Introduction

Scent marking plays a fundamental role in communication in many mammal species, allowing information such as the species, sex, individual identity, health, and reproductive status of a signaller to persist in the environment (Johansson and Jones 2007; Wyatt 2003). Many mammalian taxa use urine as the primary source of scent, because it is a waste product and so its use as a scent mark does not incur additional metabolic costs (Eisenberg and Kleiman 1972; Ralls 1971), but understanding usage of urine as scent marks in wild animals can be difficult.

Direct observation of urine marking provides the best opportunity to study patterns, but data are particularly difficult to obtain for nocturnal or crepuscular free-ranging species. Whilst faecal marking is easier to study because faeces are often displayed on or in latrines (e.g. badgers Meles meles; Roper et al. 1986) or on prominent substrates (e.g. wolves Canis lupus on plants; Barja et al. 2005), urine marking is more problematic. Urine marking can be studied where it leaves a visible trace (e.g. stains: Smith et al. 1989), where habituated individuals are directly observed (Begg et al. 2003; Jordan 2007) or where enhanced with biomarkers (e.g. Hutchings et al. 2001), but urine marking is studied less commonly than faecal marking for cryptic species. The use of video cameras can provide a non-invasive method to allow greater study of fine-scale behaviours, e.g. fox–badger interactions (Macdonald et al. 2004), and therefore could be used to monitor urine marking. In addition, long-term monitoring can be beneficial to examine seasonal patterns that otherwise might be missed using techniques limited to particular seasons, e.g. snow tracking (Henry 1980; Goszczyński 1990).

Red foxes (Vulpes vulpes) regularly urine mark objects and have been reported to use urine both as a “book keeping system” and potentially as territorial markers (Baker et al. 2000; Henry 1977, 1980; Goszczyński 2002; Macdonald 1979a). However, their nocturnal and cryptic behaviour makes it extremely hard to establish the role of urine marking in foxes. Previous studies in captivity (Blizard and Perry 1979; Macdonald 1979a), occasional field observations (Henry 1980; Goszczyński 1990; Macdonald 1979a) and experimental field studies (Arnold et al. 2011; Whitten et al. 1980) have highlighted the importance of urine in the communication and social systems of foxes. Whilst red foxes are generally regarded as territorial (e.g. Goszczyński 2002; Lloyd 1980; White and Harris 1994) and recent experimental evidence suggests that females are less territorial than males (Arnold et al. 2011), understanding how this translates into variation in urine marking rates is unclear.

We studied urine marking patterns of wild red foxes at a feeding site at an urban–rural interface in Hampshire, UK. Using year-round, night-time video recording, we aimed to assess how urine marking rates varied across the year in relation to key seasonal life history events, and whether these patterns differed between sexes.

Materials and methods

Study site and video setup

The study was conducted within a largish garden in the New Forest, Hampshire. The surrounding habitat was fairly widely spaced detached housing adjacent to open heathland. Foxes regularly attended a feeding site approximately 15 m × 20 m wide of mainly rough lawn backed by shrubs (Fig. 1). Food, mostly left over from domestic use, was placed each evening on a patio. Foxes generally arrived through a gap under a fence, though other routes were sometimes used.

Fig. 1
figure 1

Schematic representation of the layout, indicating placement and direction of cameras (C1 and C2) and the 6 infrared lamps (L1-L6). Foxes usually accessed from a hole under a fence and typically marked near food remains or at three locations: by the feeding site (U1), at shrubs near the outbuilding (U2), and near the main access (U3) usually on departing

Full size image

Two infra-red/colour video cameras, with microphones and with focal length set to about 35 mm, recorded the area from approximately 1 h before sunset throughout each night. Cameras were located near the food, so as to maximise detail when foxes were present, but the depth of field allowed observation of distant activity. Six infra-red lamps provided even illumination (Fig. 1). There was continuous recording from the cameras to two DVD-recorders and the results were monitored each following morning. Passages of interest were copied to discs for permanent retention and to facilitate confirmation of fox identities. Cameras and feeding site were used over 1 year before the start of data collection to allow optimisation of methodology and habituation of the foxes to the set-up.

Data collection

Individual foxes usually have a distinctive combination of morphological features such as white tag, ear and facial scarring, ear spots, or black colouration (e.g. Fig 2a–e). Through this method regular attenders are easily identifiable, though those visiting infrequently could be mis-identified as new individuals.

Fig. 2
figure 2

Illustrations derived from daylight or infra-red video footage, showing different identifiable features: a ear damage, b short tail without white tag, c long tail with white tag, d discernible penis and e permanently kinked tail and identifiable nipples (i.e. a female)

Full size image

Where possible, a fox’s sex was established by observing the genitalia or conspicuous nipples (Fig. 2d, e), and this was facilitated by placing the cameras only 25 cm above ground level. Recordings of urination sometimes showed its anatomical origin. Females having bred were identified from clear signs of lactation. General morphology and size of the fox are not dependable as sole criteria of gender. Similarly, posture of urination alone is not an adequate indication because females, as well as males, sometimes “cock” a leg over vegetation and males frequently squat to urinate on the ground, although usually not as low as females (see also Henry 1977, 1980). So a combination of appearance and behaviour was often necessary for reliable identification of gender, which was relatively easy as most individuals attended throughout a substantial sequence of visits. Juveniles (<12 months old) and individuals with unclear gender were excluded from analyses.

From the DVD recordings, we noted when each fox urinated, its physical state (including any injuries) and behavioural or other events (including defaecation, food caching, and intra- and inter-specific interactions). Often, there was a fairly sharp distinction between foxes rigorously abstaining from urination even during prolonged presence and those urinating several times during even brief visits. Urine marking was most often at food remains and at three other sites, shown in Fig. 1. Urinations typically occupied ≤2 s and a range of postures was used (see also Henry 1977). Data were documented according to whether an individual was seen to urine mark or to completely abstain during attendance that night.

Data analysis

Data were collated for each individual observed attending the feeding site on a daily basis. Urination was categorised as binomial (0 = no urination, 1 = urination observed), irrespective of the number of urine marks observed. To assess sex and monthly patterns of urination, we carried out three separate binomial general linear mixed-effect models with urination (0, 1) as a dependent variable. In the first model, we included month (January–December), sex (male, female) and their interaction (sex × month) as fixed factors to assess sexual differences in year-round marking pattern. In the second model, we included fortnightly intervals (0–4) during January and February to cover the peak period of female oestrous (the last 2 weeks of January), sex and their interaction (sex × period) to assess any change in urine marking during the mating period. In the third model, we again assessed urine marking at fortnightly intervals during January and February, but this time included only those females later seen to lactate (so presumed to have mated).

In all models, we included fox and year as random factors. Models were run using the lme4 package (Bates and Maechler 2009) in R 2.14.1 (R Development Core Team 2011). For results and discussion, months are grouped into seasons: spring (March–May), summer (June–August), autumn (September–November) and winter (December–February). We assessed repeatability of individual and yearly urine marking rates using the rptR package (Nakagawa and Schielzeth 2010) run in R, and quantified individual heterogeneity (Hamel et al. 2012) in urine marking patterns, from the lme4 package.

Results

Data collection

We collected data across 63 months (1 January 2007–31 March 2012) from 34 foxes (23 females, 11 males) comprising 5,020 observations (3,629 female, 1,391 male) during about 22,000 h of recording (Table 1). Urine marking occurred during 24.6 % of the total observations.

Table 1 Monthly median (lower quartile/upper quartile) percentage urine marking rates of male and female red foxes Vulpes vulpes
Full size table

Seasonal and sex differences in urine marking rate

Urine marking varied significantly across months, being lower in June–August (Tables 1, 2; Fig. 3). Also, there was a strong interaction between month and sex (Table 2); marking rates were generally higher for males than females in August–November, but not significantly in the rest of the year (Tables 1, 2), except during March when males urine marked significantly more frequently than females (Tables 1, 2).

Table 2 Binomial GLMM results for the relationships between urine marking and month (January–December), sex (male, female) and their interaction
Full size table
Fig. 3
figure 3

Median + IQR monthly urine marking rates for both sexes combined. As most of the lower interquartile ranges (like the medians for May–August) are zero, only the upper interquartiles are shown

Full size image

We did not find any significant effect of period, sex or their interaction during the period before, during and after the mating season (Table 3; Fig. 4). Even when restricting these data to just females later observed lactating, there was still no evidence of any increase in urine marking rate during the mating period (Table 3).

Table 3 Binomial GLMM results for the relationships between urine marking and (A) sex (male, female), period (period 1: 1–14 January, period 2: 15–31 January; period 3: 1–14 February, period 4: 15–28 February) and their interaction and (B) period for females later observed lactating
Full size table
Fig. 4
figure 4

Median + IQR individual urine marking rates at fortnightly intervals during January and February. Females (white) and males (grey) are shown separately. As most of the lower interquartiles are zero, only the upper interquartiles are shown

Full size image

Heterogeneity and repeatability of urine marking patterns

There was significant repeatability of individual urine marking rates (R logitM = 0.42, SE = 0.07, P < 0.001), and significant, but much lower, repeatability of urine marking rates between years (R logitM = 0.12, SE = 0.06, P < 0.001). This between-year effect could be largely driven by high between-individual marking rates (σ = 2.89, SD = 1.70) and therefore which foxes were present in each year would determine how repeatable urine marking was between years.

Discussion

Previous studies on red foxes have not provided a year-round view of red fox urine marking, their results being limited to the winter period (Goszczyński 1990; Henry 1980) or very limited sample sizes (n = 2) across the year (Macdonald 1979a). Though our study is limited to a single site, our results provide the first comprehensive overview of red fox urine marking rates for both sexes across the year.

We demonstrated that urine marking rates decline during summer, which supports previous data (Macdonald 1979a). Reduced urine marking in summer does not necessarily mean that foxes reduce territorial defence or reduce responsiveness to scent marks, as red fox territories are stable and mutually exclusive year-round (Henry et al. 2005; White et al. 1996), with little seasonal difference in inter-group interactions (White and Harris 1994). However, it may be predicted that the need to urine mark and demarcate a territory depends on season. For example, during spring, parents and other group members are involved with cub rearing (Robertson et al. 2000). Fox cubs have restricted movement near the den and can be vulnerable to infanticide (Robertson et al. 2000; Vergara 2001), so maintaining urine marking rates in spring may reduce or discourage intruders. Similarly, juveniles (6–12 months) mainly disperse during autumn and winter, and adult males during the mating period (January–February) may make movements away from their resident territories to seek extra-pair copulations (Iossa et al. 2008; Soulsbury et al. 2011; White and Harris 1994). Juvenile dispersal also continues into the spring (Baker et al. 2001; Soulsbury et al. 2011). In consequence, the risks of intrusion and therefore the benefits of urine marking are highest during autumn, winter and spring. It is only during summer that the risk of intrusions into territories declines. During summer, cubs become nutritionally independent from their parents and range increasingly further from the natal den, but still remain in the natal territory (Baker et al. 2001; Robertson et al. 2000; Soulsbury et al. 2008). Additionally, intrusions by non-resident adults are also lowest during summer (Baker et al. 2001). This reduced likelihood of intrusions by either adults or juveniles is probably the primary reason for reduced urine marking during summer.

Our results also reinforce the view that there are sex differences in territorial behaviour of foxes (Arnold et al. 2011). Urine marking rates by males were generally higher throughout the late summer and autumn in comparison to females, but not significantly during most of winter and spring. Previous sex comparisons have been limited to the winter period when, in line with our results, no sexual differences were found (Goszczyński 1990; Henry 1980). In red foxes, male territory size depends on male body mass and pressure from surrounding males, whereas female territory size depends on the male with which she associates (Iossa et al. 2008). Boundary pressure from competing males, coupled with higher pressure from male dispersers, is probably the main reason for higher urine marking rates in males.

Though males show a seasonal pattern similar to females’, their higher urine marking rates probably reflect the greater risk of male intrusions in general, particularly during August–November when male juvenile dispersal is at its highest (Baker et al. 2001; Harris and Trewhella 1988; Soulsbury et al. 2011). Male urine marking was also higher than females’ during one month in spring, March, the main period of female parturition. Pre-parturition increases in male urine marking have been observed in other species, such as arctic foxes Vulpes alopex (Kullberg and Angerbjörn 1992). Females are generally restricted to near the den site during the period of parturition (Doncaster and Macdonald 1997; Iossa 2005), so it is possible that males increase their own marking to reinforce territory boundaries when cubs are very vulnerable.

There is mixed evidence for increased urine marking during the mating period in red foxes. Henry (1980) found no temporal or sexual differences in urine marking rates before, during and after the mating period, whereas Goszczyński (1990) found differences in urine marking rates in open but not forested habitat. More detailed counts of a single female’s urine marking rates showed a sudden decline after mating, though mean monthly rates showed little evidence of change (Macdonald 1979a). We did not find any evidence of increased urine marking during the mating period (late January to early February) or any differences between sexes. As foxes urine mark frequently, perhaps there is no value in further increasing already high marking rates. Instead, additional forms of communication, such as increased vocalisations, may be important during the winter to communicate reproductive activity (Newton-Fisher et al. 1993). Also, urine marking rates may be directed at specific areas of the territory (e.g. Goszczyński 1990). Our method examines urine marking as present/absent on a daily basis, but this may mask changes in frequency of urine marking within days (Macdonald 1979a). Hence, a more detailed investigation is needed into urine marking rates during the mating period.

Urine marking rates by individuals showed a moderate level of repeatability, but there was also considerable variation in urine marking rates between individuals. Several factors may alter individual urine marking rates, such as dominance status; dominants in many species mark more than subordinates (Asa et al. 1990; Gese and Ruff 1997; Sillero-Zubiri and Macdonald 1998). Red foxes have a flexible social system with group sizes that vary from 2 to 10 individuals (Baker and Harris 2004; Cavallini 1996). A group is usually an extended family comprising a dominant pair and non-dispersing offspring (Iossa et al. 2009). These non-dispersing offspring can be of either sex but are most commonly female (Baker and Harris 2004; Macdonald 1979b). In our study, we have limited information on social group size, composition or dominance status, and nor is it generally known whether dominance status influences urine marking rates in red foxes, though it has been suggested (Macdonald 1979a). Whether dominance status among red foxes plays an important role in driving between-individual differences in marking rates and whether this may be related to sex and seasonal patterns is a key topic for future study.

More generally, our results are similar to those for studies in other canids. For example, feral dogs Canis familiaris show a similar seasonal pattern of urine marking, with males marking more frequently than females and with lower marking rates during summer (Pal 2003). However, in contrast to our results for the red fox, urine marking rates are typically higher in the mating season in other canid species (wolf: Asa et al. 1990; coyote Canis latrans: Gese and Ruff 1997; arctic fox: Korhonen and Alasuutari 1995; feral dog Canis familiaris: Pal 2003). Our results demonstrate that red fox urine marking rates show patterns similar to other canid species’, and provide a basis for further examination of urine marking and its role in this species.

Conclusions

We show that urine marking rates decline during summer and that male foxes urine mark more frequently than females during summer and autumn. We did not find any increase in urine marking during the mating season.

We have also shown that the use of video cameras can be valuable in analysing hard-to-observe behaviours such as urine marking, and has provided useful data on urine marking patterns that will aid further elucidation of the role of scent in the territorial and communication systems of red foxes. Future work will need to expand monitoring across multiple sites within territories to test whether the patterns we observe at one site are found across the whole territory.