Introduction

The placement of semiochemicals on the substrate to communicate with conspecifics, defined as scent marking, is a common trait among mammals (Johnson 1973). Scent marking typically entails faeces deposition, urine and/or dragging a particular region of the body with scent-producing glands (e.g. the anogenital region) on the substrate (Eisenberg and Kleiman 1972; Ferkin 2001; Ferkin and Johnston 1995), and seems to serve multiple and, occasionally, overlapping functions (Becker et al. 2012; Ferkin et al. 2001, 2004). Scent marking is used mainly for territorial defence, as well as mate advertisement and self-advertisement (Johnston 1983; Thomas and Wolff 2002). Given the high costs of scent marking for both scent donors and receivers (Gosling et al. 2000; Hughes et al. 2010; Koivula and Korpimaki 2001; Mason et al. 2005), it is not surprising that animals may modulate their scent-marking behaviour, reducing costs when scent marking is not profitable (Roberts et al. 2001). Notwithstanding, the plasticity of mammals’ scent-marking behaviour at the species level is poorly studied. Only a few studies have focused on the intra-specific plasticity of scent-marking behaviour; these studies show that scent-marking behaviour may respond powerfully to changed social conditions (Becker et al. 2012; Le Roux et al. 2008). For instance, in yellow mongooses (Cynictis penicillata) the population density seems to affect the subordinates’ intensity of scent-marking behaviour for territorial defence purposes: territorial scent marking was intensive in high-density populations and almost absent in low density populations (Le Roux et al. 2008). Another study has shown that pair bonding reduces the scent-marking behaviour of California mouse males (Peromyscus californicus) towards unfamiliar females (Becker et al. 2012).

In the current study, we have investigated the scent-marking behaviour and its function(s) in pair bonded Cabrera voles (Microtus cabrerae Thomas, 1906). The Cabrera vole is a rare rodent endemic from the Iberian Peninsula with patchy and highly fragmented populations (Mira et al. 2008; Palomo and Gisbert 2002; Pita et al. 2006, 2007). It is considered a habitat specialist occurring in patches of tall and dense grass cover associated with high soil moisture where individuals are grouped in subpopulations (Fernandez-Salvador 1998; Landete-Castillejos et al. 2000; Pita et al. 2006, 2007; Santos et al. 2005, 2006). The highest densities that have been reported for Cabrera vole populations range between 115 individuals/ha (Fernandez-Salvador et al. 2005) and 250 individuals/ha (Landete-Castillejos et al. 2000). However, recent surveys using a spatially explicit capture-recapture method adequate for population density estimation in the patchy fragmented populations (Borchers and Efford 2008) of Cabrera voles found a density of about 8 individuals/ha in Southern Portugal (Mira et al., unpublished). Cabrera voles often move along above-ground runways, which they build through the surface vegetation. Although reproduction may occur throughout the year, it peaks in winter/spring and declines in the driest months (Fernandez-Salvador et al. 2005; Ventura et al. 1998). This species is presumably monogamous with a K-type reproduction strategy (Fernandez-Salvador 1998; Fernandez-Salvador et al. 2001; Ventura et al. 1998) where individuals seem to establish pair bonds with opposite-sex conspecifics, as suggested by several ecological evidence: (1) they form breeding pairs, with relatively high residence of over 10 months, where one male shares the nest with one female and pups, not showing aggressive reactions to them (Fernandez-Salvador 1998; Fernandez-Salvador et al. 2001, 2005); (2) they present a reduced sexual dimorphism in body dimensions and relatively small testis of adult males (Ventura et al. 1998) and (3) have been recently described a low inter-sexual variation of home-range sizes coupled with a high inter-sexual overlap (Pita et al. 2010). Non-paired Cabrera voles of both sexes seem to use scent marking primarily to increase the likelihood of finding a mate (Gomes et al. 2013), by advertising their presence to the opposite sex. In fact, when exposed simultaneously to substrates previously marked by conspecific males, substrates previously marked by conspecific females and unmarked substrates both sex voles showed a preference to scent mark with urine on substrates, which were marked previously by donors of the opposite sex (Gomes et al. 2013).

Scent marking as a mate-finding strategy should have a great value to non-paired individuals. Once the so-called “monogamous breeding pairs” have been formed the value of the opposite sex selective marking decreases as voles are no longer searching for a potential mate. Accordingly, we hypothesised that pair bonded Cabrera voles may not use scent marks for mate advertisement. In addition, pair bonding may increase aggressive behaviour towards unfamiliar individuals particularly associated with territorial defence (Bowler et al. 2002; Carter and Getz 1993; Wang et al. 1997; Winslow et al. 1993). Therefore, we hypothesised that scent marking of pair bonded Cabrera voles may serve for territorial defence as has been suggested for pair bonded California mouse (Becker et al. 2012).

To disclose the influence of pair bonding on scent-marking behaviour, and assess the scent-marking function(s) in pair bonded Cabrera voles, we conducted two distinct tests. First, we compared scent-marking behaviour of both non-paired and pair bonded voles towards a clean substrate. In the subsequent test, we assess the scent-marking behaviours in response to scent marks from donors of each sex. If Cabrera voles stop using scent marks to increase the mate-finding likelihood when pair bonded, we expected that pair bonded voles do not show a preference to scent mark the substrates previously marked by opposite-sex conspecifics and/or over-mark opposite-sex scent marks. Additionally, if they use such scent marking for territorial defence, we predicted that they will exhibit a preference to scent mark the substrates previously marked by same-sex conspecifics and/or over-mark same-sex marks.

Methods

Study animals

We used wild voles from southern Portugal caught throughout all the year (Gomes et al. 2013). It was assumed that the season in which the animals were caught did not influenced significantly their scent-marking behaviour, given that Cabrera voles seems to interrupt their breeding activity only in particular dry conditions related to the lack of green herbs (Fernandez-Salvador 1998; Ventura et al. 1998), which is not the case of the trapping sites at the time voles were caught. The criteria to select animals to be tested in the experiments were adult, non-pregnant for females and being caught in different patches. None of the females used in the experiments showed visible signs of oestrus. Once in captivity, voles were housed individually in glass cages (70 × 40 × 40 cm) without contact to conspecifics for at least 7 days (non-paired phase). After being tested as non-paired, each animal was pair-housed with a random selected opposite-sex individual in new cage (140 × 40 × 40 cm) (pair formation phase; Becker et al. 2012). The voles were considered pair bonded after 9 days since pairing procedure. The cages were prepared with soil and vegetation to mimic their original habitat and kept in a room at 18–21 °C under a long photoperiod (14:10 h, L/D). All tests were carried out during daytime. Food (diverse grass species collected daily, supplemented with apples) and water were provided ad libitum (Fernandez-Salvador et al. 2001). All procedures were done with the permission of Portuguese governmental authorities (i.e. Institute for Nature and Biodiversity Conservation) and conformed to the guidelines approved by the American Society of Mammalogists for the use of wild mammals in research (Gannon et al. 2007), and to the guidelines for the treatment of animals in behavioural research (ASAB 2006). Moreover, at the end of the experiments, the animals were released at the same location where they had been caught.

Experiment 1

The first experiment was carried out to test the differences in scent marking among non-paired and pair bonded voles in response to a clean substrate. For both conditions, 15 adult males and 15 adult females were used. In each test, one individual was placed in a small glass arena (63 × 29.7 × 40 cm) with the bottom covered with a clean paper sheet (substrate); scent-mark behaviour was allowed for 30 min. Then, the individual was removed from the arena and returned to the respective home cage, and the substrate dried for 10 min. The dried substrate was then observed in a dark-room and scent marks were identified and accounted under UV light (Desjardins et al. 1973; Ferkin et al. 2001). Each mark was traced lightly with a pencil and then photographed with a 5 Mpx Digital Camera for assessment of scent-marked areas using Quantum-GIS software (QGis 2011).

According to studies in other vole species, different kinds of marking are distinguished by the different characteristics they show under UV light: urine appears as bluish spots, faeces appears as darkish brown streaks or boli, and anogenital secretion marks typically appear as silver-white slivers between 0.5 and 3 cm long, and 0.1 to 0.3 cm wide (Wolff et al. 2002). To prevent contamination with human scents disposable latex gloves were used to manipulate all substrates. Scent-marked substrates were stored in clean plastic pockets at −20 °C until use in the second experiment.

Experiment 2

The second experiment was carried out to assess the function of scent marking in pair bonded Cabrera voles by investigating the scent-marking behaviour of each sex when exposed simultaneously to clean substrates, substrates previously scent marked by males and substrates previously scent marked by females.

The two pre-marked substrates, one from each sex, obtained from non-paired individuals in experiment 1, and an equal-size unmarked substrate (i.e. clean paper sheet) were placed side-by-side at the bottom of a glass arena (89.1 × 63 × 40 cm). In each replicate, one paired male or one paired female was placed in the centre of the arena inside a bottomless mesh cage for 2 min. Then, the cage was removed using a remote device to avoid visual contact with the observer allowing the tested animal to scent mark for 30 min. New marks in each test substrate (clean, female donor and male donor) were identified and analysed following the procedure described in experiment 1. Initial and secondary marks were distinguishable by the colour difference under UV light and overlapping areas between initial and secondary marks were identified and measured (Ferkin and Pierce 2007; Johnston 1999; Wolff et al. 2002). The experiment was replicated with 15 paired adults of each sex.

Data analysis

Scent-mark analysis was carried out with frequency data for faecal boli (Thomas and Kaczmarek 2002; Thomas and Wolff 2002; Wolff et al. 2002) and total marked area (i.e. the sum of the areas of all marks) for urine. Urine mark area was preferred to frequency since the shape and size of such marks were highly variable. Wilcoxon signed-rank tests were used to compare scent-marking frequency (faecal boli) and total urine-marked area over clean substrate between non-paired and paired condition (experiment 1). Scent-marking data from each sex in experiment 2, i.e. faecal boli frequency and total urine area, were compared among the three types of substrate (i.e. male pre-marked; female pre-marked; non-pre-marked) using the Friedman's rank test.

To determine the occurrence of over-marking (Ferkin and Pierce 2007), the observed overlapping areas were compared with those expected randomly, for each combination (sex of tested animal: sex of initial marks donor) using Wilcoxon signed-rank tests. Expected overlapping areas were accounted by multiplying the proportion of substrate area occupied by the initial marks with the area of all secondary marks placed over the same substrate (Gomes et al. 2013).

All statistical procedures were performed using the statistical package PASW version 18 (SPSS for Windows 18 2009).

Results

Experiment 1

When exposed to clean substrate, no significant differences were observed in the number of faecal boli (z = −0.12, n = 15, P = 0.906) and urine-marked area (z = −0.28, n = 15, P = 0.776) for males in both pairing conditions (Fig. 1). A similar result was observed for females (faecal boli: z = −0.44, n = 15, P = 0.659; urine-marked area: z = 0.17, n = 15, P = 0.865; Fig. 1). No anogenital marks were detected.

Fig. 1
figure 1

Mean (±SE) number of faecal boli and total urine-marked area produced over 30-min trials by 15 females and 15 males over clean substrate, in both testing conditions (non-paired versus paired)

Full size image

Experiment 2

Pair bonded males and females did not show a significant preference to scent mark with urine a specific substrate type (males: χ 2 = 0.13, df = 2, n = 15, P = 0.936; females: χ 2 = 2.13, df = 2, n = 15, P = 0.344; Fig. 2). Similarly, no significant differences were detected in the number of faecal boli placed across the three substrate types (males: χ 2 = 0.26, df = 2, n = 15, P = 0.880; females: χ 2 = 0.038, df = 2, n = 15, P = 0.981; Fig. 2).

Fig. 2
figure 2

Number of faecal boli and total area (mean ± SE) of urine secondary scent marks located over male pre-marked substrate (male donor), female pre-marked substrate (female donor) and non-pre-marked substrate (control), for paired males (n = 15) and paired females (n = 15) in experiment 2

Full size image

No over-marking was detected between pre-existent and secondary scent marks (scent donor vs. test subject; males vs. males: z = −0.73, n = 15, P = 0.460; males vs. females, z = 0.39, n = 15, P = 0.695; females vs. males, z = −1.59, n = 15, P = 0.112; females vs. females, z = 0.72, n = 15, P = 0.470).

Discussion

The scent-marking behaviour is often associated with a strategy to attract or find mates, commonly, by counter-marking (i.e. the placement of scent marks near to or on top of marks previously left by a conspecific) or/and over-marking the marks of opposite-sex conspecifics (Ferkin et al. 2001, 2004; Ferkin and Pierce 2007; Gomes et al. 2013). Both counter and over-marking of the opposite-sex marks may be carried out to advertise individual presence and availability to mate and/or reveal their interest in previous scent donors, thus increasing the likelihood of finding a mate (Ferkin et al. 2001, 2004; Hurst and Rich 1999; Woodward et al. 2000). Furthermore, in some rodent species females increase their likelihood of mating with a “good quality” male by selectively counter-marking some male marks rather than others (Hurst and Rich 1999; Johnston 1999). In non-paired Cabrera voles the counter-marking of opposite-sex marks was documented, which suggested that they use scent marks for mate-finding (Gomes et al. 2013). However, in pair bonded voles, the benefits provided by the scent marking required for mate-finding function may be lower or inexistent, since they have a stable relationship with an opposite-sex conspecific and may not breed outside the pair as suggested for the congener prairie vole (Aragona and Wang 2004). In these circumstances, it is likely that the high costs (e.g. energetic costs and predation risk) carried by scent marking (Gosling et al. 2000; Koivula and Korpimaki 2001; Mason et al. 2005) largely overcome the benefits resulting in a selection against scent marking to finding a mate. Thus, pair bonded voles may suppress their mate-finding scent marking. Consistent with this idea, pair bonded Cabrera voles from both sexes urine-marked the different substrate types in similar ways, without preferentially counter-marking or over-marking opposite-sex conspecifics’ marks. A recent study also supports this hypothesis for male California mice, where male marking behaviour towards unfamiliar females was lower in pair bonded males than in non-paired males (Becker et al. 2012). Furthermore, physiologically, the association between pair bonding and scent marking suggested by our results is not totally surprising, since arginine vasopressin, a neuropeptide found in most mammals, regulates scent-marking behaviour (Albers and Bamshad 1999; Ferris et al. 1984; Goodson and Bass 2001) and is simultaneously a critical mediator of partner preferences during pair bond formation (Cho et al. 1999; Cushing et al. 2001; Goodson and Bass 2001; Insel and Hulihan 1995; Williams et al. 1994; Winslow et al. 1993; Young and Wang 2004).

Pair bonded voles usually show preferential contact with a familiar partner rather than unfamiliar ones (Carter et al. 1986, 1992; Williams et al. 1992, 1994). Besides the partner preference, pairing also leads to an increase in aggressive behaviour toward unfamiliar conspecifics in both sexes (selective aggression; Bowler et al. 2002; Wang et al. 1997; Winslow et al. 1993), that serves to guard mate and territory (Carter and Getz 1993). Scent-marks provide a physical evidence of dominance and are commonly used to delineate territorial limits and to intimidate or deter conspecific intruders (Brown and McDonald 1985; Gosling 1982; Hurst 1993; Roberts and Dunbar 2000). Pair bonded voles may use scent marking to defend the territory and their investment in mates and offspring by counter and/or over-marking same-sex scent marks (Humphries et al. 1999; Hurst 1990; Johnston et al. 1995, 1997; Wolff 1993). However, our results failed to support the territorial defence hypothesis in pair bonded Cabrera voles, since voles did not preferentially counter or over-mark with urine the substrate previously scent marked by a same-sex conspecific. In fact, our results do not support any of the classical hypotheses for the use of scent marking in pair bonded Cabrera voles. The pair bonded Cabrera voles also did not avoid overlapping urine marks of other individuals, suggesting no scent-blending avoidance strategy required for the self-advertising function (Gomes et al. 2013; Thomas and Kaczmarek 2002; Thomas and Wolff 2002). Likely, the reasons why both unpaired and pair bonded voles do not seem to use scent marking for territorial defence and self-advertising functions are the same. As has been proposed previously for non-paired individuals, the patchy and highly fragmented distribution of Cabrera voles, wherein encounters between conspecifics are rare, makes scent marking for both territorial defence and self-advertisement functions useless (Gomes et al. 2013).

In conclusion, the current findings suggest the intra-specific plasticity of Cabrera voles’ scent-marking behaviours related with pair cohabitation. The differences verified on scent-marking response to the opposite-sex marks between paired and non-paired voles of both sexes suggest that pair bonding may phase out their investment towards finding a new mate, particularly in mate-advertisement strategies. Despite these differences, pair bonding did not affect scent marks’ type and the amount of marks placed by voles over clean substrates. Cabrera voles do not appear to use scent marking for any of the most usually considered functions; however, that does not exclude its use for other communication purposes. Perhaps, there may be other functions that we are unable to identify with the present paradigm. It should be borne in mind that our results were produced in laboratory conditions, and that they may not resemble natural habitat conditions. These conditions may have affected the behaviour of tested animals to some extent, so further studies on semi-natural conditions are required to validate the current results. Nevertheless, we believe that Cabrera voles’ scent-marking behaviour was not significantly influenced by fear or stress promoted by the actual laboratorial set-up, given that: (1) during the initial experiments we observed tested voles and detected a low frequency of fear or stress behaviours, and a great amount of time spent in exploration-like activities; (2) under the same test set-up, non-paired Cabrera voles responded differently to scent marks from distinct donors (Gomes et al. 2013), suggesting that these laboratorial conditions did not prevent voles to respond to pre-existing scents.

Lastly, the results of the present study also give strong support for the existence of a monogamous mating system in the Cabrera vole, a trait that has been suggested for this species (Fernandez-Salvador 1998; Fernandez-Salvador et al. 2001, 2005; Pita et al. 2010; Ventura et al. 1998).