Animal Behaviour 82 (2011) 399e404

Contents lists available at ScienceDirect

Animal Behaviour
journal homepage: www.elsevier.com/locate/anbehav

Snifng with the right nostril: lateralization of response to odour stimuli by dogs
Marcello Siniscalchi a, *, Raffaella Sasso a,1, Anna M. Pepe a,1, Salvatore Dimatteo a,1, Giorgio Vallortigara b, 2, Angelo Quaranta a,1
a b

Department of Animal Production, University of Bari Aldo Moro, Italy Centre for Mind/Brain Sciences, University of Trento, Italy

a r t i c l e i n f o
Article history: Received 2 February 2011 Initial acceptance 15 March 2011 Final acceptance 13 May 2011 Available online 25 June 2011 MS. number: 11-00105 Keywords: animal welfare Canis familiaris dog emotion laterality olfaction physiology

Lateralization in dogs, Canis familiaris, has been reported for paw usage and response to visual and acoustic stimuli. Surprisingly, however, no investigation of possible lateralization for the most relevant sensory domain of dogs, namely olfaction, has been carried out. Here we investigated left and right nostril use in dogs freely snifng different emotive stimuli in unrestrained conditions. When snifng novel nonaversive stimuli (food, lemon, vaginal secretion and cotton swab odours), dogs showed initial preferential use of the right nostril and then a shift towards use of the left nostril with repeated stimulus presentation. When snifng arousal stimuli such as adrenaline and veterinary sweat odorants, dogs showed a consistent right nostril bias all over the series of stimulus presentations. Results suggest initial involvement of the right hemisphere in processing of novel stimuli followed by the left hemisphere taking charge of control of routine behaviour. Sustained right nostril response to arousal stimuli appears to be consistent with the idea that the sympathetic hypothalamicepituitaryeadrenal axis is mainly under the control of the right hemisphere. The implications of these ndings for animal welfare are discussed. 2011 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.

Brain lateralization, both structural and functional, has been observed in several nonhuman species (Bradshaw & Rogers 1993; Rogers & Andrew 2002; Vallortigara & Rogers 2005), including canine species (Wells 2003; Quaranta et al. 2004, 2007; Siniscalchi et al. 2008, 2010a; Guo et al. 2009). Functional lateralization at the population level has been reported for auditory and visual sensory modalities. For instance, an advantage of the right ear in processing conspecic vocalizations and of the left ear in processing threatening stimuli has been observed (Siniscalchi et al. 2008); visual stimuli of high emotional valence have been shown to elicit preferential turning to the left side, probably as a result of selective activation of contralateral brain structures (Siniscalchi et al. 2010c). As of yet, however, no lateralization research has been conducted on a most crucial sensory domain for dogs, Canis familiaris: that is, olfaction.

* Correspondence and present address: M. Siniscalchi, Department of Animal Production, University of Bari Aldo Moro, Strada Prov. le per Casamassima, Km 3, 70010 Valenzano, Italy. E-mail address: m.siniscalchi@veterinaria.uniba.it (M. Siniscalchi). 1 R. Sasso, A. M. Pepe, S. Dimatteo and A. Quaranta are at the Department of Animal Production, University of Bari Aldo Moro, Strada Prov.le per Casamassima, Km 3, 70010 Valenzano, Italy. 2 G. Vallortigara is at the Centre for Mind/Brain Sciences, University of Trento, Corso Bettini 31, 38068 Rovereto, Italy.

Studies suggest the presence of a lateralized process in the analysis of odours in both vertebrate and invertebrate species (Rogers & Andrew 2002; Royet & Plailly 2004; McGreevy & Rogers 2005; De Boyer Des Roches et al. 2008; Rogers & Vallortigara 2008). Domestic chicks, Gallus gallus, for example, show better discrimination of imprinting olfactory stimuli when using their right nostril (Vallortigara & Andrew 1994; Rogers et al. 1998) and stronger head shaking to a noxious odour under right nostril stimulation (Burne & Rogers 2002). Homing pigeons, Columba livia, show impaired initial orientation when olfactory input is conned to the right nostril only (Gagliardo et al. 2007). Horses, Equus caballus, show a population bias to use the right nostril rst in response to stallion faeces (McGreevy & Rogers 2005) and in response to novel objects (De Boyer Des Roches et al. 2008). Honeybees show behavioural, electrophysiological and morphological asymmetries during olfactory learning favouring their right antenna during initial memory recall (Anfora et al. 2010; Frasnelli et al. 2010a,b). In humans, despite ndings demonstrating that basic olfactory perceptual processes appear lateralized between the hemispheres (summarized in Brancucci et al. 2009), some controversy exists over the hemispheric role in the processing of odours: although Broman et al. (2001) reported a right-nostril advantage for perceived odour familiarity, human neonates respond to an emotional positive odour (smells with a high affective valence and presumably with a high degree of familiarity) presented to the left

0003-3472/$38.00 2011 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.anbehav.2011.05.020

400

M. Siniscalchi et al. / Animal Behaviour 82 (2011) 399e404

nostril (left hemisphere) by head turning but do so signicantly less often when the odour is presented to the right nostril (Olko & Turkewitz 2001). Furthermore in humans, a right-nostril dominance was also reported in unfamiliar but not familiar odour discrimination performance (Savic & Berglund 2000). The aim of the present study was to investigate whether dogs show nostril asymmetries in processing odorants that differ in terms of familiarity and emotional valence during free snifng behaviour under unrestrained conditions.

were conducted daily for a walk (about 40 min per day) into an exercise area inside the kennel. The experiment was carried out in a large room (4 5 m and 3 m high) isolated from the rest of the kennel, where each dog was kept on a lead separately during the test. Testing Apparatus and Procedure The study involved presentation of dogs with different olfactory stimuli to assess whether there was any tendency to use one nostril preferentially rather than another during snifng behaviour. Odorants used as test stimuli were presented on cotton swabs commonly used for canine vaginal cytology (FL Medical, Torreglia, Italy). Odorants used comprised: (1) food (complete commercial food for adult dogs; Royal Canin, Milan, Italia); (2) sweat of the veterinarian associated with the kennel; (3) lemon; (4) adrenaline (Adrenaline F.U. 1 mg/1 ml Galenica Senese, Siena, Italy); (5) oestrous bitch vaginal secretions; and (6) the cotton swab without any particular odour. Odorants were changed regularly to keep the odour quality constant. The veterinarian was instructed not to use deodorant/antiperspirant for 2 days before the experiment until after the collection of the swab was over, and to take only a shower on the morning of the experiment. On the day of the collection, the cotton swab was placed under the armpits of the vet for 10 min and then stored at 80  C until testing (several samples were collected). Lemon odour samples were obtained directly from natural lemon juice and then stored at 80  C. Oestrous secretion was sampled by inserting the cotton swab directly through the vulvar lips into the vagina of a healthy bitch who showed both clinical and behavioural signs of oestrous. Different samples were collected and then stored at 80  C until the experiment. The cotton swab impregnated with different odours was installed on a digital video camera located on a tripod in the centre of the testing area (see Fig. 1); the tripod and the entrance of the

METHODS Subjects and Housing The subjects were 30 neutered mongrel dogs (15 males and 15 females) kept in a large rural kennel associated with the Faculty of Veterinary Medicine of Bari University, Italy. All dogs, aged between 2 and 8 years (5.4 1.04; mean years SE), were housed individually in a large area (270 m2) with 30 runs separated by cement block partition walls (2.50 m high), concrete oors and an exposed rafter ceiling (2.50 m high). Each run comprised two sections: an indoor insulated section (sleeping area, 3 m2) and an outdoor section (6 m2) joined by a gangway that could be closed by a metal door. Dogs were conned to the sleeping area overnight (1900e 0700 hours) to prevent extremes of temperature (temperatures range between an average minimum of 17  C in winter and an average maximum of 25  C in summer). Suitable bedding equipment was also provided and was periodically cleaned and sanitized. During the day, the environment was illuminated with a combination of natural and uorescent light. Complete commercial food for adult dogs was provided twice during the day at approximately 0700 and 1800 hours, whereas water was available ad libitum. Clinical examination and coprological analyses for parasites were carried out for all the dogs in the sample. In addition, dogs

Figure 1. Schematic representation of the testing apparatus.

M. Siniscalchi et al. / Animal Behaviour 82 (2011) 399e404

401

testing room were aligned in a straight line. Two plastic panels (30 cm high, 50 cm deep) were located on the two sides of the tripod to favour a centred position of the dog with respect to the cotton swab and the video recording area during the experiment. On the day of the experiment, each dog, in turn, was taken to the testing area on a lead, and after the dog was introduced into the testing room the lead was removed. The behaviour of the animal was recorded continuously during olfactory inspection. Each dog was tested during a single session of 2 h, twice a week over 3.5 weeks, until a set of seven trials each stimulus each dog was collected. Each session consisted of the presentation of all six different olfactory stimuli in random order (if dogs did not start snifng within 2 min the test was stopped) and there was a 1 min interval between each presentation. Each presentation lasted 3 min. Over 10 weeks, the experimental group completed all the trials. Data Analyses All video recording was analysed frame by frame, and left and right nostrils use was scored using different indexes (below) by a trained observer who was blind to the odour stimulus presented to the animal (see Movie S1 in the Supplementary Material). Lateral asymmetries in the use of the rst nostril were computed using the index: LIf (rst) (L R/L R), where L and R indicate, respectively, the number of times in which each dogs used the left or the right nostril as the rst during the seven stimulus presentations. Hence a score of 1.0 indicated exclusive use of the left nostril as rst and a score of 1.0 indicated exclusive use of the right nostril as rst. An LI score of 0 indicated equal numbers of left and right nostril rst use. Signicant departures from chance level (0) were estimated by two-tailed one-sample t tests. In a similar way, lateral asymmetries in last nostril use before the end of the trial were computed using the following index: Lil (last) (L R/L R), where L and R indicate, respectively, the number of times in which each dog used the left or the right nostril as the last one during the seven trials. Signicant departures from chance level (0) were estimated by two-tailed one-sample t tests. Lateral asymmetries in the total time spent snifng with each nostril were computed using the index: Lit (total) (L R/L R), where L and R indicated, respectively, the total time (in seconds) spent snifng with the left and the right nostril. Signicant departures from chance level (0) were estimated by two-tailed onesample t tests.

Finally, the total odour investigation time of all odour stimuli (irrespectively of the left/right nostril use) during the seven trials was calculated. For all statistical tests, SPSS software (SPSS Inc., Chicago, IL, U.S.A.) was used, and the results were considered statistically signicant for P < 0.05. The analysis of variance was used on the various indexes to determine the effects of sex and repeated trials. The experiments were conducted according to the protocols approved by the Italian Minister for Scientic Research in accordance with EC regulations. RESULTS Results for rst nostril use are shown in Fig. 2 (white bars). The analysis of variance revealed a signicant main effect of the type of odour (ANOVA: F5,140 3.382, P 0.006); no signicant effect of sex (F1,28 2.240, P 0.106) or sex*type of odour interaction (GLM F5,140 2.049, P 0.075) were apparent. Dogs consistently used their right nostril at rst to sniff the sweat of the veterinarian (two-tailed one-sample t test: t29 3.070, P 0.005), the lemon odour (t29 3.427, P 0.002), the adrenaline stimulus (t29 4.904, P < 0.001) and the vaginal secretion (t29 2.847, P 0.008); there was a similar trend for the neutral stimulus (t29 2.027, P 0.052), whereas response to food did not reveal any bias (t29 1.150, P 0.882). Use of the last nostril (Fig. 2, black bars) also showed a signicant main effect of the type of odour (ANOVA: F5,140 15.918, P < 0.001); there was no effect of sex (F1,28 0.867, P 0.360) and sex*type of odour interaction (F5,140 1.682, P 0.143) in last nostril use. Preferential use of the left nostril was apparent for the food stimulus (two-tailed one-sample t test: t29 6.916, P < 0.001); whereas preferential use of the right nostril use was apparent for the sweat of the veterinarian (t29 3.588, P 0.001) and the adrenaline odour stimuli (t29 6.267, P < 0.001). No signicant biases were observed for vaginal secretion (t29 1.521, P 0.139), lemon (t29 0.855, P 0.400) and cotton swab (t29 0.065, P 0.949). Total time spent using either the left or right nostril is shown in Fig. 3. There was a signicant main effect of trials (ANOVA: F6,168 12.646, P < 0.001), odours (F5,140 15.047, P < 0.001), and trials*odours interaction (F30,840 1.678, P 0.013); no signicant effect of sex (F1,28 1.803, P 0.190), sex*trials (F6,168 0.951, P 0.460) or sex*odours (F5,140 1.571, P 0.172) and sex*trials*odours interactions (F30,840 0.989, P 0.484) were revealed.

Left nostril

0.5 0.4 0.3 * First nostril Last nostril

Laterality index

0.2 0.1 0 0.1 0.2 0.3 0.4 * * Lemon Cotton swab * * Vet odour

Right nostril 0.5

* * Adrenaline

Food

Vaginal secretion

Figure 2. Laterality index of the rst and last nostril used during inspection of different odorants (group means averaged across all seven trials with SEM are shown). Asterisks indicate signicant biases (*P < 0.05, two-tailed one-sample t tests).

402

M. Siniscalchi et al. / Animal Behaviour 82 (2011) 399e404

0.7 0.5 0.3 0.1 0.1 0.3 0.5 0.7 1 2 3

Food

* *

0.7 0.5 0.3 0.1

* Oestrous bitch vaginal secretion *

0.1 0.3 0.5 0.7

* * Cotton swab *

Left nostril

0.7 0.5

Lemon *

0.7 0.5 0.3 0.1

Laterality index

(total time)

0.3 0.1 0.1 0.3 0.5 * * 1 2 3 4 5 6 7

0.1 0.3 0.5 0.7 0.7 0.5 0.3 0.1

Right nostril 0.7 0.7 0.5 0.3 0.1 0.1 0.3 0.5 0.7 1

* *

Vet odour

Adrenaline

0.1 0.3 0.5

* *

0.7 Presentations

* *

Figure 3. Laterality index in the total time spent snifng with the right and left nostril with repeated stimulus presentations (group means with SEM are shown). Asterisks indicate signicant biases (*P < 0.05; **P < 0.01, two-tailed one-sample t tests; P value adjusted using Bonferroni correction).

There was signicant heterogeneity associated with trials for the food (F6,120 7.640, P < 0.001), the vaginal secretion (F6,66 4.977, P < 0.001), the lemon (F6,108 3.265, P 0.005) and the cotton swab (F6,90 4.091, P 0.001) odorants. Food stimulus showed a trend in the use of the right nostril during the rst presentation, no bias from the second to the fourth presentation and then a signicant left bias during the last three presentations; both vaginal secretion and cotton swab stimuli showed a right nostril bias during the rst two presentations, no bias from the third to the sixth presentation and then a left bias on the last presentation (see Fig. 3); similarly, lemon stimulus showed a right nostril bias during the rst three presentations, no bias from the fourth to the sixth presentation, and a left bias on the last presentation. With adrenaline stimulus and sweat of the veterinarian, no signicant variation with repeated stimulus presentation was observed (respectively, F6,84 0.416, P 0.866; F6,48 0.880, P 0.517). Dogs showed a consistent right nostril preference over the test (adrenaline: two-tailed one-sample t test: t29 6.058, P < 0.001; sweat of the veterinarian: t29 6.335, P < 0.001).

Regarding the total odour investigation time, the analysis of variance revealed a signicant main effect of the type of the stimulus (F5,130 30.073, P < 0.001). Total time spent snifng was higher with the vaginal secretion (mean SD, 1.90 1.03) and food stimuli (mean SD, 1.38 0.69) than with the other odour stimuli (post-hoc analysis Fishers protected LSD: P < 0.01); there was also a statistically signicant difference between vaginal secretion and food odours (post-hoc analysis Fishers protected LSD: P < 0.01); dogs tended to spend less time snifng at neutral stimulus compared with all other odours (mean SD, 1.90 1.03; post-hoc analysis Fishers protected LSD P < 0.01 in all measures). Finally, there was no statistically signicant difference between lemon, adrenaline and vet odours (mean SD, lemon: 0.87 0.04; adrenalin: 0.79 0.07; vet odour: 0.79 0.07). There was also a signicant main effect of session (F6,156 13.524, P < 0.001); contrasts revealed a signicant decreasing in total odour investigation time from session 1 to session 2 (F1,26 13.77, P 0.001) and from session 2 to session 3 (F6,156 4.510, P 0.043). There was no signicant effect of sex (F1,26 0.789, P 0.382).

M. Siniscalchi et al. / Animal Behaviour 82 (2011) 399e404

403

(a)

(b)

Total odour investigation time (s)

Food Vet odour Lemon Adrenaline Vaginal secretion Cotton swab

Presentations
Figure 4. Total odour investigation time with repeated stimulus presentations in (a) male and (b) female dogs (group means with SE are shown).

Results revealed a signicant interaction effect between sex and type of odour stimulus (F6,156 13.524, P < 0.001; Fig. 4). To break down this interaction, contrasts were performed comparing the total investigation time of each stimulus to average snifng time of odours across male and female subjects. These revealed that males tended to spent more time snifng at vaginal secretion than female (F1,26 4.530, P 0.043; male: means SD, 2.165 0.260; female: means SD, 1.481 0.260) and that female dogs tended to spent more time snifng at food stimulus than male (F6,156 7.336, P 0.012; male: means SD, 1.180 0.186; female: means SD, 1.514 0.186). There was also a signicant interaction effect between stimulus type and session (F30,780 4.601, P < 0.001). Contrasts revealed a signicant interaction when comparing session 1 with session 2 when lemon stimulus is compared with the neutral one (F1,26 10.405, P 0.003); the second contrast revealed a signicant interaction when comparing session 1 with session 2 when vaginal secretion is compared with the neutral one (F1,26 4.386, P 0.046). These data indicate that the decreasing total odour investigation time from session 1 to session 2 with respect to the neutral one (control stimulus) is stressed more for lemon and vaginal secretion stimuli. Finally, there was no signicant session*sex (F6,156 1.546, P 0.167) and stimulus*session*sex (F30,780 0.990, P 0.483) interactions. DISCUSSION During snifng of nonaversive stimuli (food, lemon, vaginal secretion and cotton swab), dogs showed initial use of the right nostril and then a shift towards use of the left nostril. This pattern was very clear for time spent at snifng, less so for initial snifng in which food and lemon did not show any clear bias, possibly because of their reduced valence as novel stimuli. Although the nasal cycle (i.e. the periodic engorgement of the erectile tissue surrounding the turbinates via autonomic nervous control) has been reported also in canine species (Webber et al. 1987), it is not expected to affect temporal changes in nostril preference related to odour perception, as in dogs the lamina

propria below the olfactory epithelium does not contain a rich vascular network and the thickness of the olfactory mucosa does not change appreciably during the nasal cycle (Craven et al. 2007). Given that in mammals the olfactory system ascends mainly ipsilaterally, with most receptor information from each nostril projecting, via the olfactory bulb, to the primary olfactory cortex in the same hemisphere (Royet & Plailly 2004), the pattern of nostril use suggests initial involvement of the right hemisphere followed by involvement of the left hemisphere. The very same pattern has been observed repeatedly in the vertebrate brain, in a variety of species (e.g. birds: Vallortigara et al. 1999a; sh: Sovrano 2004), with the right hemisphere taking charge of novel information followed by the left hemisphere taking charge of behaviour when routine responses to stimuli emerge as a result of familiarization/categorization (reviewed in Vallortigara et al. 1999b, 2011; Vallortigara 2000; Rogers & Andrew 2002; MacNeilage et al. 2009). By contrast, adrenaline and veterinary sweat odorants were associated with a consistent right nostril bias all over the series of stimulus presentations. These are clearly arousal stimuli to dogs. In particular, the right nostril (right hemisphere) bias in response to adrenaline stimulus is consistent with the idea that the hypothalamicepituitaryeadrenal axis is mainly under the control of the right hemisphere, which has been associated with the expression and control of intense emotions, such as aggression, escape behaviour and fear (Andrew & Rogers 2002). In particular, it has been suggested that the right hemisphere is preferentially involved in sympathetic activation and the left hemisphere is preferentially involved in parasympathetic activity associated with reduced tension or calm responses (Craig 2005). Although the sweat of the veterinarian associated with the kennel was likely to be quite a familiar odour for the dogs, it was associated with stressful activities required by veterinary routine examinations in the kennel (blood sampling, vaccine administration, etc.). Hence the vet odour was likely to increase the arousal state of the dog and thus the activity of the right hemisphere (see also Siniscalchi et al. 2010b).

404

M. Siniscalchi et al. / Animal Behaviour 82 (2011) 399e404 Cain, W. S. 1982. Odor identication by males and females: predictions vs. performance. Chemical Senses, 7, 129e142. Craig, A. D. 2005. Forebrain emotional asymmetry: a neuroanatomical basis? Trends in Cognitive Sciences, 912, 566e571. Craven, B. A., Neuberger, T., Paterson, E. G., Webb, A. G., Josephson, E. M., Morrison, E. E. & Settles, G. S. 2007. Reconstruction and morphometric analysis of the nasal airway of the dog (Canis familiaris) and implications regarding olfactory airow. Anatomical Record, 290, 1325e1340. De Boyer Des Roches, A., Richard-Yris, M. A., Henry, S., Ezzaoua, M. & Hausberger, M. 2008. Laterality and emotions: visual laterality in the domestic horse (Equus caballus) differs with objects emotional value. Physiology & Behavior, 94, 487e490. Frasnelli, E., Anfora, G., Trona, F., Tessarolo, F. & Vallortigara, G. 2010a. Morphofunctional asymmetry of the olfactory receptors of the honeybee (Apis mellifera). Behavioural Brain Research, 209, 221e225. Frasnelli, E., Vallortigara, G. & Rogers, L. J. 2010b. Response competition associated with righteleft antennal asymmetries of new and old olfactory memory traces in honeybees. Behavioural Brain Research, 209, 36e41. Gagliardo, A., Pecchia, T., Savini, M., Odetti, F., Ioale, P. & Vallortigara, G. 2007. Olfactory lateralization in homing pigeons: initial orientation of birds receiving a unilateral olfactory input. European Journal of Neuroscience, 25, 1511e1516. Guo, K., Meints, K., Hall, C., Hall, S. & Mills, D. 2009. Left gaze bias in humans, rhesus monkeys and domestic dogs. Animal Cognition, 12, 409e418. McGreevy, P. D. & Rogers, L. J. 2005. Motor and sensory laterality in thoroughbred horses. Applied Animal Behaviour Science, 92, 337e352. MacNeilage, P. F., Rogers, L. J. & Vallortigara, G. 2009. Origins of the left and right brain. Scientic American, 301, 60e67. Olko, C. & Turkewitz, G. 2001. Cerebral asymmetry of emotion and its relationship to olfaction in infancy. Laterality, 6, 29e37. Pierman, S., Douhard, Q., Balthazart, J., Baum, M. J. & Bakker, J. 2006. Attraction thresholds and sex discrimination of urinary odorants in male and female aromatase knockout (ArKO) mice. Hormones and Behavior, 49, 96e104. Quaranta, A., Siniscalchi, M., Frate, A. & Vallortigara, G. 2004. Paw preference in dogs: relations between lateralised behaviour and immunity. Behavioural Brain Research, 153, 521e525. Quaranta, A., Siniscalchi, M. & Vallortigara, G. 2007. Asymmetric tail-wagging responses by dogs to different emotive stimuli. Current Biology, 17, 199e201. Rogers, L. J. & Andrew, R. J. 2002. Comparative Vertebrate Lateralization. Cambridge: Cambridge University Press. Rogers, L. J. & Vallortigara, G. 2008. From antenna to antenna: lateral shift of olfactory memory in honeybees. PloS One, 3, e2340. Rogers, L. J., Andrew, R. J. & Burne, T. H. 1998. Light exposure of the embryo and development of behavioural lateralisation in chicks, I: olfactory responses. Behavioural Brain Research, 97, 195e200. Royet, J. P. & Plailly, J. 2004. Lateralization of olfactory processes. Chemical Senses, 29, 731e745. Savic, I. & Berglund, H. 2000. Right-nostril dominance in discrimination of unfamiliar, but not familiar, odours. Chemical Senses, 25, 517e523. Siniscalchi, M., Quaranta, A. & Rogers, L. J. 2008. Hemispheric specialization in dogs for processing different acoustic stimuli. PloS One, 3, e3349. doi: 0.1371/ journal.pone.0003349. Siniscalchi, M., Franchini, D., Pepe, A. M., Sasso, R., Vallortigara, G. & Quaranta, A. 2010a. Volumetric assessment of cerebral asymmetries in dogs. Laterality, 6, 1e9. Siniscalchi, M., Sasso, R., Pepe, A. M., Dimatteo, S., Vallortigara, G. & Quaranta, A. 2010b. Catecholamine plasma levels following immune stimulation with rabies vaccine in dogs selected for their paw preferences. Neuroscience Letters, 476, 142e145. Siniscalchi, M., Sasso, R., Pepe, A. M., Vallortigara, G. & Quaranta, A. 2010c. Dogs turn left to emotional stimuli. Behavioural Brain Research, 208, 516e521. Sovrano, V. A. 2004. Visual lateralization in response to familiar and unfamiliar stimuli in sh. Behavioural Brain Research, 152, 385e391. Vallortigara, G. 2000. Comparative neuropsychology of the dual brain: a stroll through left and right animals perceptual worlds. Brain and Language, 73, 189e219. Vallortigara, G. & Andrew, R. J. 1994. Olfactory lateralization in the chick. Neuropsychologia, 32, 417e423. Vallortigara, G. & Rogers, L. J. 2005. Survival with an asymmetrical brain: advantages and disadvantages of cerebral lateralization. Behavioral and Brain Sciences, 28, 575e589. Vallortigara, G., Regolin, L. & Pagni, P. 1999a. Detour behaviour, imprinting, and visual lateralization in the domestic chick. Cognitive Brain Research, 7, 307e320. Vallortigara, G., Rogers, L. J. & Bisazza, A. 1999b. Possible evolutionary origins of cognitive brain lateralization. Brain Research Reviews, 30, 164e175. Vallortigara, G., Chiandetti, C. & Sovrano, V. A. 2011. Brain asymmetry (animal). Wiley Interdisciplinary Reviews: Cognitive Science, 2, 146e157. doi:10.1002/wcs.100. Webber, R. L., Jeffcoat, M. K., Harman, J. T. & Ruttimann, U. E. 1987. MR demonstration of the nasal cycle in the beagle dog. Journal of Computer Assisted Tomography, 11, 869e871. Wells, D. L. 2003. Lateralised behaviour in the domestic dog, Canis familiaris. Behavioural Processes, 61, 27e35.

Regarding the total odour investigation time, neutral odour was the less investigated stimulus, as expected, and more intriguingly vet odour and adrenaline stimuli had exactly the same investigation time, which was less than those for the other odours, stressing the emotionally arousing component of these stimuli. In addition, males tended to sniff vaginal secretion odour more than females, in line with the work of Pierman et al. (2006), which showed that in mice same-sex urine odours are less salient cues that went unnoticed when more attractive, opposite-sex urine odours were presented rst. Regarding food odour, females tended to have higher total investigation time than males: the presence of such sex difference has also been reported in humans since Cain (1982) rst observed a general female superiority at identifying food odours. This is the rst evidence that dogs show striking asymmetries of nostril use in completely unrestrained and relatively natural conditions; further investigations are required to establish whether this asymmetry is observed also at the structural levels (e.g. asymmetric receptor expression on olfactory epithelium) as reported in other animal models (Frasnelli et al. 2010a). Apart from contributing to our understanding of the evolution of brain asymmetry in animals, the very existence of such biases open the door to their exploitation in animal welfare and the design of food dispenser and other objects, which are commonly used when negotiating with our pet dogs. For example, the constant use of the right nostril during olfactory inspection of a stimulus (an object or especially a human being) could reveal an increased arousal state of dogs also in absence of clear behavioural signs. This could be extremely useful in dogs used in animal-assisted therapy and activities, as these animals must possess advanced behavioural control skills to help them handle unexpected or stressful situations, and consequently it is not always easy to detect arousal increase in these subjects directly from behavioural signs. Acknowledgments This research is part of the following projects: The asymmetrical mind of the dog: Behavioural lateralization in animal welfare, supported by the Waltham Foundation through a nancial grant to G.V. and A.Q.; Nostril preference in food preference, supported by Royal Canin Italia through a nancial grantship to A.Q.; Brain asymmetries in dogs supported by the University of Bari, Italy through a nancial grant to M.S. Supplementary Material Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.anbehav.2011.05.020. References
Andrew, R. J. & Rogers, L. J. 2002. The nature of lateralization in tetrapods. In: Comparative Vertebrate Lateralization (Ed. by L. J. Rogers & R. J. Andrews). Cambridge: Cambridge University Press, pp. 94e125. Anfora, G., Frasnelli, E., Maccagnani, B., Rogers, L. J. & Vallortigara, G. 2010. Behavioural and electrophysiological lateralization in a social (Apis mellifera) and in a non-social (Osmia cornuta) species of bee. Behavioural Brain Research, 206, 236e239. Bradshaw, J. L. & Rogers, L. J. 1993. The Evolution of Lateral Asymmetries, Language, Tool Use, and Intellect. New York: Academic Press. Brancucci, A., Lucci, G., Mazzatenta, A. & Tommasi, L. 2009. Asymmetries of the human social brain in the visual, auditory and chemical modalities. Philosophical Transactions of the Royal Society B, 364, 895e914. Broman, D. A., Olsson, M. J. & Nordin, S. 2001. Lateralization of olfactory cognitive functions: effects of rhinal side of stimulation. Chemical Senses, 26, 1187e1192. Burne, T. H. & Rogers, L. J. 2002. Chemosensory input and lateralization of brain function in the domestic chick. Behavioural Brain Research, 133, 293e300.