- Email: [email protected]
Ultrasonic vocalization by female mice in the presence of a conspecific carrying food cues
ANIMAL BEHAVIOUR, 2000, 60, 689–694 doi:10.1006/anbe.2000.1504, available online at http://www.idealibrary.com on
Ultrasonic vocalization by female mice in the presence of a conspecific carrying food cues ANNA MOLES & FRANCESCA R. D’AMATO
Istituto di Psicobiologia e Psicofarmacologia del CNR (Received 27 September 1999; initial acceptance 23 November 1999; final acceptance 12 June 2000; MS. number: 6362R)
In female mice, Mus domesticus, reunion with a same-sex conspecific is associated with intense ultrasonic vocalization. We examined whether the palatability of a familiar food eaten by a demonstrator mouse and the motivational state of the conspecific observer could modulate the number of ultrasonic calls uttered during female–female interaction in NMRI mice. A pilot study indicated that these calls were uttered almost exclusively by the observer member of the pair. Observers were either food deprived or not deprived and demonstrators were offered either no food, a palatable diet or an unpalatable diet. We monitored both the number of ultrasounds (range 65–75 kHz) uttered and a series of behavioural parameters during the first 3 min of social interaction after 24 h of separation. Observers investigated the nose area of demonstrators fed on the unpalatable diet more than the same area of demonstrators not given food. No differences were found in demonstrators’ behaviour. Ultrasonic calls were given immediately after female–female reunion and were affected by both the motivational state of the observer and the salience of the information carried by the conspecific. These results suggest that the motivational state of the observer affects ultrasonic calling towards a demonstrator conspecific. Nondeprived animals produced more calls towards demonstrators fed on palatable food, whereas food-deprived subjects vocalized more to fed conspecifics, independently of the palatability of the food eaten by the demonstrator. We suggest that ultrasonic vocalization in female mice can facilitate proximity with a conspecific and the number of these calls is modulated by the salience of the information carried by the companion.
during intersexual encounters. The function of UVs in females is still poorly understood. They may serve as signals of sex recognition (Wysocki et al. 1982). Furthermore, since they are uttered principally during female– female social investigation, Maggio & Whitney (1985) suggested that they can contribute to the establishment of dominance hierarchies within demes. The lack of studies exploring the contexts in which UVs occur and analysing the variables influencing them in females makes such explanations highly speculative. In female mice (Maggio & Whitney 1985), as in males (Sales 1972), the majority of vocalizations occur during the first few minutes of interaction, when high levels of social investigation occur. Olfactory investigation allows animals to gain information on the identity (i.e. sex, rank, reproductive state) of a conspecific (Brown 1985). Furthermore, the results of several studies are consistent with the view that during social interaction adult observer mice and rats, Rattus norvegicus, use olfactory cues from demonstrator animals to acquire information about the food they are eating (Galef & Wigmore 1983; Valsecchi & Galef 1989; Galef 1994; Valsecchi et al. 1994
Ultrasonic vocalizations (UVs) by adult female mice, Mus domesticus, have been little investigated. For a long time they were thought to be a typical infant behaviour (Noirot 1966), but have also been reported in adulthood, during male–female interactions (Sales 1972). In the latter context, UVs are uttered mainly by the male (Whitney et al. 1973; White et al. 1998), and their emission is affected by testosterone administration (Nyby et al. 1977; Nunez et al. 1978; Nunez & Tan 1984). The observation that female mice spend more time close to vocalizing than devocalized males suggested that these calls are a component of male courtship behaviour: they promote female proximity to facilitate copulation (Pomerantz et al. 1983), although they are not necessary for copulation to occur (White et al. 1998). However, sporadic female UVs have been reported (D’Udine et al. 1982); moreover, Maggio & Whitney (1985) showed that females utter UVs during adulthood, mostly during interactions between females, rather than Correspondence: F. R. D’Amato, Istituto di Psicobiologia e Psicofarmacologia del CNR, Viale Marx, 43, 00137 Roma, Italy (email: [email protected]). 0003–3472/00/110689+06 $35.00/0
2000 The Association for the Study of Animal Behaviour
689
2000 The Association for the Study of Animal Behaviour
690
ANIMAL BEHAVIOUR, 60, 5
for reviews). This behaviour has a strong adaptive value in central place foraging and generalist species such as rats and mice. Galef & Whiskin (1998) reported that the impact of the demonstrator on the observer’s food choice is a function of the relative palatability of the food eaten by the demonstrator. Therefore, it is reasonable to assume that the salience of the information to be acquired from a conspecific changes with the palatability of the diet. We hypothesized that the quality of food eaten by a conspecific and the motivational state of the observer could modulate the number of UVs uttered during female–female interaction. We investigated whether the number of UVs uttered by the observer, when interacting with a recently fed demonstrator, could be linked to the palatability of the food. Moreover, we explored whether manipulating the motivational state of the observer with respect to food stimuli would change UV emission. For this reason, one group of observer mice was food deprived while the other was not. We expected food-deprived animals to be less choosy in terms of food preferences and that this would be mirrored in their UVs during interaction with the demonstrator. GENERAL METHODS
Subjects We used NMRI (Swiss-type) nulliparous females (60–90 days old). After arrival they were housed for 2 weeks in groups of four in Plexiglas cages (3313 cm and 13 cm high) with food (standard pellets, Harlan Teklad TRM, Mucedola S.r.l., Settimo Milanese, Italy) and water always available. Room temperature (211C) and a 12:12 h light:dark cycle (lights on at 0700 hours) were kept constant.
Diets The palatable diet consisted of 70 g of powdered standard diet mixed with 30 g of Nutella (sugar, vegetable oils, nuts, cocoa and milk, Ferrero S.p.A., Alba, Italy) and 100 g of distilled water. The unpalatable diet was made by mixing 94 g of powdered standard diet with 6 g of ground fennel seeds and 100 g of distilled water. Fennel seeds taste unpleasant to mice (Mainardi et al. 1989), and in a binary choice food preference test the palatable diet was highly preferred to the unpalatable one (mean percentage of palatable dietSE eaten by 10 subjects in a 30-min binary food preference test: 92.86.2).
Experimental Procedure Evaluation of baseline On day 0 two females from each cage group of four were marked with fur dye and housed individually in clean cages to serve as observers, while the remaining two animals were left in their cage to serve as demonstrators. Observers and demonstrators were thus familiar to each other. On day 1 one demonstrator was introduced into the home cage of the first observer and left there, while
the other was introduced into the home cage of the second observer and left there. We counted the number of UVs uttered during the first 3 min of interaction. Pairs uttering fewer than 10 UVs were excluded. Pairs uttering more than 300 UVs in 3 min were also excluded to prevent loss of accuracy in counting UVs.
Familiarization with test conditions After baseline measurements, each observer/ demonstrator pair was left undisturbed in the observer’s cage for 4 days. To ensure that observers could evaluate the palatability of the food eaten by the demonstrator in a short period of social interaction we familiarized both mice to the new foods. On days 5 and 6 we deprived each observer/demonstrator pair of food from 0800 to 1500 hours (7 h). At 1500 hours on each day each pair was moved to a new cage (feeding cage, 3313 cm and 13 cm high, scattered with home cage bedding) for 1 h. Half of the pairs were offered the palatable diet on both days 5 and 6 and the other half the unpalatable one. The food was given ad libitum during this 1-h familiarization period. After this period both animals were returned to the observer’s cage with water and standard food ad libitum until the next day at 0800 hours. On day 7 from 0800 to 1500 hours, the pair underwent the usual period of food deprivation. Then at 1500 hours, we separated the members of each pair for 24 h. The demonstrator was placed in the feeding cage where, according to the experimental group, it was offered either the standard, the palatable or the unpalatable diet. Meanwhile the observer was left alone in its home cage and fed standard diet. On day 8, observers were either fed standard diet or food deprived from 0800 hours until the test at 1500 hours. From 0800 to 1500 hours, demonstrators were food deprived; thereafter two-thirds of them were fed for 30 min according to their experimental group (one third palatable diet, one-third unpalatable diet) and placed back in the observer’s home cage for a 3-min social interaction test. The remaining one-third were left food deprived for an additional 30 min before their 3-min social interaction test. Water was available ad libitum throughout the familiarization procedure.
Pilot Study We did a preliminary experiment to identify which member of the dyad (if only one) was mainly responsible for UVs. From day 0 to day 4 the procedure was the same as that described in the general procedure. From day 5 to day 8 the experimental procedure was similar to the general procedure, but (1) the mice were never food deprived and (2) in the feeding cage they were offered standard pellets instead of the palatable or unpalatable diets. On the test day (day 8), after 24 h of separation, either an anaesthetized demonstrator (N=8) was introduced into the home cage of its awake observer partner or an awake demonstrator was introduced into the home cage of its anaesthetized partner (N=8). This procedure was used by Whitney et al. (1973), and their results were
MOLES & D’AMATO: ULTRASOUNDS AND FOOD CUES IN MICE
confirmed by Nunez et al. (1985) in devocalized animals. The anaesthetic used was chloral hydrate (500 mg/kg, administered by intraperitoneal injection). We detected UVs during the first 3 min almost exclusively when the observer was awake (8/8 vocalizing pairs versus 1/8 vocalizing pairs when the observer was anaesthetized). There were fewer UVs at the baseline level, suggesting that an active animal elicits more calls than an anaesthetized one. These data together with observations that the observers tended to investigate the demonstrators, whereas the latter tended to explore the environment, suggest that UVs were almost exclusively uttered by observers.
Experimental Groups Six groups of observer/demonstrator pairs were formed on the basis of the number of UVs recorded at baseline. Groups not differing for this behavioural parameter were obtained (ANOVA: F5,59 =0.10, NS). There were 10–12 pairs per group. These groups differed in the observer’s food deprivation condition on day 8 (deprived and not deprived) and the demonstrator’s food on day 7 and 8: no food (standard diet on day 7 and no food on day 8); palatable diet (fed with palatable diet on days 7 and 8); unpalatable diet (fed with unpalatable diet on days 7 and 8).
Behavioural Observations We videotaped the behaviour of the mice in the observer’s cage (Plexiglas apparatus, 3313 cm and 13 cm high) for the first 3 min of social interaction. From the tapes, an experimenter (F.R.D.A.) blind to the experimental situation recorded the behaviour by the focal sample method (Altmann 1974), using a keyboard connected to a Macintosh LC III computer with software written in our laboratory to record the duration of several behaviours. We recorded time (s) spent by the observer in the following behaviours (Grant & Mackintosh 1963): sniff nose (orientation to the demonstrator’s mouth or nose); sniff body (orientation to its body and anogenital areas); exploration (orientation to the physical environment either standing or in motion); and self-grooming (lick and scratch own body). The demonstrator’s behaviours included time spent in social sniffing (nose and body), exploration and self-grooming. We counted the number of UVs uttered in each minute of the 3-min test with a hand tally counter; for this we used the audible output of a bat detector (QMC Instruments, London, U.K.) set to a frequency of 70 (5) kHz and suspended 10–12 cm above the cage.
Statistical Analysis To allow parametric assumptions to be met, a squareroot transformation was computed on both behavioural and UV data. Behavioural profiles were analysed by twoway MANOVAs with observer food deprivation condition (two levels: deprived and not deprived) and demonstrator
food (three levels: no food, palatable, unpalatable) as factors. The observer’s UVs were analysed by a repeated measures ANOVA with observer food deprivation condition (two levels) and demonstrator food (three levels) as between-subjects factors and time (three levels: minutes 1, 2, 3) as a within-subjects factor. Post hoc comparisons were carried out by Tukey–Kramer tests.
Ethical Note All mice anaesthetized for the pilot study recovered from the drug administration after 20–30 min. No adverse effects of the procedure were noted. New food is not easily accepted, especially when unpalatable and offered in an unfamiliar cage. To familiarize the animals with the new diets and test conditions we used several procedures: food deprivation; 3 days of training; feeding cage scattered with home cage shavings; and the presence of the familiar partner. We did a preliminary experiment to verify the effect of 7 h of food deprivation on subsequent food intake. This level of food deprivation resulted in a mean intakeSE of 1.3450.117 and 1.3330.141 g of palatable and unpalatable food, respectively (both diets were offered wet to facilitate ingestion in a short time). In the light of these results, mice deprived of food for 7 h were expected to eat the same amount of both types of food, enough to carry information on that food. To take account of the daily feeding pattern of mice, the period of food deprivation started at 0800 hours, during the light phase and after the peak of feeding activity. The animals fed on the unpalatable diet did not show any signs of distress, did not lose any weight and, owing to the mild food deprivation, consumed the same amount of food as mice fed on the palatable diet. All mice were checked daily by the animal care technicians and ourselves. The weight range of the mice before and after the experiment was 26–31 g. The lack of differences in behaviour of demonstrators in the three food conditions in the 3-min test (see Results) also suggests the treatments did not cause any distress. There were no obvious adverse effects on the health of the mice. This study was carried out under a Ministero della Sanita` Italiano licence for animal care. RESULTS During the 3 min of interaction after reunion, no significant differences were found in the behaviour of demonstrators either in relation to observer condition or demonstrator food group (effects: observer: Wilk’s 3,57 =0.97, NS; demonstrator: Wilk’s 6,114 =0.87, NS; interaction: Wilk’s 6,114 =0.93, NS). In contrast, the observer’s behaviour varied with the demonstrator’s food group (Wilk’s 8,112 =0.74, P=0.02). Observers spent more time sniffing the nose/mouth area of demonstrators fed on the unpalatable diet than they did with unfed demonstrators (Tukey–Kramer: P<0.05). There were no effects of the observer’s food deprivation condition (Wilk’s 4,56 =0.94, NS) or of the observer*demonstrator interaction (Wilk’s 8,112 =0.95 , NS; Table 1).
691
ANIMAL BEHAVIOUR, 60, 5
Table 1. Time (s) spent in different behaviours by female mice (observers) during 3 min of social interaction with their cagemate (demonstrators) after 24 h of separation Demonstrator Observer
No food
Palatable diet
Unpalatable diet
Not food deprived Sniff nose Sniff body Self-groom Explore N
10.58±2.71 65.92±10.42 2.17±0.85 101.75±12.34 12
12.92±2.00 71.92±9.24 1.17±0.49 93.67±8.77 12
20.70±3.73 67.60±8.56 4.80±4.15 87.60±9.31 10
Food deprived Sniff nose Sniff body Self-groom Explore N
6.00±1.38 55.30±10.41 2.80±1.68 116.90±11.19 10
8.64±1.56 82.09±11.06 2.91±1.21 87.27±11.01 11
17.70±5.17 62.10±9.97 2.00±1.34 98.60±13.30 10
Means are given±SE. Observers were either deprived of food for 7 h or not deprived. Demonstrators were either not fed or were given palatable or unpalatable food for 30 min before the 3-min test.
100
100 (a)
(b) *†
80 Number of UVs
692
80
60
60 †
*‡
40
40
20
20
0
1
2
3
0
*
1
2
3
Time during test (min) Figure 1. Mean±SEM number of ultrasounds (UVs) emitted by (a) not food-deprived or (b) food-deprived observer female mice after reunion with a demonstrator that was either offered no food ( ) or fed on a palatable ( ) or an unpalatable ( ) diet. Statistical analysis was performed on the square-root transformed data. P levels (Tukey–Kramer) refer to comparisons within the same minute of observation. *Palatable diet versus no food; †palatable diet versus unpalatable diet; ‡unpalatable diet versus no food. All Ps<0.05.
In the light of our pilot study we assumed that the observer uttered the majority of UVs within a pair of observers/demonstrators. Observers uttered fewer UVs when food deprived than when not deprived (F1,59 =7.94, P<0.01; Fig. 1). The number of UVs was strongly affected by the palatability of the food eaten by the demonstrator (F2,59 =6.55, P<0.01) and this effect was dependent on the observer’s motivational state (interaction: F2,59 =3.60, P<0.05). The number of UVs decreased across the 3 min of observation (F2,118 =54.3, P<0.001). In the first minute no differences were found in UVs between the demonstrator groups, both for nondeprived (Fig. 1a) and deprived observers (Fig. 1b). During the second minute nondeprived observers uttered more UVs with demonstrators fed palatable food than with unfed demonstrators (Tukey–Kramer: P<0.05). They also uttered more UVs in the presence of demonstrators fed palatable food (in both
the second and the third minutes) than with demonstrators fed unpalatable food (Tukey–Kramer: P<0.05). Deprived observers uttered more UVs in the presence of fed demonstrators (palatable diet: second and third minutes; unpalatable diet: second minute) than with unfed demonstrators (Tukey–Kramer: P<0.05). DISCUSSION Ultrasonic communication in adult female mice has long been underinvestigated since it was considered a typical male behaviour. Our study shows that it is also an important component of the females’ behavioural repertoire. The UVs in our study were almost entirely uttered by the observer member of the pair, that is, the one that remained in its home cage. The number of UVs during the 3 min after reunion is comparable to the
MOLES & D’AMATO: ULTRASOUNDS AND FOOD CUES IN MICE
number of UVs of males of the same strain during courtship behaviour (D’Amato 1991). These calls were uttered immediately after reunion and were affected both by the motivational state of the vocalizing individual and the salience of the information carried by the conspecific. The first few minutes of interaction between female mice were characterized by intense social investigation and high levels of ultrasonic calling by the observer, whereas the demonstrator was mainly interested in the physical environment. Motivational state (food deprivation condition) did not affect the observer’s behaviour towards the demonstrator, although it strongly modulated the number of calls. A difference in the number of UVs emerged in the observers: food-deprived animals called less. Indeed, ultrasonic calling is a highly uneconomical behaviour (Nyby & Whitney 1978) and would be particularly costly for food-restricted animals. In contrast, the food eaten by the demonstrators affected both the behaviour and the number of calls of the observers. Observers differed in which region of the partner they investigated: the nose area of demonstrators fed the unpalatable food was investigated more than the same area of other demonstrator groups. Body sniffing, in contrast, did not vary with the demonstrator’s food. This suggests that some trace of this unpalatable food probably remained in the nose/mouth area, eliciting additional sniffing. However, the tendency of food-deprived observers to investigate this region less bears out the hypothesis that this behaviour is not related to the feeding motivational state of the animals. During the first minute of interaction, UVs did not differ significantly between groups, although nondeprived observers tended to vocalize less in response to a demonstrator fed on the unpalatable food. During the second and third minutes, statistically significant differences were detected in observers’ UVs according to both demonstrator group and food deprivation condition. In food-deprived observers more UVs were elicited in the second minute in the presence of fed animals irrespective of the quality of the food eaten by the partner. In the presence of a conspecific fed on the palatable food, vocalization was high until the end of the test. In contrast, nondeprived observers called more when the demonstrators were fed palatable food and this effect persisted until the last minute of the test. We conclude that the motivational state affects UVs towards a demonstrator conspecific, with nondeprived animals calling more towards demonstrators fed on palatable food, whereas food-deprived subjects vocalized more to fed conspecifics, independently of the palatability of the food eaten. From a functional viewpoint, UVs emitted in male– female pairs may facilitate mating by promoting proximity (Pomerantz et al. 1983). Bean et al. (1986) proposed the primary function of 70-kHz vocalizations to be ‘to announce the presence of another mouse and to attract other mice’. We suggest that these calls have an affiliative function, rather than a role in rank acquisition (Maggio & Whitney 1985). Both the infantile origin of this behaviour and the absence of UVs during male–male interactions (Sales 1972; Whitney et al. 1973) support this
hypothesis. In female mice UVs can facilitate proximity, helping relevant information to be acquired. Our study provided a context in which proximity allowed observers to identify the demonstrator’s food source. When the observers first approached the companion to gain general information (i.e. individual/sex recognition), UVs were unaffected by the quality of the demonstrator’s food. In contrast, after these data were acquired observers vocalized more to conspecifics carrying more attractive cues. In this study we used the palatability of food as a cue to make demonstrators more or less attractive to their observers. The possible role of UVs in the social acquisition of food preferences should be investigated. Further studies are needed to explore this behaviour in observers not familiar with the foods eaten by their conspecific demonstrators. As Galef (1993) pointed out, social interaction serves primarily as a means of expanding feeding repertoires rather than maintaining food habits. Acknowledgments A.M. was supported by a CNR scholarship Bando di Concorso 201.16.14. We thank two anonymous referees for their comments on the manuscript. References Altmann, J. 1974. Observational study of behavior: sampling methods. Behaviour, 49, 227–266. Bean, N. J., Nunez, A. & Wysocki, C. J. 1986. 70-kHz vocalizations by male mice do not inhibit aggression in lactating mice. Behavioral and Neural Biology, 46, 46–53. Brown, R. E. 1985. The rodents II: suborder Myomorpha. In: Social Odours in Mammals. (Ed. by R. E. Brown & D. W. MacDonald), pp. 389–457. Oxford: Clarendon Press. D’Amato, F. R. 1991. Courtship ultrasonic vocalizations and social state in mice. Animal Behaviour, 41, 875–885. D’Udine, B., Robinson, D. J. & Oliverio, A. 1982. An analysis of single gene effects on audible and ultrasonic vocalizations in the mouse. Behavioral and Neural Biology, 36, 197–203. Galef, B. G. Jr. 1993. Functions of social learning about food: a causal analysis of effects of diet novelty on preference transmission. Animal Behaviour, 46, 257–265. Galef, B. G. Jr. 1994. Olfactory communication about foods among rats: a review of recent findings. In: Behavioral Aspects of Feeding (Ed. by B. G. Galef, Jr, M. Mainardi & P. Valsecchi), pp. 83–101. Chur: Harwood Academic. Galef, B. G. Jr & Whiskin, E. E. 1998. Limits on social influence on food choice of Norway rats. Animal Behaviour, 56, 1015–1020. Galef, B. G. Jr & Wigmore, S. W. 1983. Transfer of information concerning distant foods: a laboratory investigation of the ‘information-centre’ hypothesis. Animal Behaviour, 31, 748–758. Grant, E. C. & Mackintosh, J. H. 1963. A comparison of the social postures of some common laboratory rodents. Behaviour, 21, 246–259. Maggio, J. & Whitney, G. 1985. Ultrasonic vocalizing by adult female mice, Mus musculus. Journal of Comparative Psychology, 99, 420–436. Mainardi, M., Poli, M. & Valsecchi, P. 1989. Ontogeny of dietary selection in weanling mice: effects of early experience and mother’s milk. Biology of Behavior, 14, 185–194.
693
694
ANIMAL BEHAVIOUR, 60, 5
Noirot, E. 1966. Ultrasounds in young rodents. II Changes with age in albino mice. Animal Behaviour, 14, 129–134. Nunez, A. A. & Tan, D. T. 1984. Courtship ultrasonic vocalizations in male SwissWebster mice: effects of hormones and sexual experience. Physiology and Behavior, 32, 717–721. Nunez, A. A., Nyby, J. & Whitney, G. 1978. The effects of testosterone, estradiol and dihydrotestosterone on male mice (Mus musculus) ultrasonic vocalizations. Hormones and Behavior, 11, 264–272. Nunez, A. A., Pomerantz, S. M., Bean, N. J. & Youngstrom, T. G. 1985. Effects of laryngeal denervation on ultrasound production and male sexual behavior in rodents. Physiology and Behavior, 34, 901–905. Nyby, J. & Whitney, G. 1978. Ultrasonic communication of adult miomorph rodents. Neuroscience Biobehavioural Reviews, 2, 1–14. Nyby, J., Dizinno, J. & Whitney, G. 1977. Sexual dimorphism in ultrasonic vocalization in mice: gonadal hormone regulation. Journal of Comparative and Physiological Psychology, 93, 957–975. Pomerantz, S. M., Nunez, A. A. & Bean, N. J. 1983. Female behavior is affected by male ultrasonic vocalizations in house mice. Physiology and Behavior, 31, 91–96.
Sales, G. D. 1972. Ultrasound and mating behaviour in rodents with some observations on other behavioural situations. Journal of Zoology, 168, 149–164. Valsecchi, P. & Galef, B. G. Jr. 1989. Social influences of the food preferences of house mice (Mus musculus). International Journal of Comparative Psychology, 2, 245–256. Valsecchi, P., Moles, A., Mainardi, D. & Mainardi, M. 1994. Individual and social experiences in the establishment of food preferences in mice. In: Behavioral Aspects of Feeding (Ed. by B. G. Galef, Jr, M. Mainardi & P. Valsecchi), pp. 103–124. Chur: Harwood Academic. White, N. R., Prasad, M., Barfield, R. J. & Nyby, J. 1998. 40 and 70-kHz vocalizations of mice, Mus musculus, during copulation. Physiology and Behavior, 63, 467–473. Whitney, G., Coble, J., Stockton, M. & Tilson, E. 1973. Ultrasonic emissions: do they facilitate courtship of mice? Journal of Comparative and Physiological Psychology, 84, 445–452. Wysocki, C. J., Nyby, J., Whitney, G., Beauchamp, G. K. & Katz, Y. 1982. The vomeronasal organ: primary role in mouse chemosensory gender recognition. Physiology and Behavior, 29, 315–327.
Ultrasonic vocalization by female mice in the presence of a conspecific carrying food cues ANNA MOLES & FRANCESCA R. D’AMATO
Istituto di Psicobiologia e Psicofarmacologia del CNR (Received 27 September 1999; initial acceptance 23 November 1999; final acceptance 12 June 2000; MS. number: 6362R)
In female mice, Mus domesticus, reunion with a same-sex conspecific is associated with intense ultrasonic vocalization. We examined whether the palatability of a familiar food eaten by a demonstrator mouse and the motivational state of the conspecific observer could modulate the number of ultrasonic calls uttered during female–female interaction in NMRI mice. A pilot study indicated that these calls were uttered almost exclusively by the observer member of the pair. Observers were either food deprived or not deprived and demonstrators were offered either no food, a palatable diet or an unpalatable diet. We monitored both the number of ultrasounds (range 65–75 kHz) uttered and a series of behavioural parameters during the first 3 min of social interaction after 24 h of separation. Observers investigated the nose area of demonstrators fed on the unpalatable diet more than the same area of demonstrators not given food. No differences were found in demonstrators’ behaviour. Ultrasonic calls were given immediately after female–female reunion and were affected by both the motivational state of the observer and the salience of the information carried by the conspecific. These results suggest that the motivational state of the observer affects ultrasonic calling towards a demonstrator conspecific. Nondeprived animals produced more calls towards demonstrators fed on palatable food, whereas food-deprived subjects vocalized more to fed conspecifics, independently of the palatability of the food eaten by the demonstrator. We suggest that ultrasonic vocalization in female mice can facilitate proximity with a conspecific and the number of these calls is modulated by the salience of the information carried by the companion.
during intersexual encounters. The function of UVs in females is still poorly understood. They may serve as signals of sex recognition (Wysocki et al. 1982). Furthermore, since they are uttered principally during female– female social investigation, Maggio & Whitney (1985) suggested that they can contribute to the establishment of dominance hierarchies within demes. The lack of studies exploring the contexts in which UVs occur and analysing the variables influencing them in females makes such explanations highly speculative. In female mice (Maggio & Whitney 1985), as in males (Sales 1972), the majority of vocalizations occur during the first few minutes of interaction, when high levels of social investigation occur. Olfactory investigation allows animals to gain information on the identity (i.e. sex, rank, reproductive state) of a conspecific (Brown 1985). Furthermore, the results of several studies are consistent with the view that during social interaction adult observer mice and rats, Rattus norvegicus, use olfactory cues from demonstrator animals to acquire information about the food they are eating (Galef & Wigmore 1983; Valsecchi & Galef 1989; Galef 1994; Valsecchi et al. 1994
Ultrasonic vocalizations (UVs) by adult female mice, Mus domesticus, have been little investigated. For a long time they were thought to be a typical infant behaviour (Noirot 1966), but have also been reported in adulthood, during male–female interactions (Sales 1972). In the latter context, UVs are uttered mainly by the male (Whitney et al. 1973; White et al. 1998), and their emission is affected by testosterone administration (Nyby et al. 1977; Nunez et al. 1978; Nunez & Tan 1984). The observation that female mice spend more time close to vocalizing than devocalized males suggested that these calls are a component of male courtship behaviour: they promote female proximity to facilitate copulation (Pomerantz et al. 1983), although they are not necessary for copulation to occur (White et al. 1998). However, sporadic female UVs have been reported (D’Udine et al. 1982); moreover, Maggio & Whitney (1985) showed that females utter UVs during adulthood, mostly during interactions between females, rather than Correspondence: F. R. D’Amato, Istituto di Psicobiologia e Psicofarmacologia del CNR, Viale Marx, 43, 00137 Roma, Italy (email: [email protected]). 0003–3472/00/110689+06 $35.00/0
2000 The Association for the Study of Animal Behaviour
689
2000 The Association for the Study of Animal Behaviour
690
ANIMAL BEHAVIOUR, 60, 5
for reviews). This behaviour has a strong adaptive value in central place foraging and generalist species such as rats and mice. Galef & Whiskin (1998) reported that the impact of the demonstrator on the observer’s food choice is a function of the relative palatability of the food eaten by the demonstrator. Therefore, it is reasonable to assume that the salience of the information to be acquired from a conspecific changes with the palatability of the diet. We hypothesized that the quality of food eaten by a conspecific and the motivational state of the observer could modulate the number of UVs uttered during female–female interaction. We investigated whether the number of UVs uttered by the observer, when interacting with a recently fed demonstrator, could be linked to the palatability of the food. Moreover, we explored whether manipulating the motivational state of the observer with respect to food stimuli would change UV emission. For this reason, one group of observer mice was food deprived while the other was not. We expected food-deprived animals to be less choosy in terms of food preferences and that this would be mirrored in their UVs during interaction with the demonstrator. GENERAL METHODS
Subjects We used NMRI (Swiss-type) nulliparous females (60–90 days old). After arrival they were housed for 2 weeks in groups of four in Plexiglas cages (3313 cm and 13 cm high) with food (standard pellets, Harlan Teklad TRM, Mucedola S.r.l., Settimo Milanese, Italy) and water always available. Room temperature (211C) and a 12:12 h light:dark cycle (lights on at 0700 hours) were kept constant.
Diets The palatable diet consisted of 70 g of powdered standard diet mixed with 30 g of Nutella (sugar, vegetable oils, nuts, cocoa and milk, Ferrero S.p.A., Alba, Italy) and 100 g of distilled water. The unpalatable diet was made by mixing 94 g of powdered standard diet with 6 g of ground fennel seeds and 100 g of distilled water. Fennel seeds taste unpleasant to mice (Mainardi et al. 1989), and in a binary choice food preference test the palatable diet was highly preferred to the unpalatable one (mean percentage of palatable dietSE eaten by 10 subjects in a 30-min binary food preference test: 92.86.2).
Experimental Procedure Evaluation of baseline On day 0 two females from each cage group of four were marked with fur dye and housed individually in clean cages to serve as observers, while the remaining two animals were left in their cage to serve as demonstrators. Observers and demonstrators were thus familiar to each other. On day 1 one demonstrator was introduced into the home cage of the first observer and left there, while
the other was introduced into the home cage of the second observer and left there. We counted the number of UVs uttered during the first 3 min of interaction. Pairs uttering fewer than 10 UVs were excluded. Pairs uttering more than 300 UVs in 3 min were also excluded to prevent loss of accuracy in counting UVs.
Familiarization with test conditions After baseline measurements, each observer/ demonstrator pair was left undisturbed in the observer’s cage for 4 days. To ensure that observers could evaluate the palatability of the food eaten by the demonstrator in a short period of social interaction we familiarized both mice to the new foods. On days 5 and 6 we deprived each observer/demonstrator pair of food from 0800 to 1500 hours (7 h). At 1500 hours on each day each pair was moved to a new cage (feeding cage, 3313 cm and 13 cm high, scattered with home cage bedding) for 1 h. Half of the pairs were offered the palatable diet on both days 5 and 6 and the other half the unpalatable one. The food was given ad libitum during this 1-h familiarization period. After this period both animals were returned to the observer’s cage with water and standard food ad libitum until the next day at 0800 hours. On day 7 from 0800 to 1500 hours, the pair underwent the usual period of food deprivation. Then at 1500 hours, we separated the members of each pair for 24 h. The demonstrator was placed in the feeding cage where, according to the experimental group, it was offered either the standard, the palatable or the unpalatable diet. Meanwhile the observer was left alone in its home cage and fed standard diet. On day 8, observers were either fed standard diet or food deprived from 0800 hours until the test at 1500 hours. From 0800 to 1500 hours, demonstrators were food deprived; thereafter two-thirds of them were fed for 30 min according to their experimental group (one third palatable diet, one-third unpalatable diet) and placed back in the observer’s home cage for a 3-min social interaction test. The remaining one-third were left food deprived for an additional 30 min before their 3-min social interaction test. Water was available ad libitum throughout the familiarization procedure.
Pilot Study We did a preliminary experiment to identify which member of the dyad (if only one) was mainly responsible for UVs. From day 0 to day 4 the procedure was the same as that described in the general procedure. From day 5 to day 8 the experimental procedure was similar to the general procedure, but (1) the mice were never food deprived and (2) in the feeding cage they were offered standard pellets instead of the palatable or unpalatable diets. On the test day (day 8), after 24 h of separation, either an anaesthetized demonstrator (N=8) was introduced into the home cage of its awake observer partner or an awake demonstrator was introduced into the home cage of its anaesthetized partner (N=8). This procedure was used by Whitney et al. (1973), and their results were
MOLES & D’AMATO: ULTRASOUNDS AND FOOD CUES IN MICE
confirmed by Nunez et al. (1985) in devocalized animals. The anaesthetic used was chloral hydrate (500 mg/kg, administered by intraperitoneal injection). We detected UVs during the first 3 min almost exclusively when the observer was awake (8/8 vocalizing pairs versus 1/8 vocalizing pairs when the observer was anaesthetized). There were fewer UVs at the baseline level, suggesting that an active animal elicits more calls than an anaesthetized one. These data together with observations that the observers tended to investigate the demonstrators, whereas the latter tended to explore the environment, suggest that UVs were almost exclusively uttered by observers.
Experimental Groups Six groups of observer/demonstrator pairs were formed on the basis of the number of UVs recorded at baseline. Groups not differing for this behavioural parameter were obtained (ANOVA: F5,59 =0.10, NS). There were 10–12 pairs per group. These groups differed in the observer’s food deprivation condition on day 8 (deprived and not deprived) and the demonstrator’s food on day 7 and 8: no food (standard diet on day 7 and no food on day 8); palatable diet (fed with palatable diet on days 7 and 8); unpalatable diet (fed with unpalatable diet on days 7 and 8).
Behavioural Observations We videotaped the behaviour of the mice in the observer’s cage (Plexiglas apparatus, 3313 cm and 13 cm high) for the first 3 min of social interaction. From the tapes, an experimenter (F.R.D.A.) blind to the experimental situation recorded the behaviour by the focal sample method (Altmann 1974), using a keyboard connected to a Macintosh LC III computer with software written in our laboratory to record the duration of several behaviours. We recorded time (s) spent by the observer in the following behaviours (Grant & Mackintosh 1963): sniff nose (orientation to the demonstrator’s mouth or nose); sniff body (orientation to its body and anogenital areas); exploration (orientation to the physical environment either standing or in motion); and self-grooming (lick and scratch own body). The demonstrator’s behaviours included time spent in social sniffing (nose and body), exploration and self-grooming. We counted the number of UVs uttered in each minute of the 3-min test with a hand tally counter; for this we used the audible output of a bat detector (QMC Instruments, London, U.K.) set to a frequency of 70 (5) kHz and suspended 10–12 cm above the cage.
Statistical Analysis To allow parametric assumptions to be met, a squareroot transformation was computed on both behavioural and UV data. Behavioural profiles were analysed by twoway MANOVAs with observer food deprivation condition (two levels: deprived and not deprived) and demonstrator
food (three levels: no food, palatable, unpalatable) as factors. The observer’s UVs were analysed by a repeated measures ANOVA with observer food deprivation condition (two levels) and demonstrator food (three levels) as between-subjects factors and time (three levels: minutes 1, 2, 3) as a within-subjects factor. Post hoc comparisons were carried out by Tukey–Kramer tests.
Ethical Note All mice anaesthetized for the pilot study recovered from the drug administration after 20–30 min. No adverse effects of the procedure were noted. New food is not easily accepted, especially when unpalatable and offered in an unfamiliar cage. To familiarize the animals with the new diets and test conditions we used several procedures: food deprivation; 3 days of training; feeding cage scattered with home cage shavings; and the presence of the familiar partner. We did a preliminary experiment to verify the effect of 7 h of food deprivation on subsequent food intake. This level of food deprivation resulted in a mean intakeSE of 1.3450.117 and 1.3330.141 g of palatable and unpalatable food, respectively (both diets were offered wet to facilitate ingestion in a short time). In the light of these results, mice deprived of food for 7 h were expected to eat the same amount of both types of food, enough to carry information on that food. To take account of the daily feeding pattern of mice, the period of food deprivation started at 0800 hours, during the light phase and after the peak of feeding activity. The animals fed on the unpalatable diet did not show any signs of distress, did not lose any weight and, owing to the mild food deprivation, consumed the same amount of food as mice fed on the palatable diet. All mice were checked daily by the animal care technicians and ourselves. The weight range of the mice before and after the experiment was 26–31 g. The lack of differences in behaviour of demonstrators in the three food conditions in the 3-min test (see Results) also suggests the treatments did not cause any distress. There were no obvious adverse effects on the health of the mice. This study was carried out under a Ministero della Sanita` Italiano licence for animal care. RESULTS During the 3 min of interaction after reunion, no significant differences were found in the behaviour of demonstrators either in relation to observer condition or demonstrator food group (effects: observer: Wilk’s 3,57 =0.97, NS; demonstrator: Wilk’s 6,114 =0.87, NS; interaction: Wilk’s 6,114 =0.93, NS). In contrast, the observer’s behaviour varied with the demonstrator’s food group (Wilk’s 8,112 =0.74, P=0.02). Observers spent more time sniffing the nose/mouth area of demonstrators fed on the unpalatable diet than they did with unfed demonstrators (Tukey–Kramer: P<0.05). There were no effects of the observer’s food deprivation condition (Wilk’s 4,56 =0.94, NS) or of the observer*demonstrator interaction (Wilk’s 8,112 =0.95 , NS; Table 1).
691
ANIMAL BEHAVIOUR, 60, 5
Table 1. Time (s) spent in different behaviours by female mice (observers) during 3 min of social interaction with their cagemate (demonstrators) after 24 h of separation Demonstrator Observer
No food
Palatable diet
Unpalatable diet
Not food deprived Sniff nose Sniff body Self-groom Explore N
10.58±2.71 65.92±10.42 2.17±0.85 101.75±12.34 12
12.92±2.00 71.92±9.24 1.17±0.49 93.67±8.77 12
20.70±3.73 67.60±8.56 4.80±4.15 87.60±9.31 10
Food deprived Sniff nose Sniff body Self-groom Explore N
6.00±1.38 55.30±10.41 2.80±1.68 116.90±11.19 10
8.64±1.56 82.09±11.06 2.91±1.21 87.27±11.01 11
17.70±5.17 62.10±9.97 2.00±1.34 98.60±13.30 10
Means are given±SE. Observers were either deprived of food for 7 h or not deprived. Demonstrators were either not fed or were given palatable or unpalatable food for 30 min before the 3-min test.
100
100 (a)
(b) *†
80 Number of UVs
692
80
60
60 †
*‡
40
40
20
20
0
1
2
3
0
*
1
2
3
Time during test (min) Figure 1. Mean±SEM number of ultrasounds (UVs) emitted by (a) not food-deprived or (b) food-deprived observer female mice after reunion with a demonstrator that was either offered no food ( ) or fed on a palatable ( ) or an unpalatable ( ) diet. Statistical analysis was performed on the square-root transformed data. P levels (Tukey–Kramer) refer to comparisons within the same minute of observation. *Palatable diet versus no food; †palatable diet versus unpalatable diet; ‡unpalatable diet versus no food. All Ps<0.05.
In the light of our pilot study we assumed that the observer uttered the majority of UVs within a pair of observers/demonstrators. Observers uttered fewer UVs when food deprived than when not deprived (F1,59 =7.94, P<0.01; Fig. 1). The number of UVs was strongly affected by the palatability of the food eaten by the demonstrator (F2,59 =6.55, P<0.01) and this effect was dependent on the observer’s motivational state (interaction: F2,59 =3.60, P<0.05). The number of UVs decreased across the 3 min of observation (F2,118 =54.3, P<0.001). In the first minute no differences were found in UVs between the demonstrator groups, both for nondeprived (Fig. 1a) and deprived observers (Fig. 1b). During the second minute nondeprived observers uttered more UVs with demonstrators fed palatable food than with unfed demonstrators (Tukey–Kramer: P<0.05). They also uttered more UVs in the presence of demonstrators fed palatable food (in both
the second and the third minutes) than with demonstrators fed unpalatable food (Tukey–Kramer: P<0.05). Deprived observers uttered more UVs in the presence of fed demonstrators (palatable diet: second and third minutes; unpalatable diet: second minute) than with unfed demonstrators (Tukey–Kramer: P<0.05). DISCUSSION Ultrasonic communication in adult female mice has long been underinvestigated since it was considered a typical male behaviour. Our study shows that it is also an important component of the females’ behavioural repertoire. The UVs in our study were almost entirely uttered by the observer member of the pair, that is, the one that remained in its home cage. The number of UVs during the 3 min after reunion is comparable to the
MOLES & D’AMATO: ULTRASOUNDS AND FOOD CUES IN MICE
number of UVs of males of the same strain during courtship behaviour (D’Amato 1991). These calls were uttered immediately after reunion and were affected both by the motivational state of the vocalizing individual and the salience of the information carried by the conspecific. The first few minutes of interaction between female mice were characterized by intense social investigation and high levels of ultrasonic calling by the observer, whereas the demonstrator was mainly interested in the physical environment. Motivational state (food deprivation condition) did not affect the observer’s behaviour towards the demonstrator, although it strongly modulated the number of calls. A difference in the number of UVs emerged in the observers: food-deprived animals called less. Indeed, ultrasonic calling is a highly uneconomical behaviour (Nyby & Whitney 1978) and would be particularly costly for food-restricted animals. In contrast, the food eaten by the demonstrators affected both the behaviour and the number of calls of the observers. Observers differed in which region of the partner they investigated: the nose area of demonstrators fed the unpalatable food was investigated more than the same area of other demonstrator groups. Body sniffing, in contrast, did not vary with the demonstrator’s food. This suggests that some trace of this unpalatable food probably remained in the nose/mouth area, eliciting additional sniffing. However, the tendency of food-deprived observers to investigate this region less bears out the hypothesis that this behaviour is not related to the feeding motivational state of the animals. During the first minute of interaction, UVs did not differ significantly between groups, although nondeprived observers tended to vocalize less in response to a demonstrator fed on the unpalatable food. During the second and third minutes, statistically significant differences were detected in observers’ UVs according to both demonstrator group and food deprivation condition. In food-deprived observers more UVs were elicited in the second minute in the presence of fed animals irrespective of the quality of the food eaten by the partner. In the presence of a conspecific fed on the palatable food, vocalization was high until the end of the test. In contrast, nondeprived observers called more when the demonstrators were fed palatable food and this effect persisted until the last minute of the test. We conclude that the motivational state affects UVs towards a demonstrator conspecific, with nondeprived animals calling more towards demonstrators fed on palatable food, whereas food-deprived subjects vocalized more to fed conspecifics, independently of the palatability of the food eaten. From a functional viewpoint, UVs emitted in male– female pairs may facilitate mating by promoting proximity (Pomerantz et al. 1983). Bean et al. (1986) proposed the primary function of 70-kHz vocalizations to be ‘to announce the presence of another mouse and to attract other mice’. We suggest that these calls have an affiliative function, rather than a role in rank acquisition (Maggio & Whitney 1985). Both the infantile origin of this behaviour and the absence of UVs during male–male interactions (Sales 1972; Whitney et al. 1973) support this
hypothesis. In female mice UVs can facilitate proximity, helping relevant information to be acquired. Our study provided a context in which proximity allowed observers to identify the demonstrator’s food source. When the observers first approached the companion to gain general information (i.e. individual/sex recognition), UVs were unaffected by the quality of the demonstrator’s food. In contrast, after these data were acquired observers vocalized more to conspecifics carrying more attractive cues. In this study we used the palatability of food as a cue to make demonstrators more or less attractive to their observers. The possible role of UVs in the social acquisition of food preferences should be investigated. Further studies are needed to explore this behaviour in observers not familiar with the foods eaten by their conspecific demonstrators. As Galef (1993) pointed out, social interaction serves primarily as a means of expanding feeding repertoires rather than maintaining food habits. Acknowledgments A.M. was supported by a CNR scholarship Bando di Concorso 201.16.14. We thank two anonymous referees for their comments on the manuscript. References Altmann, J. 1974. Observational study of behavior: sampling methods. Behaviour, 49, 227–266. Bean, N. J., Nunez, A. & Wysocki, C. J. 1986. 70-kHz vocalizations by male mice do not inhibit aggression in lactating mice. Behavioral and Neural Biology, 46, 46–53. Brown, R. E. 1985. The rodents II: suborder Myomorpha. In: Social Odours in Mammals. (Ed. by R. E. Brown & D. W. MacDonald), pp. 389–457. Oxford: Clarendon Press. D’Amato, F. R. 1991. Courtship ultrasonic vocalizations and social state in mice. Animal Behaviour, 41, 875–885. D’Udine, B., Robinson, D. J. & Oliverio, A. 1982. An analysis of single gene effects on audible and ultrasonic vocalizations in the mouse. Behavioral and Neural Biology, 36, 197–203. Galef, B. G. Jr. 1993. Functions of social learning about food: a causal analysis of effects of diet novelty on preference transmission. Animal Behaviour, 46, 257–265. Galef, B. G. Jr. 1994. Olfactory communication about foods among rats: a review of recent findings. In: Behavioral Aspects of Feeding (Ed. by B. G. Galef, Jr, M. Mainardi & P. Valsecchi), pp. 83–101. Chur: Harwood Academic. Galef, B. G. Jr & Whiskin, E. E. 1998. Limits on social influence on food choice of Norway rats. Animal Behaviour, 56, 1015–1020. Galef, B. G. Jr & Wigmore, S. W. 1983. Transfer of information concerning distant foods: a laboratory investigation of the ‘information-centre’ hypothesis. Animal Behaviour, 31, 748–758. Grant, E. C. & Mackintosh, J. H. 1963. A comparison of the social postures of some common laboratory rodents. Behaviour, 21, 246–259. Maggio, J. & Whitney, G. 1985. Ultrasonic vocalizing by adult female mice, Mus musculus. Journal of Comparative Psychology, 99, 420–436. Mainardi, M., Poli, M. & Valsecchi, P. 1989. Ontogeny of dietary selection in weanling mice: effects of early experience and mother’s milk. Biology of Behavior, 14, 185–194.
693
694
ANIMAL BEHAVIOUR, 60, 5
Noirot, E. 1966. Ultrasounds in young rodents. II Changes with age in albino mice. Animal Behaviour, 14, 129–134. Nunez, A. A. & Tan, D. T. 1984. Courtship ultrasonic vocalizations in male SwissWebster mice: effects of hormones and sexual experience. Physiology and Behavior, 32, 717–721. Nunez, A. A., Nyby, J. & Whitney, G. 1978. The effects of testosterone, estradiol and dihydrotestosterone on male mice (Mus musculus) ultrasonic vocalizations. Hormones and Behavior, 11, 264–272. Nunez, A. A., Pomerantz, S. M., Bean, N. J. & Youngstrom, T. G. 1985. Effects of laryngeal denervation on ultrasound production and male sexual behavior in rodents. Physiology and Behavior, 34, 901–905. Nyby, J. & Whitney, G. 1978. Ultrasonic communication of adult miomorph rodents. Neuroscience Biobehavioural Reviews, 2, 1–14. Nyby, J., Dizinno, J. & Whitney, G. 1977. Sexual dimorphism in ultrasonic vocalization in mice: gonadal hormone regulation. Journal of Comparative and Physiological Psychology, 93, 957–975. Pomerantz, S. M., Nunez, A. A. & Bean, N. J. 1983. Female behavior is affected by male ultrasonic vocalizations in house mice. Physiology and Behavior, 31, 91–96.
Sales, G. D. 1972. Ultrasound and mating behaviour in rodents with some observations on other behavioural situations. Journal of Zoology, 168, 149–164. Valsecchi, P. & Galef, B. G. Jr. 1989. Social influences of the food preferences of house mice (Mus musculus). International Journal of Comparative Psychology, 2, 245–256. Valsecchi, P., Moles, A., Mainardi, D. & Mainardi, M. 1994. Individual and social experiences in the establishment of food preferences in mice. In: Behavioral Aspects of Feeding (Ed. by B. G. Galef, Jr, M. Mainardi & P. Valsecchi), pp. 103–124. Chur: Harwood Academic. White, N. R., Prasad, M., Barfield, R. J. & Nyby, J. 1998. 40 and 70-kHz vocalizations of mice, Mus musculus, during copulation. Physiology and Behavior, 63, 467–473. Whitney, G., Coble, J., Stockton, M. & Tilson, E. 1973. Ultrasonic emissions: do they facilitate courtship of mice? Journal of Comparative and Physiological Psychology, 84, 445–452. Wysocki, C. J., Nyby, J., Whitney, G., Beauchamp, G. K. & Katz, Y. 1982. The vomeronasal organ: primary role in mouse chemosensory gender recognition. Physiology and Behavior, 29, 315–327.