Breeding status affects territorial behaviour to simulated intruders in captive wolves

Grey wolves (Canis lupus) rely on olfaction to communicate and acquire information about their environment. Olfactory signals delivered via urine, faeces, and glandular secretions, serve critical functions in reproduction, social interactions, and territorial defence. We studied the role of olfactory signals in mediating inter-pack conflict, a leading cause of mortality in free-ranging wolves, by exposing zoo-housed wolves to the scent of an unfamiliar individual, mimicking an intruder. We predicted breeding adults to exhibit higher investigatory and territorial behaviours, accompanied by an increased patrol effort in response to this simulated intruder than other individuals due to their dependence on territorial integrity and reproductive investment. Consistent with this prediction, breeders displayed significantly higher rates of investigatory, territorial and locomotor behaviours to conspecific than heterospecific urine than non-breeders. These findings highlight the specialised role that breeders play in wolf packs, particularly in relation to territorial defence, and demonstrate the deeply conserved patterns of territorial behaviours in wolves, even in populations that have been living under human care for a long time. Our findings have implications for theories of mammalian olfactory signalling and potential applications for conservation, including scent-based bio-fencing, to reduce human–wildlife conflicts.

Wolves (Canis lupus) and other canids possess a highly developed olfactory system, which they rely upon to acquire information about the environment and to communicate with conspecifics (Ortiz-Leal et al., 2022; Ortiz-Leal et al., 2024). Olfactory communication is primarily mediated through scent-marking via urine, faeces, and glandular secretions (Peters and Mech, 1975; Mech and Boitani, 2003; Brennan and Kendrick, 2006; Clapperton et al., 2017; Torres et al., 2023). Scents are long-lasting individual markers with functions in reproduction, rank relations, and territorial defence (Paquet, 1991; Vilà et al., 1994; Sillero-Zubiri and Macdonald, 1998; Mech and Boitani, 2003; Brennan and Kendrick, 2006; Torres et al., 2023).

Canid scent-marking is well studied, with chemical compositions of scents analysed in African wild dogs (Lycaon pictus) (Apps et al., 2012), red foxes (Vulpes vulpes) (Jorgenson et al., 1978; McLean et al., 2021), coyotes (C. latrans) (Schultz et al., 1988), dingos (C. l. dingo) (Robley et al., 2015; Lepan et al., 2023), Iberian wolves (C. l. signatus) (Martín et al., 2010; Barja Núñez et al., 2023) and grey wolves (C. lupus) (Raymer et al., 1986; Wolfram, 2013). As a behaviour, scent-marking has been described primarily from the perspective of the sender, in terms of spatial variables (deposition sites) (Rothman and Mech, 1979; Mech and Boitani, 2003; Zub et al., 2003; Barja Núñez et al., 2004, 2005; Claase et al., 2022, 2024), marking postures (Peters and Mech, 1975; Mertl-Millhollen et al., 1986; Barja Núñez and De Miguel, 2004), and social variables (Gese and Ruff, 1997; Gese, 2001; Mech and Boitani, 2003; Parker, 2010; Cassidy et al., 2015; Smith et al., 2015; York et al., 2019).

Comparably less is known, however, about how receivers respond to scents. Some work has shown how prey species avoid or show defensive behaviour to wolf urine (Apfelbach et al., 2005; Osada et al., 2014, 2015; Kashiwayanagi et al., 2017; Wikenros et al., 2017; Haswell et al., 2018; Ginkel et al., 2019; van Beeck Calkoen et al., 2021; Sunde et al., 2022), and there is some general information available of how individuals respond to conspecific cues (Arnold et al., 2011; Shivik et al., 2011; Jackson et al., 2012; Ausband et al., 2013; Anhalt et al., 2014; Robley et al., 2015; Ghaskadbi et al., 2022; Hansen et al., 2024), with relatively little experimental evidence (Barja Núñez and De Miguel, 2004; Wolfram, 2013; Stępniak et al., 2025).

Olfactory information about the presence of unfamiliar individuals may be more important for some individuals than others. Wolves are seasonal breeders with typically only one breeding pair per pack. During the breeding season, competition for mates and territories among conspecifics increases and is accompanied by elevated territorial marking (Peters and Mech, 1975; Jedrzejewski et al., 2001). Breeders may be especially interested in confronting territory intruders if they are about to have vulnerable offspring (Zub et al., 2003; Smith et al., 2015). Intrusions occur regularly from roaming young individuals from neighbouring packs in search of a mate and suitable territory to form their own pack (Morales-González et al., 2022). These recurrent events present a considerable challenge for established territory holders and a major cause of wolf mortality (Mech, 1994; Mech and Boitani, 2003; Smith et al., 2015; Cassidy et al., 2017; SunderRaj et al., 2022). Here, scent-marking likely operates a cost-effective strategy to minimise territorial disputes, by deterring nearby conspecifics and roaming individuals from trespassing and potentially fatal confrontations (Gosling, 1982; Kaufmann, 1982; Robley et al., 2015; Torres et al., 2023). Urine marking, the main modality for delineating territorial boundaries, is performed primarily by breeding individuals with stereotyped postures, such as raised-leg urination (RLU) and flexed-leg urination (FLU) (Peters and Mech, 1975; Mertl-Millhollen et al., 1986; Barja Núñez and De Miguel, 2004), suggesting that olfactory communication is especially important for them. Field studies have shown that wolves do not contribute equally to inter-pack conflicts: participation in territorial aggression depends on age, sex, and individual traits. In particular, Cassidy et al. (2017) showed that males chase intruders more than females, wolves with a grey coat are more aggressive than black ones, and male wolves get involved in conflicts more with increasing age; proposing that pack members differ in their involvement in defending resources and territory. Wolf packs, in other words, are not homogenous units, but groups of individuals with different social-ecological interests, also when it comes to commitment to territorial defence (Smith et al., 2015; Cassidy et al., 2017).

Given the above considerations, we predicted variation in individual behavioural responses to odour-simulated intrusions. Free-ranging wolves have been observed to respond to experimentally placed conspecific scent-marks (Stępniak et al., 2025), although the role of individual identity, breeding status, and or pack composition has not been investigated. We addressed this gap by simulating the presence of a conspecific intruder to breeding and non-breeding subjects. Overall, we predicted that subjects would show different behavioural responses to conspecific versus heterospecific urine. We further predicted that breeders would be more reactive than non-breeders, showing more territorial and investigatory behaviours along with differences in patrolling behaviour. In the wild, scents from an unfamiliar conspecific can lead to boundary avoidance (Mech and Boitani, 2003; Ausband et al., 2013; Schlägel et al., 2017), but captive individuals without previous intergroup experience may instead approach the scent out of investigative interest.

Materials

For the conspecific treatment, commercial wolf urine was purchased from ® Leg Up Enterprises (Lovell, ME, USA) as in Ausband et al. (2013). The commercial wolf urine used in this study has been successfully employed in previous research (Ausband et al., 2013), however, no further information was available concerning the sex, age or breeding status of the donor or donors.

As the heterospecific treatment, female human urine was collected cumulatively over a 24-hour period under controlled conditions using pre-sterilised glass jars and was immediately frozen at -18 °C. On October 3, 2024, two replicates of each treatment were prepared for each zoo by pipetting 3 mL aliquots of wolf and human urine into separate labelled plastic tubes (Corning 15 mL centrifuge tube, 430791), then stored frozen at -18 °C.

Inspired by Robley et al. (2015), eight scent stations were fabricated, consisting of an aluminium plate (7.7 cm x 9.7 cm x 0.03 cm), to prevent any chemical interaction with urine, with a central depression designed to hold 3 mL of urine to be exposed for behavioural testing (Figure 1A). This plate was affixed to a steel base, which was attached to an aluminium tube and positioned 30 cm above the ground to simulate RLU. To protect the plate from the influence of the weather, custom-made birdhouse-style wooden protection was built for each of the devices (Figure 1B).

Study sites and subjects

Data was collected in four zoos in Switzerland: Tierpark Goldau, Tierpark Bern, Biel Zoological Parks, and Les Marécottes Alpine Zoo. A total of five enclosures (corresponding to five wolf packs) were tested, each varying in size and number of individuals housed (Table 1). In total, thirteen wolves from these five packs were observed for this study, comprising six breeders and seven non-breeders. Three packs consisted of breeding pairs, while two consisted of a single adult and their offspring (Tables 1, 2). The mean age of breeders was 5.67 years, whereas non-breeders averaged 5.43 years.

“Breeders” were defined as adult wolves currently in a bonded, mixed-sex dyad with breeding potential (i.e., individuals not surgically or chemically neutered or otherwise prevented from mating). In contrast, “non-breeders” were defined as wolves not belonging to a bonded pair, either due to their subordinate status within the pack or because they had lost a previously bonded mate, or because they had been neutered, thus lacking the potential to reproduce.

The individuals in Bern, Goldau, and Les Marécottes were European grey wolves (C. l. lupus), while the ones in Biel belonged to the Mongolian grey wolf subspecies (C. l. chanco). All individuals were born in captivity and had lived in their respective home enclosures for at least 2 years (6 months in case of one female individual, ID13).

Ethics statement

The experiments in this study were conducted in accordance with the guidelines outlined by the Swiss Animal Welfare Act (TSchG, SR 455, Kap.3, Art. 22). An animal testing licence was obtained from the relevant authorities before conducting this study (Nr. BE89/2023).

Test procedures

The locations of scent stations were strategically determined through a week-long preliminary observation period. This involved using 24-hour motion-triggered cameras and consulting animal keepers’ expertise to identify wolf-frequented areas along the enclosure periphery. Scent stations were positioned along sites typically used by wolves for travel and territorial marking, such as junctions of established trails, close to conspicuous objects and away from feeding areas (Rothman and Mech, 1979; Mech and Boitani, 2003; Zub et al., 2003; Barja Núñez et al., 2004; Barja Núñez, 2009), thereby simulating natural marking. Following previous studies, four scent stations were installed 30–40 cm from the enclosure fences (Wolfram, 2013) (Figures 1D, E). Where enclosure design or zoo regulations limited placement, adjustments were made accordingly. A pilot study was conducted three months before the main experiment to validate the setup and materials (SI).

Treatments

The purchased wolf urine served as a treatment to test the territorial behavioural reactions, while the human urine was used as a control. The latter was used to ensure that behavioural reactions to the wolf urine were due to intraspecific interest and to control for the novelty effect of the stimulus (Ryon et al., 1986; Petak, 2010).

The exposure to the treatments was randomised for each enclosure. To avoid confounding wolves’ reactions and mimic their natural behaviour of refreshing territorial marking (Peters and Mech, 1975; Mech and Boitani, 2003; Ausband et al., 2013), new treatments were introduced at regular five-day intervals. With each subsequent exposure, the previous treatment was not removed but allowed to degrade naturally until the end of the experiment. This approach aimed to simulate realistic territorial marking by wolves more accurately (Mech and Boitani, 2003; Ausband et al., 2013) and prevent potential reactions caused by the sudden disappearance of a stimulus and the device.

Devices containing replicates of the same treatment were positioned at least 10 meters apart to avoid potential habituation to the stimulus in the same location. Furthermore, a minimum separation distance of 5 meters was maintained between devices with different treatments to minimise cross-treatment interference. This spacing ensured that wolves’ behaviour near one device was primarily in response to that specific treatment, rather than being influenced by nearby scents from other treatments.

Behavioural data collection

To monitor behavioural reactions to urine marks, a single camera trap (Browning Recon Force Elite HP4 Trail Cameras - detection range and angle 24 m, 53.5°) was placed behind each scent station. The cameras were mounted at a height of 40–50 cm above the ground, aligned with the nose level of wolves (Figure 1D). The camera traps were set to record 30-second videos during the day and 20-second videos at night, upon motion activation. Each camera trap operated continuously for 24 hours a day over a period of 5 days. To minimise false triggers, vegetation around the device was removed. Videos in which the individual could not be clearly identified were discarded.

Wolves in each zoo were recorded for 20 consecutive days; the observation period spanned from April 3 to June 3, 2024, encompassing a total of 55 hours of recorded footage.

Behavioural analyses

Six specific scent-marking and investigatory behaviours (Table 3; Figure 2) were deliberately selected and modified after Wolfram (2013) to address the research question. These focal behaviours recorded are typically displayed in the context of territory maintenance (Sillero-Zubiri and Macdonald, 1998; Wolfram, 2013; Schlägel et al., 2017) and confirmed to occur during preliminary observations of the study population.

Behavioural occurrences were counted using a discrete event sampling method as in Wolfram (2013). Each instance of a specific behaviour was recorded as a separate occurrence. If a wolf interrupted a behaviour (e.g., stopped sniffing, turned its head, or walked briefly) and then resumed the same or initiated a different behaviour, it was counted as a new occurrence. This method allowed to capture the total number of times wolves engaged in each behaviour, rather than the duration of behaviours or continuous activity. The occurrence of these behaviours was recorded per each single individual.

To assess the reliability of the data coding and to control for observer bias, a second observer who had no information about the treatment condition of each video coded 20% of the recordings. Interobserver reliability was subsequently calculated using percent agreement and intra-class correlation coefficient (ICC). Agreement was 98.1% for behaviours (ICC = 0.8) and 81.2% for passages (ICC = 0.9), indicating a score of interobserver consistency between good and excellent (Koo and Li, 2016).

In addition, the number of passages of a wolf by a scent station was counted. A passage was defined as one complete crossing of the path directly in front of the scent station within the camera frame, moving from one side to the other (either right to left or vice versa). Each crossing was counted as a distinct event, regardless of the time elapsed between crossings by the same individual.

Statistical analyses

To provide a more standardised comparison across individuals, rates of behaviours and passages were calculated by dividing the total number of behaviours and passages exhibited by each individual wolf by the total number of camera activations for each treatment and individual. The results of these calculations are referred to in the text as the “rate of behaviours” and the “rate of passages”. In this way, it was possible to account for differences in enclosure sizes and individual overall activity. In addition, for better visualisation, results are shown on a logarithmic scale.

Statistical analyses were conducted using two distinct models to assess: (1) the behavioural (scent-marking and investigatory) responses and (2) the passage rates of wolves in relation to the two treatments.

The behavioural responses and the visits to the scented area (passages) of the wolves to the commercial conspecific urine were recorded as counts. All recorded behaviours were grouped together for analysis due to the small number of occurrences in each category on its own (Supplementary Figure 1).

To account for repeated measures and count data, a Generalised Linear Mixed Model with negative binomial error distribution (NB-GLMM) was fitted (glmer.nb function, lme4 package (Bates et al., 2015)) in R (v4.3.2; R Core Team, 2023). A full-null model comparison was conducted using a likelihood ratio test to evaluate the overall significance of the predictors.

The behavioural model was constructed as follows: the response variable was the count of observed behaviours for each individual wolf (BehavCount). To test how breeding status might influence wolves’ reactions to the treatments, the interaction effect of treatment (i.e. wolf urine vs human urine) and whether or not the wolf belonged to a breeding pair (BreedingPair) was used as the test predictor. To account for the hierarchical structure of the data, individual (ID) and zoos (Zoo) were included as random effects. An offset term was incorporated to account for the number of camera activations by each individual wolf (IDCamActiv). Moreover, the model’s goodness of fit was assessed by checking for overdispersion, which was not the case.

On the other hand, in the model that considered the passages, only the response variable was changed to the number of passages made in proximity to the scented area.

In both models, membership in a breeding pair was chosen as the test predictor over sex, as it was expected to have a stronger influence on territorial behaviours and number of passages, since breeding pairs are primarily responsible for territory marking compared to other pack members (Macdonald, 1980). Accordingly, the inclusion of BreedingPair as a predictor was found to improve model performance compared to using Sex as an alternative predictor. Sex influence on both behavioural and visitation response was also tested, resulting in no significant difference between males and females (Supplementary Figures 2, 3). Likelihood ratio tests were performed to compare models with different predictor combinations (Treatment interacting with either BreedingPair or Sex). The final models (mod_be and mod_ps) were determined to be the best fitting based on Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), log-likelihood, and deviance.

Pairwise post-hoc comparisons of Estimated Marginal Means (EMMs) were conducted using emmeans package (Lenth et al., 2024) for both models to assess significant differences across treatment levels within each breeding condition category (Supplementary Tables 1, 2). P-values were adjusted using Tukey’s method for multiple comparisons.

Finally, models’ stability was assessed by comparing model estimates through case deletion diagnostics (Supplementary Figures 4, 5), and confidence intervals were obtained through parametric bootstrapping (Supplementary Figures 6, 7; Supplementary Tables 3, 4).

Investigatory and scent-marking behaviours

Individual responses varied considerably, with some wolves showing increased behavioural rates (investigatory and scent-marking) in response to wolf urine while others exhibited decreased rates compared to the human urine control (Figure 3). These individual differences may reflect a series of factors such as previous experiences with conspecifics (ID9 and ID4) (Table 2) (Christensen et al., 2016; Christensen and Radford, 2018), frequent experiences with visitors bringing domestic dogs to the wild park (Stępniak et al., 2025), group size effects (Cassidy et al., 2015; Ghaskadbi et al., 2022) and changes in hormonal status due to reproduction and gestation (Mertl-Millhollen et al., 1986; Smith et al., 2015). Six individuals showed decreased or minimal change in response rates (ratios ≤ 2: IDs 1, 2, 3, 6, 10, 11), while six others exhibited notably increased responses to wolf urine (ratios > 2: IDs 4, 5, 7, 8, 9, 12) as compared to human urine. Additionally, one individual (ID13) showed responses exclusively to wolf urine, displaying no targeted behavioural response to human urine. Importantly, results showed distinct patterns between breeding and non-breeding pairs in response to urine treatments (Figure 4). Specifically, there was a significant interaction between treatment and breeding pair membership (z = 2.703, P = 0.007), demonstrating that breeding status significantly affected the territorial reaction to conspecific scent-marks. Estimated marginal means indicated that breeding pairs exhibited a substantially stronger behavioural response to wolf urine (42.6 ± 16.7 behaviours) compared to human urine (3.3 ± 1.5 behaviours), after accounting for other predictors in the model. Pairwise comparisons confirmed this difference to be highly significant (response ratio = 0.077, p < 0.001), indicating that wolf urine elicited ~13 times more behaviours than human urine in breeders. In contrast, non-breeders displayed relatively consistent behavioural rates across both treatments, with 7.72 (± 2.55) and 4.96 (± 1.91) behaviours per individual across the 5-day treatment period for wolf and human urine, respectively, showing no significant difference between treatments (ratio = 0.642, P = 0.819). While all behaviours were pooled for the primary analysis due to low individual frequencies, the detailed breakdown reveals that, breeders exhibited sniffing as the most common behaviour (73% of the total), followed by rubbing (20%), raised-leg urination and defecation (2%), and flexed-leg urination and scraping (1%), respectively. In contrast, non-breeders predominantly exhibited exploratory sniffing with minimal marking behaviours (96% SN, 3% FLU and 1% DF) (Supplementary Figure 1).

Visit to scent station

Analysis of passage rates at scent-stations revealed a significant effect of breeding pair membership (z = 4.835, P < 0.001) and a minimal, non-significant decrease in passage rates with wolf urine compared to human urine (z = -0.083, P = 0.934) (Figure 5). The estimated marginal means showed that breeding pairs maintained consistently higher passage rates regardless of treatment, with 312 (± 48.3) passages when exposed to human urine and 308 (± 49.8) passages with conspecific urine. In contrast, non-breeders showed lower rates, with 118 (± 19.5) passages with human urine exposure and 116 (± 19.7) passages with wolf urine. While there was no significant difference between treatments within each group (ratio = 1.013, P = 0.100), there was a significant difference in passage rates between breeders and non-breeders for both human (ratio = 0.378, P < 0.001) and wolf urine treatments (ratio = 0.378, P < 0.001).

The present study aimed to investigate how zoo-housed wolves of known breeding status react to conspecific urine scent-marks in comparison to a heterospecific control stimulus (human urine). We predicted more scent-marking and investigatory behaviours in response to conspecific than human urine, with stronger reactions in breeders than in non-breeders. We further predicted either (1) reduced visits to the scented area, consistent with avoidant reaction of territorial frontiers, or alternatively, (2) an increase in visits to the scented area in line with higher interest in and efforts to maintain territorial boundaries in response to a potential intruder. In both scenarios, we expected the breeders to show the most pronounced spatial response.

In line with our predictions, conspecific urine elicited higher investigatory and scent-marking behaviours compared to the control condition. Our results also show that there is a significant difference in behavioural reactions and interest in conspecific scent-marks between breeding and non-breeding wolves. As expected, breeders exhibited significantly higher behavioural responses to conspecific urine compared to human urine, while non-breeding individuals showed minimal responses to both treatments. Breeders also maintained significantly higher passage rates at scent stations compared to non-breeders, regardless of treatment. Contrary to both predictions, passage rates did not differ significantly between conspecific and human urine treatments in either group.

Responses to territory intrusions are influenced by the position of a wolf within the pack (Desjardins et al., 2008; Cassidy et al., 2017; Christensen and Radford, 2018). The social structure of wolf packs is highly organised, with individuals fulfilling precise family-style roles that shape both their level of interest in intruders (Cassidy et al., 2017; Christensen and Radford, 2018) and their involvement in territorial defence in intergroup encounters (Desjardins et al., 2008; Cassidy et al., 2015; York et al., 2019). The importance of role differentiation in shaping territorial responsiveness was strongly supported by the study’s results: breeders reacted significantly more frequently to conspecific urine than non-breeders (Figure 4), displaying both higher behavioural reaction rates and a wider range of scent-marking behaviours (Supplementary Figure 1). This breeding status-dependent pattern is consistent with Stępniak et al. (2025), who similarly showed that breeding wolf pairs expressed a wider range of behaviours, both marking and investigative, towards conspecific scent, while non-breeding juveniles showed mainly investigatory responses. The heightened and differentiated responsiveness likely reflects breeders’ fundamental role as the core unit of an established pack, responsible for territory defence and social stability (Gese, 2001; Mech and Boitani, 2003; Brainerd et al., 2008; Borg et al., 2015; York et al., 2019). Higher rates of territorial responsiveness in breeders may reflect the greater benefit of detecting and responding to intruders, as observed in other canids where dominant breeders exhibit elevated scent-marking rates (e.g., coyotes, Gese and Ruff, 1997; Ethiopian wolves, Sillero-Zubiri and Macdonald, 1998; African wild dogs, Apps et al., 2022; Claase et al., 2024a). Indeed, breeding pairs bear the greatest costs to territorial intrusions: they risk not only their territory but also their reproductive opportunities and the resources needed to raise their offspring (Smith et al., 2015; Cassidy et al., 2017). This territorial investment was further evidenced in territorial patrols, with breeding pairs’ significantly higher passage rates to scent stations compared to non-breeders. Additional, albeit anecdotal, support for the effect of breeding status comes from observations of individual ID13 tested while she underwent different social roles in the pack. During pilot testing in December 2023, when ID13 occupied a subordinate non-breeder position in pack A (ID13.1), preliminary results showed no response to conspecific urine. However, after becoming the breeding female in pack E (ID13) in February 2024, the same individual exhibited substantially stronger reactions to conspecific urine compared to the control treatment (Supplementary Figure 8). This behavioural response persisted in follow-up testing six months later (June 2024) (Supplementary Figures 8, 9), after acclimation to the new enclosure, suggesting the change was not just due to novelty effects (Petak, 2010).

On the contrary, non-breeding wolves showed reduced responses to both treatments, primarily engaging in investigation rather than other marking behaviours (Supplementary Figure 10), and passing by the scent stations fewer times than breeders, regardless of treatment. Notably, both packs lacking breeding pairs (B, C) included an adult individual that had previously held a breeding role but had lost its partner. Other studies showed that widowed breeders are more tolerant and likely to accept another individual of the opposite sex as a potential mate, possibly explaining the lack of a strong territorial response (Stahler et al., 2002; Mech and Boitani, 2003; Brainerd et al., 2008). The loss of a breeding partner can lead to pack dissolution, with remaining wolves either re-establishing the territory or being displaced by neighbouring packs (Brainerd et al., 2008). The absence of a breeder reduces a pack’s social cohesion (Brainerd et al., 2008; Borg et al., 2015; Cassidy et al., 2023) and territorial defence capacity (Brainerd et al., 2008), thus weakening its territorial claims. The low territorial response of non-breeders may reflect a shift from territorial defence towards re-establishing social bonds and the pack’s reproductive unit.

In non-breeding wolves (i.e., only groups among which offspring were present), age-related differences in behavioural responses could not be investigated due to limited sample size. Nevertheless, it is important to highlight that age is inherently linked to breeding status: young wolves remain subordinate and non-breeders within their native packs, acquiring the status of breeder only at the age of 1 or 2 years after dispersal and pair formation (Mech and Boitani, 2003; Gable et al., 2019; Morales-González et al., 2022), a process that does not occur in captivity.

While breeding status appears to be a key factor in determining different behavioural responses, considerable individual variation was observed, ranging from very low to extremely high reaction rates. This variability might reflect differences in prior social experiences with conspecifics (Christensen and Radford, 2018), dog presence (Stępniak et al., 2025), or group size effects (Cassidy et al., 2015; Ghaskadbi et al., 2022). Captive wolves are unlikely to have had previous experience of territorial conflicts with intruders, and most of the tested individuals had never experienced conspecific fights, potentially attenuating their responses. Notably, higher response rates were observed in D, particularly from ID9 that was separated from ID4 following dominance-related aggression, potentially increasing general territorial sensitivity and vigilance. Additionally, among the tested zoos, policies regarding dog access varied, with zoo B being the only facility that allowed dogs to enter the park. As Stępniak et al. (2025) showed that wolves respond to domestic dog urine markings, the regular exposure to dog scent at zoo B may have habituated these wolves to unfamiliar canid odours and potentially attenuated their response to the experimental wolf urine stimulus.

Individual response rates varied substantially, ranging from 0.2% (non-breeder female, human urine: 1 behaviour in 553 camera activations) to 25% (breeder female, wolf urine: 9 behaviours in 36 camera activations), reflecting the combined influence of breeding status and stimulus type on territorial responsiveness. Despite this variation, behavioural responses remained relatively low overall (201 behaviours across 6181 camera activations, 3.3%), as most observations resulted in simple passages without overt investigatory or scent-marking displays. This limited response might be explained by our use of single, isolated scent-marks, whereas natural wolf territories are typically defined by multiple scent-marks creating clear boundaries (Peters and Mech, 1975; Mech and Boitani, 2003; Gable et al., 2022, 2025). Furthermore, while we predicted either avoidance or increased visits to the conspecific scent compared to the human scent, passage rates did not differ significantly between treatments. Previous research has shown that single scent-marks elicit minimal responses in African wild dogs unless overmarked by a breeder (Parker, 2010), and that simulating the presence of both breeding individuals may be necessary for effective territorial signals (Rothman and Mech, 1979; Ausband et al., 2013). Besides the quantity of scents-marks, the quality of the urine used as a treatment may have influenced the wolves’ responses in this study. Since canid scent signals vary substantially with reproductive state, sex, and social rank (Raymer et al., 1984; 1986; Asa et al., 1990; Wolfram, 2013), the unknown origin of the commercial wolf urine represents a limitation. Ausband et al. (2013) suggested that clear and consistent marker identity is essential for a strong territorial signal, and ambiguity can weaken its deterrent effect over time. In our study is unclear whether the stimulus was perceived as a mark from a dominant breeder or a subordinate individual, or even whether it originated from one or multiple donors. Future studies should use urine of verified origin, ideally from individuals of known sex and breeding status to better control for stimulus identity and disentangle the contributions of chemical composition from social context.

Furthermore, it’s known that smaller groups may react differently to intruder threats than bigger groups (Cassidy et al., 2015; Ghaskadbi et al., 2022; Claase et al., 2024b; Hansen et al., 2025). In this study breeder groups consisted of only two individuals (breeding pair), while the non-breeders numbered between three and four individuals. Unfortunately, it was not possible to account for this difference in group size. Nevertheless, some research suggests that the number of markings is not correlated with the number of individuals in a group (Peters and Mech, 1975; Zub et al., 2003), as the breeding pair always remain the main depositors. Moreover, the effect of group size may also vary depending on the perception of threat, whether the marking is by a neighbour or a stranger (nasty neighbour vs dear enemy effect) (Ghaskadbi et al., 2022; Wyrick et al., 2025), and the individuals composing the pack (Cassidy et al., 2015). Because of these diverse factors potentially affecting individual responses, it is not possible to determine which and to what extent they contribute to the overall variability and only hypotheses can be made; further research focussing specifically on each variable would be valuable.

The analysis of behavioural responses to simulated wolf urine marks revealed two principal findings: First, the experimental approach successfully simulated raised-leg urination, eliciting clear and functionally relevant behavioural reactions in zoo-housed wolves. Second, these responses, most notably investigatory behaviours, were significantly stronger in response to conspecific urine than to human urine among breeding individuals, although they did not translate into detectable changes in spatial movement. Importantly, these pronounced responses were observed even in captive wolves with no prior exposure to intruding conspecifics, indicating that reactions to territorial scent-marks are not experience-dependent but represent a strongly conserved tendency in wolf socio-ecology.

The costs and benefits associated with territorial defence and engagement in potential conflicts with conspecifics vary among pack members according to their social role, with breeders facing both higher gains and greater risks (Christensen and Radford, 2018; York et al., 2019). The socially differentiated responses observed here are consistent with this framework and suggest that reactions to conspecific scent-marks are socially mediated, with breeding status being a key determinant of response intensity. These findings reinforce the role of chemical signals as socially embedded cues rather than simple environmental stimuli. In territorial species, scent signals can function both as attractants and deterrents for conspecifics (Shivik et al., 2011; Ausband et al., 2013; Campos et al., 2017). In highly social and group-living species such as wolves, responses to these chemical signals are likely shaped by breeding status and pack structure, thus making it essential to account for social context when developing deterrents and scent-based conservation and management tools. The present results support the central role of the breeding pair as the primary unit and main territorial receiver and signallers of a pack, demonstrating that artificial conspecific marks can effectively mimic biologically meaningful territorial signals, even in captive settings.

Territorial marking affects both resident and dispersing wolves (Rothman and Mech, 1979; Stahler et al., 2002; Imbert et al., 2016), with dispersers accounting for a disproportionate number of depredation events (Imbert et al., 2016; Mayer et al., 2022). Expanding future work to include both resident breeders and dispersing individuals, combined with ecologically realistic marking simulations, would improve our understanding of the deterrent potential of artificial territorial signals. Nonetheless, it should be borne in mind that deterrent strategies likely require regular reinforcement, as their effectiveness may decline over time due to habituation (Ausband et al., 2013; Appleby et al., 2017; Lorand et al., 2022). Combining these cues with other non-lethal measures such as fladry, fencing or guardian dogs may help mitigate habituation and increase their applicability to management contexts and local conservation needs (Miller et al., 2016).

This study provides a methodological foundation for experimentally simulating territorial scent-marking, upon which future research can build. Systematic monitoring of behavioural reactions to chemical signals remains essential for decoding the information conveyed by scent-marks. Further investigation is needed to understand how breeding status shapes chemical communication in wolves, including how marking postures may affect chemical composition of urine marks and how such differences modulate territorial responses. Although to date, the application of conspecific odour as deterrents in wild settings remains limited – with Appleby et al. (2017) identifying only two such cases - chemical communication-based approaches show promise as integrative tools for wildlife conservation and population management (Campbell-Palmer and Rosell, 2011).

In conclusion, our findings show that simulated conspecific scent-marks elicit clear and functionally relevant behavioural responses in captive wolves, with more frequent responses in breeding individuals than in non-breeders, supporting the central role of breeders in territorial defence and inter-specific communication. These findings highlight the potential of olfactory cues as tools for both understanding and managing wolf behaviour in the wild.