al. 2005). Due to their relative isolation from the mainland, the population on Samos has likely been subjected to several short-term diversification processes caused by reduced rates of colonisation and gene flow (Blakeslee et al. 2019; Matthews and Triantis 2021). Like in mainland Greece, golden jackals on Samos faced a rapid population decline during the end of the last century. However, legal protection has resulted in a likewise rapid recovery of most populations both on Samos and on the mainland concurring with the global population range expansion of the species (Giannatos 2004; Giannatos et al. 2005; Karamanlidis et al. 2023).

The origins of the golden jackal population on Samos are not well known, since the native range of the species is yet to be determined (Stefanović et al. 2024). Nevertheless, a study on the genetic structure has found that there have been at least some connections between Samos and mainland Greece and that the genetic structure of the population on Samos is highly differentiated from populations on the mainland (Rutkowski et al. 2015). This indicates that the geographic barrier formed by the sea has resulted in a genetic distinctiveness of golden jackals living on Samos from those living on the mainland. Golden jackals on Samos Island are relatively easy to monitor due to their relatively higher population density than that in the mainland (Giannatos et al. 2005). Nonetheless, only a handful of past research efforts have been made regarding the behavioural ecology of golden jackals on Samos (Bulmer 2015; Pietroluongo, Leggett et al. 2018; Pietroluongo, Linardaki et al. 2018).

Acquiring information on golden jackal social behaviour is challenging, as this requires distinction between individuals. However, with close observations, individual golden jackals can be distinguished based on coat colouration patterns (Macdonald 1979; Wandrey 1975). Generally, golden jackals live in close family groups consisting of a monogamous dominant pair and their offspring (Macdonald 1979). By engaging in intimate social interactions, group members reinforce their social bonds and establish the social hierarchy. They cooperatively secure food sources and fend off strangers from their group's territory. Yearlings are known to sometimes remain at the nest as helpers, meaning that they aid their parents in raising subsequent generations of offspring before dispersing (Lanszki et al. 2018; Macdonald 1979; Wandrey 1975). Only recently, in one social unit, allosuckling behaviour, or the suckling of pups by a female other than the biological mother, has been documented, which may have been due to the breeding of two females within one social group (Pecorella et al. 2023).

We studied the social system of the golden jackals on Samos. A social system consists of three interrelated components: social organisation, mating system and social structure (Kappeler et al. 2013) (Figure 2). First, the social organisation of a group is determined by aspects such as its size, (genetic) composition and cohesion (Kappeler et al. 2013). Second, the mating system depends on the number of breeding males and females within a social group, and how often these individuals mate (Clutton-Brock 1989). Third, a group's social structure is the emergent outcome of all social interactions (except mating) and relationships within the group. Since social interactions and relationships strongly vary between individuals, and even within individuals throughout their lives, the social structure of a group may vary over a longer time (Hinde 1976; Kappeler et al. 2013). This knowledge is pivotal for understanding the ongoing range expansion of the species, as well as for developing effective monitoring and conservation methods.

In this study, we aimed to gain insights into the social behaviour of one group of golden jackals by exploring each of these three components. By closely monitoring this group, we planned to identify the group members and subsequently look at their social interactions and the strength of the social bonds between them. We hypothesised that individual golden jackals can be distinguished based on their unique coat colourations (Macdonald 1979). We also hypothesised that the studied group consisted of an adult pair and their offspring (Macdonald 1979) and, furthermore, expected changes in composition over time, as well as variations in behaviour depending on the individual's position within the family (Macdonald 1979; Moehlman 1987; Wandrey 1975). We explored the intricate social lives of these golden jackals by means of individual recognition based on camera trap pictures, the subsequent determination of the dominance hierarchy and social network analyses. Presented work contributes to narrowing the research gap on golden jackal social behaviour.

2 Materials and Methods

2.1 Study Area

This study was conducted on the Greek island of Samos over a period of 9 months. Located in the eastern Aegean Sea and less than 1.7 km from the Anatolian mainland, Samos has a length of about 45 km and covers a total surface area of approximately 480 km² (Stiros et al. 2000). The study area is located in Potami Mesokampos (Ποτάμι Μεσοκάμπου) in the southeast of Samos and has a total size of approximately 0.032 km². This area was selected based on the approximate estimation of the origins of group howls and the presence of golden jackal tracks. The area consists of a dried-out riverbed carved from north to south. In the west of the riverbed are small pastures and agricultural fields surrounded by dense vegetation and patches of coniferous forest that cover the eastern hill flank. Even further to the west, the research area is bordered by a military camp. However, the old barbed-wire fences do not form a barrier to the golden jackals. In the east of the riverbed, the landscape is dominated by agricultural fields with small farms. Parallel to the dried-out riverbed, a small road connects a larger road in the north with houses near the coast in the south (Figure 3).

Human activity in the study area is fairly low, with only the occasional goat shepherd passing through with his herd and dogs. Similarly, activity in the military camp in the west is relatively low. Nevertheless, litter coming from households, agriculture and river deposits can be found throughout the whole research area. This includes plastics, fabrics, metals, glass, barbed wires, agricultural tools, etc. In addition, the area to the east of the study area is subjected to high-to-medium human activity throughout the day. Besides humans, the study area is frequently visited by domestic animals, including domestic dogs and cats. On Samos, golden jackals do not face a big risk of predation, except for large birds of prey, which may potentially catch younger individuals. Other wild mammals that do occur are wild boar (Sus scrofa), European hare (Lepus europaeus), stone martens (Martes foina), least weasel (Mustela nivalis) and small rodents.

3 Results

Over 9 months (from 12 December 2023 to 30 October 2023), we have obtained a total of 1564 recordings containing golden jackals. We did not have any recordings for the month of May due to logistical issues. Of the 1.564 recordings, 609 were excluded due to bad video quality, making it impossible to identify individuals. Within the recordings, golden jackals were in frame for 7 h, 29 min and 22 s. The calculated autocorrelations of the footage lie between −0.002 and 0.058, indicating temporal independence of the data (Chatfield 2004).

3.1 Individual Characterisation

Individuals were found to indeed vary in coat colouration, making individual identification possible. Main discriminative characteristics were black spots on the front legs, colouration patterns on the chest and dark patches on the back of the tails (Figure 4). Moreover, some individuals had notably lighter or darker coat colours, and distinct marks and patches (cf. Figure 5). The total number of characterised individuals is supported by camera trap videos in which all group members were recorded. The average number of individuals per recording is 1.42. In some cases, we were not able to distinguish individual characteristics (e.g., bad light conditions, fast movement speed, bad camera angle, etc.). Finally, we identified the recorded individuals in 64.4% of the cases. Moreover, we reached a substantial intra-observer reliability score (κ = 0.804). For some of the individuals, we also managed to determine the sexes based on the footage of scent-marking behaviour and morphological characteristics (Table 3).

TABLE 3. Summary of individual group members.

Individual ID	Live stage	Sex	Rank	NormDS
IR	Adult	♀	1	4.655159
TA	Adult	♂	2	4.655159
ME	Adult	♂	3	3.382937
MO	Adult	Na.	4	2.285714
SE	Adult	♀	5	2.142857
EC	Adult	♂	6	2.142857
AT	Adult	Na.	7	1.571429
P1	Juvenile	Na.
P2	Juvenile	Na.
P3	Juvenile	Na.
P4	Juvenile	Na.
P5	Juvenile	Na.
P6	Juvenile	Na.

3.2 Social Organisation

We found that the studied group had a relatively stable composition. At the start of the study period, it consisted of seven individuals: two females, three males and two of unknown sex. The social group remained stable for 4 months (until May). At this time, a major change occurred involving the dispersion of three individuals and later the birth of a total of six pups (between 13 April and 23 June). After this point, the group remained stable in size and composition for the rest of the study period (Figure 6) (For a video containing all group members after the birth of the pups, see Video 1). From the original five subordinate golden jackals, two individuals, one female and one male, stayed with the group. We have observed them near the pups and engaging with them in social play, suggesting their role as helpers at the nest (Malcolm and Marten 1982).

This video cannot be streamed at this time, please download the video instead.

Within the social group, there were two clear dominant individuals, one female (IR) and one male (TA). Both have the highest normalised David's scores, with the female's score being slightly higher than the male's. Between the rest of the group members, values of normalised David's scores are lower and more homogeneous, indicating equal subordinate positions within the hierarchy (Table 3 and Figure 7) (for an example of behaviour related to dominance, see Video 2).

This video cannot be streamed at this time, please download the video instead.

The group had a fixed territory over time estimated to be about 0.07 km² in which they were most active, especially during the hours of daylight (Figure 3). It is within this territory that they engage in nightly group howling ceremonies and in which we have observed them raising their pups.

3.3 Mating System

Although the studied group consisted of multiple reproductively mature individuals at all times during the study period, only the dominant female and male have been observed engaging in behaviour related to pre-copulation (at the start of March). In this instance, the male was sniffing and licking the anogenital area of the female (Figure 8 and Video 3) (cf. Wandrey 1975). We did not observe the mating itself. During the month preceding this event, as well as during the same period, we observed a relatively higher proportion of the dominant pair staying in close proximity to each other (Figure 9). Additionally, we constantly observed the dominant male applying scent marks in the same spot as the dominant female.

This video cannot be streamed at this time, please download the video instead.

Moreover, in March, we recorded seven occasions in which unknown golden jackals entered the group's territory. Meanwhile, we did not observe any intruders outside the mating season. We recorded the resident golden jackals meeting the intruders only once. In this interaction, the resident individuals responded by quickly chasing the intruders away. In footage recorded in April, the dominant female appeared to be pregnant. However, it was not until the second half of June that six newborn pups were recorded for the first time. Of these pups, only five were recorded simultaneously at a later time. We observed notable size differences between some of these pups (Figure 10). At this time, the dominant female had enlarged nipples, indicating active suckling. Interestingly, similar to the dominant female, the subordinate female had enlarged nipples (Figure 11).

3.4 Social Structure

The results of the social network analyses show that in both networks (before and after dispersion and the birth of the pups), the dominant female (IR) and male (TA) have the highest eigenvector centrality scores, indicating that these were the most centralised individuals within the group (Figure 12). Nevertheless, the pups had a similarly high eigenvector centrality score in the second network. Thus, similar to both dominant individuals, the pups had a central position within the group's social network.

Accordingly, two clusters were recognised within the sociogram, with one containing the dominant female and male and the other containing all of the subordinate group members. The modularity of the social network had a value of 0.05 (Tables 4 and 5), which were below the threshold of 0.3 (30%) (Newman 2006), thus indicating a weak community structure. Therefore, the data only weakly supported the clusters within this sociogram. In the second sociogram, again, two clusters were recognised, with one containing both dominant individuals as well as the pups, and the other cluster containing the two subordinate golden jackals. However, since modularity (−6.93 × e⁻¹⁸) was very close to zero (Tables 4 and 5), the clusters within the network were not much different from what would be expected by random chance and were thus only very weakly supported.

TABLE 4. Eigenvector centrality scores of the individual jackals and modularity scores of the studied group, displaying the values before a major change in group structure involving the dispersion of three individuals (AT, ME, MO) and the birth of six pups (Pups).

Individual	Eigenvector centrality score
Dominant female (IR)	0.9747901
Dominant male (TA)	1.0000000
Subordinate 1 (ME)	0.3659128
Subordinate 2 (MO)	0.5523451
Subordinate 3 (SE)	0.6491962
Subordinate 4 (EC)	0.4826284
Subordinate 5 (AT)	0.6421497
Modularity	0.05000559

TABLE 5. Eigenvector centrality scores of the individual jackals and modularity scores of the studied group, displaying the values after a major change in group structure involving the dispersion of three individuals (AT, ME, MO) and the birth of six pups (Pups).

Individual	Eigenvector centrality score
Dominant female (IR)	1.0000000
Dominant male (TA)	0.9734171
Subordinate 3 (SE)	0.850765
Subordinate 4 (EC)	0.6743985
Pups (Pups)	0.9323426
Modularity	−6.938894E-18

4 Discussion

In this research, close monitoring of a golden jackal group living on Samos using camera trap footage has allowed us to reveal aspects of each of the three interrelated components of its social system: social organisation, mating system and social structure. Successful identification of individual group members based on physical characteristics has proven pivotal to our results. We have found evidence supporting current knowledge on golden jackal social behaviour while gaining some novel insights. The group composition of the studied group coincides with what is to be expected from a golden jackal group, with the group consisting of a dominant pair, a litter of pups, optionally one or more adult (subordinate) helpers and dispersing adults (Macdonald 1979; Moehlman and Hayssen 2018; Wandrey 1975). The fixated behaviour of the dominant male towards the dominant female and the scent-marking of the same location in quick succession are both indications of hormonal changes occurring in the dominant female that are related to the proestrus phase (Nagashima and Songsasen 2021; Wandrey 1975). The dominant male's behaviour towards the female likely serves as a signal to keep other males from mating with her (Moehlman 1987; Moehlman and Hayssen 2018). The fact that we have observed both this behaviour and anogenital licking between these two individuals only implies monogamy, as is the standard for golden jackals (Macdonald 1979; Moehlman and Hayssen 2018).

Over the study period, the group composition varied, with some yearlings dispersing from their natal group, while two yearlings, a male and a female, stayed at the nest, six pups were born in late spring. Although the exact dates of dispersal and the birth of the pups are unknown, the data suggest that both events roughly concurred. Further, we have determined a clear difference in dominance between two individuals and the rest of the group.

Behavioural changes starting in January suggested that the female was entering proestrus, a key phase before mating (Moehlman 1987; Nagashima and Songsasen 2021). These behavioural changes included the dominant male spending a larger proportion of time than usual in proximity to the dominant female. Moreover, the male immediately marked the same spot whenever the female applied her scent mark. We have observed one case of anogenital licking between the dominant male and female, a behaviour associated with pre-copulation (Wandrey 1975) (Video 2). The pups were likely born during May, since they were first observed in June (Kebede 2017; Wandrey 1975). At this time, the clearly enlarged nipples of the dominant female indicated active suckling (Jirků et al. 2018). Interestingly, like the dominant female, the subordinate female who stayed as a helper at the nest showed signs of enlarged nipples. However, except for the size difference between pups, there were no indications that she had given birth to pups herself.

SNA has revealed the strengths of the different social relationships within the group, which golden jackals occupy the most influential positions and whether there are clusters or subgroups of individuals within the group. The dominant pair was linked by the strongest tie. Initially, the rest of the network appeared to be relatively homogeneous, with social bonds between subordinates and between dominants and subordinates being roughly equal in strength. However, the birth of the pups and the dispersal of the yearlings have caused a major shift within the network. Afterwards, the pups occupied a central position within the group's social network similar to both dominant individuals. In addition, of both remaining subordinates, the female occupied a more central position than the male. This is a result of her being observed more time babysitting the pups, thus being in a social context. In both analyses, low modularity indicates that clusters within the network are only weakly supported by the data.

Enlargement of the nipples of the female helper without an apparent cause may be the result of a pseudopregnancy (De Vos 1969). Following the ovulation in the absence of mating, female jackals may behave as if they were impregnated. This behaviour is often accompanied by morphological changes including enlargement of the mammary glands, which may secrete a watery murky fluid (De Vos 1969). Hormonal shifts causing such changes could be tied to the strong social hierarchy and reproductive suppression often seen in subordinate females of social canids (Creel and Creel 2002). It is unclear whether this female also attempted to take part in ‘lactating’ the pups, as we have no recordings of any of the females lactating. Another possible explanation may be the observation of a second case of allosuckling in golden jackals in Europe (the first one, cf. Pecorella et al. 2023). This indicates that there would not be one but two litters of pups, with the dominant female being the mother of one and the helper female being the mother of the other litter. Two separate litters would also explain the observed size differences between some of the pups, although this could also be the result of differences in growth rates depending on the amount of energy consumed during the first weeks of their lives (Western 1979). Further research is needed to confirm any of these possible hypotheses.

This is the first time that social network analysis has been conducted on wild golden jackals. The strong tie between the dominant male and female was to be expected, as in golden jackals, the dominant pair generally has a strong pair bond, which lasts for a lifetime. Especially during the mating season, they spend the majority of their time together (Moehlman and Hayssen 2018; Wandrey 1975). Except for the adult pair, the golden jackals in the group do not have a major preference for associating with specific individuals. The clear difference between the dominant pair and other group members reflects the findings of previous studies on hierarchical structures in golden jackal groups (Macdonald 1979; Moehlman 1987). This study adds a new layer of understanding by using SNA, an analysis that has not previously been applied to this species in the wild. This method has allowed for a quantification of the central role of the dominant pair within the group. The finding that the strength of social bonds involving subordinates was roughly equal before the pups were born aligns with the general egalitarian nature of interactions seen in social carnivores (Creel and Creel 2002). The shift within the network after the birth of the pups is mainly the result of all individuals devoting large proportions of their time to providing care to the pups and ‘babysitting’ (Moehlman and Hayssen 2018; Pecorella et al. 2023; Wandrey 1975). The central role of the pups and the increased prominence of the female helper (likely due to her time spent babysitting) illustrate how parental care is a central organising principle in golden jackal societies, especially during pup-rearing periods (Moehlman 1987).

The lack of clusters within both social networks is likely the result of both networks consisting of only a few individuals who have all been observed interacting with each other. This differs from observations in canid species with larger social groups, where clear clusters of subgroups form within the group's social network (e.g., wolves, Mech 1999). The low modularity observed here implies that only weak subgrouping occurs. Although the frequency of interactions may not be equal between individuals, interactions do occur between all group members (Newman 2006).

Unfortunately, logistic and practical circumstances had not allowed us to continuously monitor the golden jackal group throughout the study period, particularly during May when the pups were likely born. Without continuous monitoring during this time, it is possible that key social dynamics and behavioural changes in the early phase of pup care, such as shifts in the interactions among group members or critical behaviours related to pup-rearing, may have been missed. The lack of data during May also affects our understanding of seasonal behaviours and their full impact on the social network. While the observed shift in the social network after the birth of the pups suggests significant changes, additional data from the earliest weeks following the pups' birth would have allowed for a more robust assessment of the social interactions and cooperative care roles, contributing to a more comprehensive view of how the social structure of the group fluctuates throughout the year. Although camera traps have provided valuable data, this data collection method limits the detection of more fine-scale social interactions and dynamics that may occur out of view. More camera traps would have been desirable, as they would have helped to capture more comprehensive data on individual interactions and social changes within the group. Additionally, it would have been valuable to use genetic tests to confirm whether the pups are indeed of separate litters. Such tests could provide clearer answers about allosuckling or multi-litter group structures, as has been observed in golden jackals before (Pecorella et al. 2023).

5 Conclusion

In this study, we aimed to uncover various aspects of the social system of one golden jackal group using a relatively non-invasive method. With a limited number of camera traps, we successfully identified individual group members for the first time in Europe, revealed their social relationships and monitored changes within the group throughout the year. In doing so, we proved that golden jackals can indeed be distinguished based on unique physical characteristics. As predicted, our findings support and refine our current general knowledge of golden jackal social behaviour. Moreover, our predictions regarding variation in group composition and social behaviour over time have also been proven correct.

Despite some limitations, we accomplished our goal of gaining new insights into the social behaviour of the unique population of golden jackals living on Samos Island. A key limitation was the limited number of camera traps, which restricted our ability to capture fine-scale data on individual interactions. Additionally, the absence of data in May, a critical period for pup birth and early care, has constrained our understanding of the social dynamics during this pivotal time. Nevertheless, our findings contribute to our broader knowledge of golden jackal social behaviour, as we have confirmed some general topics while revealing new insights. We hope this knowledge will contribute to a better understanding of the species and their social interactions, which are important in developing specific and effective conservation methods.

In the future, it would be desirable to repeat a similar study over a continuous period using more camera traps. This would allow for better and easier identification of individual golden jackals, and closer monitoring would give a more complete representation of the social interactions between individuals, especially during the mating season and while raising pups. In addition, using molecular techniques would allow us to confirm the relatedness of the group members.

Author Contributions

Jonas Custers: conceptualization (lead), data curation (lead), formal analysis (lead), investigation (lead), methodology (lead), project administration (equal), resources (lead), validation (equal), visualization (equal), writing – original draft (lead), writing – review and editing (lead). Jennifer Hatlauf: funding acquisition (equal), methodology (equal), supervision (lead), writing – review and editing (equal). Sem van der Niet: data curation (equal), investigation (equal), project administration (equal), validation (equal), writing – review and editing (equal). Beatriz Tintoré: conceptualization (equal), methodology (equal), project administration (equal), supervision (equal), writing – review and editing (equal). Anastasia Miliou: conceptualization (equal), project administration (equal), resources (equal), supervision (equal), validation (equal).

Conflicts of Interest

The authors declare no conflicts of interest.