Repellent effect of synanthropic house mouse urine odor on small forest mammals

Abstract

In this study, we examined the effect of synanthropic house mouse (Mus musculus) urine odor on catching probability of small mammals to live traps. We conducted a series of field experiments in August 2016 and 2017 in a natural forests of the northwestern Moscow Region (Russia). Small mammals were trapped at two 4-ha fields using capture-mark-recapture technique by setting 200 live traps (100 points, 2 traps per point) within each field. One trap in each pair was odorless (control) with bait only, whereas the other one was odor-baited with 20 μL of the urine of a synanthropic house mouse. Further analysis was based on the data collected from 2 rodent species (bank vole Myodes glareolus, herb field mouse Apodemus uralensis) and 3 shrew species (common shrew Sorex araneus, Laxmann’s shrew Sorex caecutiens, and Eurasian pygmy shrew Sorex minutus). As a result, only bank voles significantly avoided odor-baited live traps. Using generalized linear mixed models, we showed that the choice of a trap by bank voles depended on their age, whereas the probability of repeated capture to a certain live trap was related to their prior experience. We discuss the possible role of components of synanthropic house mouse urine in the population management of exoanthropic small mammals.

Various factors determine trap capture frequencies of small mammals, including population density, individual activity patterns, trap and bait type, season and weather, habitat, or trap location specificity as well as small mammals species-specific, sex- and age-related reactions to traps which usually depend on a neophobic intensity; olfactory cues, and many others (Zhigarev 1993; Kotenkova 1995a; Tasker and Dickman 2002). Trap odor greatly influences trappability (Kotenkova 1995b; Oleinichenko 2015). Given all the above, it is crucially important to 1) reveal if odors can be used as attractants/repellents for rodents (Stoddart 1982; Drickamer 1984); 2) analyze how a trap odor affects the estimation of population size and evaluation of the distribution of small mammals (Stoddart 1982; Wuensch 1982); and 3) test the results of laboratory experiments on the response to olfactory cues under natural conditions (Musso et al. 2017). Odors that highly influence a trap capture success of rodents and shrews (except a bait odor) may be classified into 4 main categories, namely, conspecific, sympatric heterospecific of similar size and ecology, potential predator, and those of human origin. A small mammal’s response to conspecific or heterospecific odor left in a live trap depends on the species, sex, age, or social status of both the odor donor and the recipient (Kotenkova 1995a; Tasker and Dickman 2002). The preference for live traps with the conspecific odor in the breeding season can change to the opposite effect outside that season (Daly et al. 1978, 1980). Predator odors act as repellents for some small mammals including being applied to a live trap (Stoddart 1976, 1982; Dickman and Doncaster 1984), which is especially well expressed when the focal small mammal species is a natural prey for this predator (Dickman 1992; Drickamer et al. 1992). Moreover, an odor of a predator can induce male-biased litters in rodents, disrupt maternal behavior, inhibit research activity, and cause other physiological and behavioral responses (Voznessenskaya and Malanina 2013; Voznessenskaya 2014).

Apparently, not only predator excreta components can make such effects. Experiments have shown that the synanthropic house mouse (Mus musculus L., 1758) odor, when exposed to semi-synanthropic East European vole (Microtus levis [syn. Microtus rossiaemeridionalis] Miller, 1908), decreased the fertility among first mated females (Kotenkova and Osadchuk 2009). Voles of this species, inclined to the facultative synanthropy, and exoanthropic bank vole (Myodes (=Clethrionomys) glareolus (Schreber, 1780) avoided synanthropic house mouse odors more than those of exoanthropic steppe mouse (Mus spicilegus Petenyi, 1882) or European rabbit (Oryctolagus cuniculus L., 1758). Voles avoid the steppe mouse odor applied in 1 compartment of Y-shaped maze versus water in the other, but to a lesser extent than the house mouse odor in a similar experiment (Bazhenov et al. 2014).

The persistent and potent house mice odor unveils their presence for both people and predators, therefore should not be maintained by natural selection. At the same time, a pungent urine smell of synanthropic house mice could be a warning, or aposematic signal for competitors, adaptive to defend the territory from other rodents in a man-made environment (Bazhenov et al. 2014). Besides, the inhibitory effect on the reproductive behavior of other rodent species could be another reason for such pungent a odor to be maintained by evolution. In this case, there is a trade-off between the benefits of the resource monopoly and costs due to increased predator pressure. House mouse is typically synanthropic species (Kotenkova and Muntyanu 2007; Bobrov et al. 2008) that do not inhabit pristine or little disturbed forests of Moscow Region. The same tendency is revealed for at least Eastern Europe. In the Moscow Region house mice inhabit only disturbed wooded areas (park-like), exclusively in summer (Zhigarev 1993, 2004; Kotenkova and Muntyanu 2007; Karaseva et al. 2008; Zhigarev et al. 2018). Since house mice inhabit not only man-made environment, there comes a question of the repellent effect of their odor on other species.

In this study, we tested the hypothesis that the synanthropic house mouse odor can repel other rodent species and shrews. The aim was to analyze small forest mammals’ responses to the house mouse odor within live traps in natural forest habitats.

Materials and Methods

The field study was carried out in August 2016 and 2017, in pine-spruce forest near Chernogolovka town (Noginsk district, Moscow Region, Russia, 56.022569, 38.446812). Capture-mark-recapture techniques were used during four 12-day live trapping sessions CMR method allows collecting datasets of observations on the reaction of individually marked mammals related to their species, sex, age, and previous experience, for a long and enough for statistical analysis. Two 4-ha plots were divided into a 10 × 10 m square grid. Live traps were placed in a checkerboard pattern at 20 m intervals, so each plot contained 100 evenly distributed live trapping points simultaneously. To prevent small mammals’ addiction to live traps, each live trap was relocated every 3 days to a new point, 10 m from the previous one, with 20 m intervals between traps being unchanged (Zhigarev, 1990; 2004). In total, all 400 trapping points of 4-ha plot were covered in 12-day trial.

During the session, each trapping point contained 2 identical live traps with a treadle release mechanism, so 200 live traps were exposed at 100 points simultaneously and checked twice a day, in the morning and in the evening. Both traps were baited with daily changed rye bread dipped in unrefined sunflower oil and a 7 mm cube of foam rubber fixed on the inner wall. One live trap from each pair was used as control (“odorless”), whereas the other was experimental, scented with 20 μL of the synanthropic house mice urine, applied to the foam every day with a BIOHIT Proline dispenser (Biohit Proline, Finland).

The urine was obtained from sexually mature male and female house mice in the laboratory. Each animal was placed for an hour in a special cage (12 × 6 × 6 cm), made of metal wire, with a Petri dish under it. Fresh urine samples were filled into 1.5 mL plastic tubes (Eppendorf) and stored at −18 °C. Being thawed at room temperature just before usage, the urine sample was never refrozen and reused again. Urine samples from 5 individuals (both sexes) were mixed before applying to a trap for the imitation of a family group odor. Preliminary experiments demonstrated that in large open enclosures bank voles responded to odor-baited with 20 μL of the urine of synanthropic house mice (mixture of 2 males and 3 females) for many hours (Osipova et al. 2018). Laboratory house mice urine was shown to be chemically very similar to that of wild type (Mucignat-Caretta et al. 2010), therefore it can be used to imitate properly the odor of wild-type synanthropic individuals.

Prior to utilization, all traps were cleaned by the Karcher SC1.020 steam generator (Kärcher Futuretech GmbH, Germany) without any detergents. Traps with a captured animal or any sign of it (fecal, urine, etc.) as well as with insects, slugs, or frogs, were changed for a clean one and then thoroughly washed in the lab.

Trapped mammals were individually marked and treated according to the standard protocol (Karaseva et al. 2008) included the identification of species, sex, age stage (adult/subadult/juvenile), reproduction stage of females (pregnant/nursing/barren). Right after that, mammals were released within 1 m radius from the trap. All the dead animals were autopsied with an accurate determination of sex, age, and reproduction stage; also, standard measurements of body size were carried out.

A total of 866 individuals of 8 species were trapped during 9600 trap-days (Table 1). A total number of captures was 1048. Whenever, both live traps (control and experimental) captured animals simultaneously, they were excluded from further analysis.

Table 1.

The species composition and number of small mammals being trapped in 2016, 2017.

Species	♂ Adult	♀ adult	Subadult	Juvenile	Total number
Bank vole (Myodes glareolus Schreber)	31	29	79	24	163
Herb field mouse (Apodemus uralensis Pallas)	13	20	67	2	102
Northern birch mouse (Sicista betulina Pallas)	0	0	2	0	2
Striped field mouse (Apodemus agrarius Pallas)	1	1	1	0	3
Common shrew (Sorex araneus L.)	33	26	427	0	486
Laxmann’s shrew (Sorex caecutiens Laxmann)	0	1	42	0	43
Eurasian pygmy shrew (Sorex minutus L.)	3	5	57	0	65
Eurasian water shrew (Neomys fodiens Pennant)	1	0	1	0	2
Totally					866

Our field experiments were based on the assumption that in natural environment forest exoanthropic species of small mammals face synanthropic house mice odor for the first time, so the reaction, if appears, would be innate but not acquired.

The effect of the odor on trap choice was analyzed by χ²-test. The null hypothesis suggests random observed differences in trap choice (no odor effect), and equal rates of capture in experimental (scented) and control (odorless) live traps. The calculations were made using STATISTICA 10.0.

General linear mixed models (GLMM) were used to test an influence (single and mixed) of sex, age, and other parameters, including time of capture and the prior experience, that is, an effect of the previous capture on trap choice in small mammals. At first, in trap choice modeling, a trap type was used as a binomial dependent variable (0—control, 1—experimental). The fixed effects were sex, age, time of capture, the prior experience (the trap type of the previous capture), and their interactions. Random effects were an individual number of a captured animal and a year. Second, for recaptured individuals of each species, trap type change (1) or no change (0) were used as the binomial dependent variables, whereas sex, age, and the prior experience were independent variables. The Akaike’s information criterion corrected for small samples (AICc) was used to rank candidate models (Anderson et al. 1994; Burnham and Anderson, 2004). Models with Δ AICc ≤ 2 were considered well supported by the data. β-coefficients were used to evaluate the significance of factors. All data analyses were performed in R (The R Core Team 2019). The package lme4 (Bates et al. 2015) to construct the GLMM and the dredge function and the MuMIn package (Barton 2016) for model selection were used.

Results

We conducted annual small mammal counts since 2000 to date (Nurimanova, Alpatov, 2009; Zhigarev, Putilova, Alpatov, 2017). In 2016 and 2017, the abundance of rodents in the Region was relatively low, whereas that of shrews, was rather high. Three of the 8 trapped species were rare during the study (Sicista betulina, Apodemus agrarius, and Neomys fodiens) and were excluded from further analysis. The rest of the trapped exoanthropic forest species of small mammals M. glareolus, Apodemus uralensis, as well as Sorex araneus, Sorex Caecutiens, and Sorex minutes (Table 1), had high frequencies.

General characteristics of the response of small exoanthropic forest mammals to the synanthropic house mice odor

Species-specific response

Control (odorless) traps were more attractive (more than 50% of choices) for almost all small mammals’ species in the field experiment, except the common shrew, whose frequency of choosing odorless traps was 48.4% (Figure 1). However, only bank vole significantly avoided traps with the synanthropic house mice odor (χ² = 9.16, P = 0.002).

Figure 1.

The ratio (%) of small mammals trapped in odorless traps (black sector and percent shown) and traps with the house mice odor (white sector). 1: Myodes glareolus, 2: Sorex minutus, 3: Apodemus uralensis, 4: Sorex caecutiens, and 5: Sorex araneus. *The rate is statistically significant (χ² = 9.16, P = 0.002).

Sex- and age-specific response

In our study, the house mice’s urine odor had no discernible and significant impact on the responses of Herb field mouse, common shrew, Laxmann’s, and Eurasian pygmy shrews related to their sex and relative age (Table 2). As for bank voles, both males and females did not significantly respond to the heterospecific urine odor (the ratio was close to equal). At the same time, immature (subadult and juvenile) bank voles considerably and significantly avoided traps with the odor (2:1, χ² = 13.4, P = 0.0002, Figure 2).

Table 2.

The ratio (%) of captures in control/experimental live traps (first line) related to the sex and relative age of trapped small mammals. Second line—χ² and P-value (in bold if statistically significant), third line—total number of trapped individuals (N).

Sex and age
Species	Adult	Subadult + juvenile	♀ All	♂ All	♀ Adult	♂ Adult	♀ Subadult + juvenile	♂ Subadult + juvenile
Myodes glareolus	52.0/48.0 0.17; 0.68 N = 96	66.4/33.6 13.4; 0.0002 N = 125	60.6/39.4 4.65; 0.03 N = 104	59.8/40.2 4.52; 0.03 N = 117	52.1/47.9 0.08; 0.77 N = 48	52.1/47.9 0.08; 0.77 N = 48	67.9/32.1 7.14; 0.008 N = 56	65.2/34.8 6.39; 0.01 N = 69
Apodemus uralensis	55.6/44.4 0.67; 0.41 N = 54	50.0/50.0 0.00; 1.00 N = 84	55.2/44.8 0.73; 0.39 N = 67	49.3/50.7 0.01; 0.91 N = 71	61.0/39.0 1.98; 0.16 N = 41	38.5/61.5 0.69; 0.41 N = 13	46.2/53.8 0.15; 0.69 N = 26	51.7/48.3 0.07; 0.79 N = 58
Sorex Araneus	53.8/46.2 0.31; 0.58 N = 52	47.7/52.3 0.84; 0.36 N = 386	48.1/51.9 0.27; 0.60 N = 183*	45.7/54.3 1.21; 0.27 N = 162*	60.0/40.0 0.80; 0.37 N = 20	50.0/50.0 0.00; 1.00 N = 32	46.6/53.4 0.74; 0.39 N = 163*	45.0/55.0 1.29; 0.26 N =131*
Sorex caecutiens	100.0/0.0 1.00; 0.32 N = 1	50.0/50.0 0.00; 1.00 N = 38	42.9/57.1 0.29; 0.59 N = 14*	53.8/46.2 0.08; 0.78 N = 13*	100.0/0.0 1.00; 0.32 N = 1	0.0/0.0 N = 0	38.5/61.5 0.69; 0.41 N = 13*	53.8/46.2 0.08; 0.78 N = 13*
Sorex minutus	50.0/50.0 0.00; 1.00 N = 8	55.1/44.9 0.51; 0.48 N = 49	65.2/34.8 2.13; 0.14 N = 23*	55.0/45.0 0.20; 0.65 N = 20*	80.0/20.0 1.80; 0.18 N = 5	0.0/100.0 3.00; 0.08 N = 3	61.1/38.9 0.89; 0.35 N = 18*	64.7/35.3 1.47; 0.23 N = 17*

Note: Sex of some immature shrews (Sorex sp.) could not be determined, so they were excluded from marked (*) samples.

Figure 2.

The ratio (%) of bank voles Myodes glareolus of different age trapped in odorless (black segment and percent) and traps with the house mice odor (white segment). 1: immature (subadult + juvenile), 2: subadult, 3: juvenile, and 4: adult. The difference is statistically significant for: * χ²= 12.1; P = 0.0005; **χ²= 13.4; P = 0.0002.

Small mammals’ response related to their previous contact with the synanthropic species odor

First-time captured bank voles on average entered odorless traps a little more often (56.0%, P > 0.05; Figure 3A, upper block). In the subsequent captures the frequency of choice of odorless traps increased and became statistically significant (62.1%, χ² = 3.88, P = 0.049; Figure 3A, middle block). The experiment design does not allow us to confidently claim that all individuals get acquainted with the house mouse odor after the first capture. Those first captured in the control trap might have chosen it either randomly, as the first on their way, or consciously, after sniffing both traps. Though rodents have rather a subtle olfactory to smell senses at distance well, we do not have enough evidence to make a confident choice in favor of either explanation. The rodents who chose the scented trap at first capture (44%), had evidently got acquainted with the house mouse odor. The percentage of subsequent trapping in odorless traps for them raised to 68.3% (χ² = 5.49, P = 0.019; Figure 3A, bottom block), which indicated an obvious avoidance of the house mice odor by bank voles after the first contact with it. This analysis reflects the average results for the studied population. However, it is also important to find out how the prior experience of an individual mammal could affect its consequent choice.

Figure 3.

The choice of 3 abundant small mammal species (A, B, C) of odorless (black segment) or scented traps (white segment) at the first (upper) and the consequent (bottom) captures in the field experiment (2016, 2017). The percentage of odorless trap captures are shown. Statistically significant are marked: * χ² = 3.88, P = 0.049; ** χ² = 5.49, P = 0.019.

In our field experiment, each trapped small mammal was marked individually, allowing us to track the sequence of trap choices for a certain individual in case of recaptures, that is, to evaluate its prior experience. After the first capture in a scented trap, the probability of being trapped in the same trap type next time was only P = 0.3 against P = 0.7 for the odorless trap (differences are significant: χ² = 4.8, P = 0.028). That could be caused by the stress due to the previous capture, which resulted in changing the trap type. However, the analysis of the voles’ choice after trapping in a control odorless trap, reveals that individuals chose the same trap type repeatedly with the probability P = 0.67 (χ² = 4.0, P = 0.046). Thus, we can confidently conclude that the bank vole’s prior experience had a pronounced effect on the further avoidance of the house mice odor.

Our data imply but do not prove conclusively that the first capture influences the subsequent choice in Ural field mice (Figure 3B). Repeatedly trapped individuals of this species after the first capture in an odorless trap chose the same trap type next time with the probability P = 0.68 (though the difference was not significant χ² = 3.57, P = 0.059).

As a result, only bank voles clearly showed their avoidance of the house mice odor. Hence, first being trapped in the scented trap, next time this individual would be trapped in the odorless trap with a higher probability. Unlike bank voles, the common shrew did not show significant changes in trap choice after the first trapping (Figure 3C). The total number of Laxmann’s and Eurasian pygmy shrews’ recaptures was insufficient for the statistical analysis.

The effect of changeable characteristics (sex, age, trapping time, and the prior experience) on the trap choice (using GLMM)

Factors affecting the trap choice

Myodes glareolus

Two of the eight GLMM models, built to evaluate the influence of changeable characteristics, had Δ AICc < 2 (Table 3). According to them, the mixed effect of relative age, the prior experience, and sex of an individual influenced the voles’ trap choice. Immature individuals entered odorless traps significantly more often than adults (β = −0.76 ± 0.29, 95% CI −1.34, −0.19). Recaptured animals (i.e., with the prior experience) were trapped in odorless traps significantly more often (β = −0.68 ± 0.31, 95% CI −1.30, −0.09), whereas sex had no effect (β = −0.11 ± 0.28, 95% CI −0.67, 0.43).

Table 3.

Competing models of the effect of changeable characteristics (sex, relative age, trapping time, the prior experience, and their interactions) on the trap choice of four small mammals’ species (only models with Δ AICc ≤ 2 are given).

Model	K	AICc	Δ AICc	AICc weight
Myodes glareolus
Age + prior experience	5	297.828	0	0.520
Age + prior experience + sex	6	299.777	1.949	0.196
Apodemus uralensis
Const	3	197.158	0	0.346
Prior experience	4	198.735	1.577	0.157
Sex	4	198.802	1.644	0.152
Age	4	198.892	1.734	0.145
Sorex Araneus
Const	3	612.805	0	0.438
Age	4	614.142	1.337	0.224
Prior experience	4	614.178	1.373	0.220
Sorex minutus
Const	2	82.802	0	0.746

“Const”—intercept only model. K—number of parameters in model.

Apodemus uralensis

Four of 8 models had Δ AICc ≤ 2, including intercept only, age, the prior experience, and sex, separately (Table 3). The effect estimations of the last 3 parameters were statistically insignificant (age: β = 0.23 ± 0.36, 95% CI −0.67, 0.94; the prior experience: β = −0.26 ± 0.35, 95% CI −0.93, 0.53; sex: β = −0.24 ± 0.35, 95% CI −0.96, 0.65). Thus, none of them influenced Ural field mice trap choice.

Sorex araneus

Three of eight models had Δ AICc ≤ 2, including intercept only, age, and the prior experience, separately (Table 3). The effect estimations of the last 2 parameters were statistically insignificant (age: β = 0.25 ± 0.30, 95% CI −0.33, 0.83; the prior experience: β = 0.28 ± 0.34, 95% CI −0.39, 0.97). Thus, none of the studied factors influenced common shrews’ trap choice.

Sorex minutus

Intercept only model of 4 being built, had Δ AICc ≤ 2 (Table 3). Thus, the sex and age of Eurasian pygmy shrews had no significant effect on their trap choice. Almost the same was found for Laxmann’s shrew.

Factors affecting the probability of recapture to the same trap type

Myodes glareolus

Two of the eight models had Δ AICc ≤ 2 (Table 4). Individuals, first captured in scented traps, significantly more often changed the trap type for the next trapping, compared with those first captured in control odorless traps (β = −1.45 ± 0.49, 95% CI −2.46, −0.51). The sex effect was statistically insignificant (β = −0.34 ± 0.49, 95% CI −1.32, 0.63).

Table 4.

Competing models of the effect of sex, age, and prior experience of an individual on changing trap type for three small mammal species (only models with Δ AICc ≤ 2 are given).

Model	K	AICc	Δ AICc	AICc weight
Myodes glareolus
Prior experience	4	109.402	0.000	0.521
Prior experience + sex	5	111.258	1.855	0.206
Apodemus uralensis
Const	3	86.849	0.000	0.322
Sex	4	87.821	0.972	0.198
Prior experience	4	88.592	1.743	0.135
Age	4	88.692	1.843	0.128
Sorex araneus
Const	3	57.342	0	0.565
Prior experience	4	59.340	1.998	0.208

“Const”—intercept only model. K—number of parameters in model.

Apodemus uralensis

Four of 8 models had Δ AICc≤ 2, including intercept only, age, the prior experience, and sex (Table 4). The effect estimations of all 3 factors were statistically insignificant (age: β = −0.34 ± 0.53, 95% CI −1.64, 0.69; the prior experience: β = −0.40 ± 0.53, 95% CI −1.54, 0.55; sex: β = 0.64 ± 0.59,95% CI −0.44, 2.01).

Sorex araneus

Two of 8 models had Δ AICc ≤ 2, including intercept only, and the prior experience (Table 4). The effect estimation of the last factor was statistically insignificant (β = 0.46 ± 0.67, 95% CI −0.83, 1.94). Two other shrew species (S. minutus and S. caecutiens) had the same results.

Discussion

Chemical communication plays the leading role in conveying information for most species of small mammals, and the response to olfactory cues can be either innate or formed and modified by an individual experience (Doty 2010; Wyatt 2014; Surov & Maltsev 2016; Kotenkova et al. 2018a, 2018b; Zhigarev et al. 2018). Today, one of the most promising areas of research concerning olfaction is interspecies chemical communication. Leading researcher teams are now focusing on decoding danger signals, (“alarm pheromones”), particularly from predator (Apfelbach et al. 2005; Ferrero et al. 2011; Kondoh et al. 2016). Our study showed that a similar effect could be achieved on at least some rodent species by chemical cues of heterospecific small mammals, not predators. The urine odor of synanthropic house mice may act as a repellent, causing immature bank voles (subadult and juvenile) to avoid scented live traps. The trap choice of adult bank voles as well as other 4 studied species of exoanthropic forest small mammals were not affected by the house mouse odor in the field experiment. However, previous laboratory experiments with a Y-shaped maze revealed that adult bank voles nonetheless avoided the synanthropic house mice odor (Bazhenov et al. 2014).

In other experiments, carried out in a 120 m² enclosure, the bank vole’s responsivity toward traps depended on its social status in a group. Only low-ranking males avoided a trap with the synanthropic house mice odor (Osipova et al. 2018). Dominant individuals chose odorless and scented traps with equal frequency and stayed there for the same period of time (Osipova et al. 2018). Sometimes the social status of an individual is not easy to identify in the field. Probably, because of this we did not find a significant avoidance response to the mice odor in adult bank voles. At the same time, the odor effect on juveniles and subadults both in the laboratory and in the field could indicate higher repellent properties of the house mice urine odor for immature animals. Is such response innate or acquired? It can be confidently asserted that all rodents within the study plots had never encountered the odor of house mouse before the experiment, as they inhabit forests where no house mouse was found for the past 20 years. This is more likely to indicate an innate response of bank voles. From one point of view, it can be a reaction of voles to an unknown and pungent smell, that is, a kind of caution (neophobia), differently expressed by different small mammal species. In our field experiments, the prior experience effect among bank voles, that is, repeated contact with the odor was proved to enhance the after odor avoidance. Thus, an innate response is boosted but not weakened by the prior experience, which refutes the assumption that the avoidance reaction is related to the neophobic effect. Another explanation of repellency of the house mice urine odor could deal with its significance for an animal, for example, as an innate warning-related smell. Our hypothesis is that pungent house mice odor is aimed at repelling other rodent species from the habitat, well settled by house mice, such as human buildings. The pronounced avoidance of this odor among young bank voles is of particular interest, as they are more prone to disperse and occupy vacant areas, including buildings (Zhigarev 1997, 2005).

The pungent smell of synanthropic mice species, sensed by humans, is largely determined by sulfur-containing compounds (Mucignat-Caretta et al. 2010), particularly, 2-sec-butyl-4,5-dihydrothiazole (SBT) (Kwak et al. 2016). SBT is a pheromone stimulating male aggression (Novotny et al. 1985), and is also confirmed to be an alarm pheromone in house mice, excreted by both males and females in case of fear (Brechbühl et al. 2013). Unlike another sulfur-containing pheromone of mice urine, SBT is not contained in food, being specific for synanthropic house mice (Kwak et al. 2016). It can be SBT, probably in combination with other sulfur-containing components of house mice urine, which acts as a repellent for other rodent species. SBT, in complex with another sulfur-contained component of house mouse urine (3,4-dehydro-exo-brevicomin), found in higher concentrations in males than in females, was shown to increase 4-times the frequency of female trapping (Musso et al. 2017).

As we noted earlier, the SBT effect can be compared with the predator odor effect on rodent behavior and reproduction (Bazhenov et al. 2014). In the laboratory, predator odor provokes an innate fear response among some mammal species, which enables to deploy synthetic analogues of carnivorous excreta as repellents (Lindgren et al. 1995; Takahashi, 2014). Recent studies found the similarity between SBT molecular structure and that of predator excreta components, induced fear response among rodents, particularly, heterocyclic sulfur- and nitrogen-containing compounds (e.g., predator pheromone 2,3,5-trimethyl-3-thiazoline (TMT); Brechbühl et al. 2013).

The same chemical is known to act as a pheromone in intraspecific communication or to be a heterotelergon, that is, a substance affecting other species (Kirshenblat, 1968), or to be a kairomone, that is, a substance released by the body into the environment and having a specific effect on another species. As an example, the sulfur-containing aminoacid l-phelinin and its derivatives, found in house cat urine, not only inhibit the reproduction of rats and mice (Voznessenskaya, 2014) but also convey information about predator sex and social status (Miyazaki et al. 2006; 2008). Field data demonstrated that the male sex pheromone of Rattus norvegicus attracts conspecific females and repels males. A synthetic blend of 6 compounds, including male sex pheromone (2-heptanone, 4-heptanone, 3-ethyl-2-heptanone, 2-octanone, 2-nonanone, and 4-nonanone), increased trappability of wild female rats 10 times more than the corresponding control boxes (Takács et al. 2016). The “counterespionage” hypothesis that predator (rats) and prey (mice) co-opt each other’s pheromone as a cue to locate prey or evade predation, was supported. Pair of male rat pheromone-baited traps captured 3.05 times fewer mice than the trap pairs with scented male mouse pheromone components, and no female mice were captured in rat pheromone-baited traps. These data indicate that male rat pheromone is aversive for mice (Varner et al. 2020).

Results, given here, provide the basis for further research on effects, including repellent, of SBT and other house mouse urine components on the behavior of exoanthropic and semi-synanthropic small mammal species, which have different potential to settle in human-made environments.

Ethics approval

All experiments with mice were performed in accordance with the rules adopted by the European Convention for the Protection of Vertebrate Animals used for Experimental and other Scientific Purposes. The experimental protocol was approved by the Bioethical Committee of the A.N. Severtsov Institute of Ecology and Evolution of the Russian Academy of Sciences.