Lateralization of social learning in the domestic chick, Gallus gallus domesticus: learning to avoid

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

http://www.elsevier.com/copyright

Author's personal copy

Lateralization of social learning in the domestic chick,Gallus gallus domesticus: learning to avoid

Orsola Rosa Salva a, Jonathan Niall Daisley a,*, Lucia Regolin a, Giorgio Vallortigara b,1

a Department of General Psychology, University of Padovab Center for Mind/Brain Sciences, University of Trento

a r t i c l e i n f o

Article history:Received 13 March 2009Initial acceptance 6 April 2009Final acceptance 27 June 2009Published online 7 August 2009MS. number: 09-00171

Keywords:chickGallus gallus domesticuslateralizationpassive avoidance learningright hemispheresocial learning

Social learning occurs through the observation of conspecifics performing biologically relevant behav-iours. Like other forms of learning, aptitude for such tasks may be influenced by cerebral lateralization.Social recognition is subject to lateralization and it seems highly likely that any lateralization effectwould be transposed onto a social-learning situation. We used a social version of a passive avoidance taskin which one chick (the demonstrator) pecked at a red bead while another chick (the observer) viewedthe demonstrator’s response. This bead was either coated in the bitter-tasting substance methylan-thranilate (MeA) or was left dry. Thirty minutes later both chicks were presented with a similar, dry, beadto determine whether learning had taken place. Experiment 1 showed that chicks were able to learn, byobservation only, not to peck the bead when it had been coated in MeA at training. Experiment 2demonstrated a role of lateralization in performing the task; specifically, male observer chicks trainedbinocularly but tested monocularly had a poorer performance in this task with the left hemisphere thanwith the right. In experiment 3, we both trained and tested chicks monocularly. This time there was nodifference between left-eyed and right-eyed chicks, suggesting a spontaneous bias for encoding relevantinformation available at training in the right hemisphere. In conclusion, chicks are able to avoid peckingjust by observing a conspecific. The social learning appears to be lateralized to the right hemisphere inmale chicks, females showing no bias in lateralization.� 2009 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.

The ability to learn from the behaviour of conspecifics confers animportant adaptive advantage, allowing an individual to benefitfrom the experiences of others without the need for direct inter-action itself. For example, social learning may result in increasedforaging efficiency by reducing the costs required for learning, bydirect experience, the features and edibility of different foodsources (Nicol 2004, 2006). Chicks of the junglefowl, Gallus gallusspadiceus (the wild ancestor of domestic chickens) do not have aninnate ability to recognize edible food items, and tend to peck atboth food and nonfood objects (Hogan 1973). Learning to recognizeedible items is thus required in this species, but direct learningarising from the consequences of ingestion does not seem to bepossible shortly after hatching, increasing the importance of sociallearning (Nicol 2004, 2006). Birds including gallinaceous species,seem to be capable of learning about the availability and quality offood sources via social mechanisms. The mother chicken influences

the feeding behaviours of her chicks by attracting their attentiontowards suitable food sources in the environment (Moffat & Hogan1992; Nicol & Pope 1996; Nicol 2004, 2006; Allen & Clarke 2005).The feeding behaviours of chickens are influenced not only bymaternal food displays, but also by the behaviour of a peerdemonstrator (McQuoid & Galef 1992, 1993, 1994; Nicol & Pope1992, 1993, 1994, 1999; Gajdon et al. 2001; Sherwin et al. 2002;Caldwell & Whiten 2003; Nicol 2004; for similar evidence of sociallearning in chicks, but with an artificial demonstrator see Barta-shunas & Suboski 1984; Suboski & Bartashunas 1984).

The ability to refrain from ingesting items that conspecificsavoid eating could have as much adaptive value as the tendency toconsume food preferred by social partners by reducing the risk ofconsuming potentially poisonous items. When hens observe chickspecking at a kind of food that they had experienced as beingunpalatable, they increase behavioural displays to direct the chicks’attention towards other, palatable food (Nicol & Pope 1996; Nicol2004; 2006). This result suggests that maternal displays and theconsequent social learning of the chicks could serve to redirect thechicks’ attention away from unpalatable food sources. Adult hensseem incapable of learning to avoid pecking at a food that eliciteda ‘disgust’ reaction in another hen (Sherwin et al. 2002). This lack ofavoidance may have been because although demonstrators eating

* Correspondence: J. N. Daisley, Department of General Psychology, University ofPadova, Via Venezia 8, 35131 Padova, Italy.

E-mail address: [email protected] (J.N. Daisley).1 G. Vallortigara is at the Center for Mind/Brain Sciences, University of Trento,

Corso Bettini 31, I-38068 Rovereto, Italy.

Contents lists available at ScienceDirect

Animal Behaviour

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

0003-3472/$38.00 � 2009 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.doi:10.1016/j.anbehav.2009.06.021

Animal Behaviour 78 (2009) 847–856


the unpalatable food showed a disgust reaction, they continuedeating, thus providing potentially conflicting information to theirobservers (Nicol 2004). This result is inconsistent with evidenceindicating the presence of social learning for pecking avoidance inday-old chicks (Johnston et al. 1998). This difference may be a resultof the different ages of the subjects: newborn chicks learning forthe first time to distinguish food with respect to nonfood itemscould be more sensitive to social learning derived from the obser-vation of the feeding behaviours of others. This sensitivity may bereduced as the chicks mature and the consequences of ingestionmay become an important source of information about foodpalatability (Nicol 2004). In addition, Johnston et al’s. (1998)procedure was radically different from that of Sherwin et al. (2002).Johnston et al. used a modified version of the one-trial passiveavoidance learning (PAL) task described by Lossner & Rose (1983).This task exploits the tendency of young chicks to peck at any smallobject presented to them. The task involves the presentation ofa coloured bead coated in a bitter-tasting substance (usuallymethylanthranilate, MeA) to a chick. Chicks usually peck at thebead but pecking MeA is an aversive experience, following whichchicks display a disgust response, which includes behaviours suchas head shaking and bill-wiping. After undergoing this training,chicks will subsequently avoid pecking at a bead of similar colourand size (but not at a bead of a different colour and/or size) for 24 hor more (Cherkin 1969; Lossner & Rose 1983).

In the PAL task, chicks are usually trained and tested in pairs (Nget al. 1991), but social information transmission between cage-mates does not seem to be present when the standard version ofthe procedure is used (Gibbs & Ng 1977). Nevertheless, Johnstonet al. (1998), who used a slightly different version of the procedure,found convincing evidence showing just such an occurrence. Theyused pairs of chicks in which one of the pair (the ‘demonstrator’)was presented with a chrome bead dipped in either MeA or water.The second chick in each pair was termed the ‘observer’, and wasprevented from pecking at the training chrome bead. During thetest phase the demonstrator and observer chicks were presented,one at a time, with a dry chrome bead, followed by the presentationof a dry white bead to determine whether the chicks’ response wasspecific to the bead or a general inhibition of pecking.

With that procedure, it was possible to prove that bothdemonstrator and observer chicks avoided pecking at the drychrome bead up to 24 h after the observer chick had seen itsdemonstrator pecking a similar bead coated in a bitter-tastingsubstance. In contrast, chicks continued to peck at the dry chromebead if, during the training phase, the demonstrator had peckeda similar bead that was coated in water and that had not elicited anydisgust reaction. This social learning occurred specifically duringtraining.

The formation of a memory for the PAL task occurs over thecourse of several hours, with a range of biochemical, physiologicaland morphological changes (Rose 2000) associated with differentmemory phases (e.g. short-term, intermediate-term and long-termmemory, see Gibbs et al. 2003) leading to a permanent memory.Localized brain circuits, such as those within the mesopallium andmedial striatum, seem to be involved in these memory phases (e.g.see Freeman & Rose 1999) with the resulting activity lateralized ina time-dependent fashion (e.g. Andrew 1999; Rose & Stewart 1999;Rickard & Gibbs 2003a, b). A flow of memory has been hypothe-sized, with a transfer of memory from the left mesopallium throughto the right mesopallium and then later to the left and right medialstriata or to the basal ganglia (e.g. Patterson et al. 1990). Bilateral orleft, but not right, pretraining mesopallium lesions interfere withthe acquisition of this task (Patterson et al. 1990) consistent withbiochemical evidence showing that the memory trace appears toconsolidate first in the left mesopallium and then in the right (Sandi

et al. 1993). Other authors have hypothesized changes in linkagesbetween distributed memory fragments, with different informa-tional content in the two hemispheres, rather than a flow ofmemory with trace transfer over time (Andrew 1999). According tothese models, successive phases of memory formation would bedetermined by changes in the left hemisphere trace, responsible forthe performance at test, possibly with moments of integration withthe contents of the right hemisphere trace (Gibbs et al. 2003).Existing evidence thus seems to indicate that the memory for thePAL task (in its standard nonsocial version) forms in the lefthemisphere (Gibbs et al. 2003).

The predominant role of the left hemisphere in memoryformation for the PAL task, in its standard nonsocial version at least,may be because of the importance of the left hemisphere incontrolling motor responses towards objects (Andrew et al. 2000).In addition, the left hemisphere performs a role in the discrimi-nation of local, specific cues associated with a target (such as thecolour of the bitter-tasting bead; see Vallortigara et al. 1996;Tommasi & Vallortigara 2001). However, the social-learning taskused by Johnston et al. (1998) was different, with respect to thestandard PAL task, in that it involved the elaboration of social cues.In chicks, as well as other animals, the right hemisphere appears tobe dominant for various aspects of social cognition, including therecognition of individual conspecifics (Vallortigara & Andrew 1991,1994; Vallortigara 1992; Deng & Rogers 2002; Andrew et al. 2004)and gaze processing (Rosa Salva et al. 2007). This differential use ofbrain hemispheres is due to a structural lateralization in the chick’sbrain that is triggered by exposure of the embryo in the egg to light(Rogers & Sink 1988). During a critical period (Rogers 2008)exposure to light produces an asymmetrical stimulation of the twoeyes and the consequent lateralization of projections of the visualpathways to the forebrain (Rogers & Bolden 1991; Rogers & Deng1999; Koshiba et al. 2003).

Nevertheless, some behaviours, including social behaviourssuch as preferences to approach a social companion are known tobe lateralized also in dark-incubated chicks (Deng & Rogers 2002;Andrew et al. 2004). An intriguing hypothesis in this regard is thatsome forms of lateralization involving the social domain mayemerge independently from environmental cues such as exposureof the embryo to light. This in turn could result from within-pop-ulation consistency in the strength and direction of lateralizationbeing extremely relevant in the social domain (Vallortigara &Rogers 2005); thus mechanisms could have evolved ensuring thatall individuals are similarly lateralized, despite changes in envi-ronmental conditions.

Thus, one issue to arise is that regarding the lateralizationpattern expected for this social-learning version of the PAL task,which requires both elaboration of behavioural cues from thedemonstrator (probably a right-hemisphere process) associatedwith the bead’s visual cues (a left-hemisphere process) and alsoeffective control of the pecking response (again a left-hemisphereprocess). Our main aims were to confirm that 2-day-old domesticchicks are capable of the social learning of pecking avoidance(Johnston et al. 1998), and to investigate the presence and directionof lateralization effects in observer chicks for the social version ofthe PAL task.

GENERAL METHODS

Subjects

We used domestic chicks of the ‘Hybro’ strain. For experiment 1chicks were obtained weekly from a local commercial hatchery(Agricola Berica, Montegalda (VI), Italy) when they were only a fewhours old. These chicks hatched from eggs incubated in the dark, so

O.R. Salva et al. / Animal Behaviour 78 (2009) 847–856848


they may not have been lateralized for some visual tasks (see theIntroduction for a discussion of visual lateralization in light- anddark-incubated chicks). For experiments 2 and 3 fertilized eggswere obtained weekly from the same commercial hatchery on the14th day of incubation. On Days 14–17 of incubation eggs were keptin an incubator (MG 70/100 Rurale). On the 17th day of incubationeggs were placed in a hatchery (MG 100). During the last 5 days ofincubation (Days 17–21) eggs were exposed to light (15 W,230–240 V), to lateralize the embryos’ visual pathways (see above).

Rearing Conditions

On arrival, chicks were immediately sexed and housed in same-sex pairs in standard metal homecages (28 � 32 cm and 40 cmhigh). Chicks were maintained at a controlled temperature (ca. 28–31�C) and humidity (ca. 70%), with food (chicken starter) scatteredliberally over the cage floor and with water available ad libitum. Thecages were lit (24 h/day) by 36 W fluorescent lamps (placed 15 cmabove the cages).

Each pair of chicks was separated by a wire partition16 � 28 cm), dividing the cage in half (28 � 16 cm � 40 cm high). Inthis way, each chick could see, hear and have some physical contactwith its cagemate, but was not able to peck at the bead stimulusused for training and testing when it was presented to its cagemate.Only one chick in each pair was trained with the red bead (seebelow). The position of the trained chick in each pair (either on thefar or the near side of the wire mesh) was randomized.

We allowed each pair of chicks to habituate to their environ-ment for about 24 h (until the morning of Day 2) before testingthem.

Apparatus and Procedure

The experiment took place in the same cages and conditions asdescribed for the rearing period. The behaviour of the subjects wasvideo recorded with a handheld video camera. To train and test thechicks, we used the procedure described by Johnston et al. (1998).Experiments 2 and 3 differed from the general procedure describedbelow only in that chicks were given eye patches immediately after(experiment 2) or before (experiment 3) the training phase. Foodcrumbs were scattered on the floor of the cages as per Rose (2000);however, young chicks do not need to feed on external sources untilapproximately the 3rd day of life, living off internal (yolk sac) storesin the meantime (Entenmann et al. 1940).

Pretraining and Training

To evaluate the propensity for each chick to peck, chicks werepresented three times with a dry white bead (5 mm diameter)attached to a white-painted metal rod (1 mm diameter) for 10 s at5 min intervals. Only pairs of chicks in which both individuals hadpecked the white bead on at least two of the three presentationsunderwent the training procedure (see exclusion criteria listedbelow).

Five minutes later half of the chicks (one from each pair) werepresented for 30 s with a similarly sized bead, this time red,attached to an unpainted metal rod (1 mm diameter). The red beadwas either dipped in a solution of 100% MeA (MeA-chicks) or wasdry (Dry-chicks).

The chicks that were trained with the red bead are referred to asthe demonstrator chicks (Dem-chicks), whereas the second chick ofeach pair is referred to as the observer chick (Obs-chick), and wasprevented from pecking the red bead during training by the pres-ence of the wire partition. Consequently, Obs-chicks could see andsmell the training stimulus as well as see and hear their Dem-chick

interacting with the training bead. To attract the attention of bothchicks towards the training bead, we gently tapped the front of thecage prior to inserting the red training bead into the Dem-chick’scompartment. Moreover, training started only after ensuring thatthe Obs-chick was seen to be looking towards the Dem-chick’ssection of the cage.

Test Procedure

The test phase started 30 min after training. The Dem-chick ofeach pair was presented for 10 s with the dry red bead attached toa metal rod. The observer was then tested using the same proce-dure. Then, following a 5 min delay, Dem- and Obs-chicks wereindividually presented (in random order) with a dry white bead ona white rod for 10 s. We recorded (1) the number of pecks at the redbead and (2) the number of pecks at the white bead during testing.

Data Analysis

Subjects were categorized as avoiders if at test they avoided thered bead but still pecked the white, and were categorized as non-avoiders if they pecked both the red and the white bead at test.Chicks that did not peck at the white bead at test were not includedin these categories or in the subsequent analysis involving thesecategories. To compare the number of avoiders and nonavoidersbetween experimental groups (MeA-chicks versus Dry-chicks;Dem-chicks versus Obs-chicks; LE-Obs-chicks versus RE-Obs-chicks for experiments 2 and 3 only; see below for details of LE- andRE-chicks) we used the chi-square test of independence.

In addition, a discrimination ratio was calculated for thenumber of pecks at the red and white beads at test using theformula: [number of pecks at the white bead]/[number of pecksat the white þ red bead] � 100. Discrimination ratios were ana-lysed using an ANOVA incorporating the between-subjects factorscondition (MeA- or Dry-chicks), role (Dem- or Obs-chicks), eye inuse at the moment of the test (left eye, LE or right eye, RE) andsex. The performance of LE- and RE-Obs-chicks was comparedusing t tests.

For experiments 2 and 3 further video analysis was carried outto allow a more precise scoring. A small proportion of the pecksperformed by the chicks appeared distinctively different froma behavioural point of view, that is they appeared to be aggressivein manner. We wanted to exclude these pecks from the analysissince we believed that the motivational state underlying anaggressive peck was opposite to that of an appetitive peck:aggressive pecks seemed to be made to remove the bead from thechick’s proximity. For this reason, pecks categorized (from a slow-motion analysis of the video recordings) as ‘aggressive pecks’ wereexcluded from the analysis. Two criteria were used to categorizea peck as aggressive: the presence of an extremely rapid peckingmovement, performed with the bill closed, and a subsequentimmediate fast backward motion of the head or entire body (toavoid close proximity to the bead). A typical nonaggressive peckwas a much slower movement and performed with an open bill; atthe same time the chick would remain in close proximity to thebead.

Exclusion Criteria

Pretraining failurePairs of chicks were excluded from the experiment if either

chick failed to peck the bead on at least two of the three pretrainingsessions (presentations of the white bead). Chicks excludedbecause of pretraining failure underwent part of the pretraining,

O.R. Salva et al. / Animal Behaviour 78 (2009) 847–856 849


but were never trained or tested (this is a standard preselectionprocedure for the PAL task).

Training failureWe excluded from the experiment those pairs whose Dem-chick

failed to peck the training stimulus or (for MeA-chicks) to show anydisgust reaction after pecking the MeA-coated bead at training(disgust reactions were defined as the presence of at least one headshake and/or bill-wiping behaviour). These chicks underwent thepretraining and training phases, but they were never tested.‘Excluded’ chicks (pretraining and training) did not undergo thecomplete experimental procedure, because some of the standardprerequisites (e.g. a propensity to peck at the beads used or theappropriate behaviour of the Dem-chick) were missing.

Failure to peck beads at testAmong the chicks that completed all experimental phases, the

number of chicks available for data analysis was different for thetwo dependent variables (discrimination ratio and number ofavoiders). In regard to the discrimination ratios, individual chicksthat did not peck either of the two beads at test were automaticallyexcluded from the sample because of the way the dependentvariable was calculated (see above). In regard to the number ofavoiders and nonavoiders, subjects that did not peck at the whitebead could not be included in either of the two categories.

Ethical Note

The experimental procedures were licensed by the Ministerodella Salute, Dipartimento Alimenti Nutrizione e Sanita PubblicaVeterinaria. Immediately after the behavioural test, chicks werecollectively caged in groups of five with food and water available adlibitum, and on the same day, a few hours later, all of the chickswere donated to local farmers providing the chicks with free-rangerearing conditions.

EXPERIMENT 1

This initial experiment was carried out to determine whetherwe could replicate the results of Johnston et al. (1998), specificallywith regard to their interpretation of a social-learning element(direct observational learning) in passive avoidance learning.

Subjects

Subjects were 98 (48 male and 50 female, see the Appendix)dark-incubated domestic chicks. They came from an initial group of132 chicks. Of these, 34 did not complete the experiment or wereexcluded from the analysis because they did not peck at beadsduring the test (see exclusion criteria listed above): (1) 18 chicksdid not complete the experiment because of pretraining failure; (2)two chicks did not complete the experiment because of trainingfailure; and (3) 14 chicks did not peck either of the two beads attest. The number of chicks available for each dependent variable inthe various experimental groups is shown in the Appendix.

Results

In the ANOVA run on discrimination ratios, the only significanteffect was that of experimental condition (F1,61 ¼ 9.791, P ¼ 0.003),with both Dem-MeA-chicks and Obs-MeA-chicks showinga stronger avoidance of the red bead than Dry-chicks (Fig. 1).

No other effects and interactions were significant: role(F1,61 ¼ 2.680, P ¼ 0.107), sex (F1,61 ¼ 0.916, P ¼ 0.342), con-dition*role (F1,61 ¼1.051, P ¼ 0.309), condition*sex (F1,61 ¼ 0.290,

P ¼ 0.592), role*sex (F1,61 ¼ 0.387, P ¼ 0.536), condition*role*sex(F1,61 ¼ 0.424, P ¼ 0.517).

Because of technical problems in video recording these datawere not available for the whole sample. Therefore, the above-mentioned ANOVA was run on a sample of 69 subjects (19 MeA-chicks and 50 Dry-chicks, equally distributed between the two rolesand sexes: see the Appendix).

In the chi-square test of independence run on the number ofavoiders and nonavoiders there was a significant effect of thetraining condition (Dry or MeA) for both Dem-chicks (c1

2 ¼ 14.108,P < 0.000; MeA: 19 avoiders and five nonavoiders; Dry: six avoidersand 18 nonavoiders) and Obs-chicks (c1

2 ¼ 7.487, P ¼ 0.006; MeA:11 avoiders and 15 nonavoiders; Dry: two avoiders and 22 non-avoiders). However, there was a significant difference in thedistribution of avoiders and nonavoiders between Dem- and Obs-MeA-chicks (c1

2 ¼ 7.065, P ¼ 0.008). In contrast, this difference wasnot significant for Dry-chicks (c1

2 ¼ 2.400, P ¼ 0.121).This pattern of results meant that a higher number of avoiders

was present for Dem- and Obs-MeA-chicks than for Dem- and Obs-Dry-chicks (the opposite was true for nonavoiders). Nevertheless,the number of avoiders was higher in Dem-MeA-chicks than inObs-MeA-chicks (again, the opposite was true for nonavoiders).

Discussion

The results of this first experiment confirmed those obtained byJohnston et al. (1998): both observers and demonstrators learnt toavoid the bitter-tasting red bead and/or preferred to peck the whitebead with respect to the red one, 30 min after an interaction (director indirect) with the red bead. This form of social learning thusappears robust across the species: our chicks were of a differentstrain to that used by Johnston et al. (1998). The observer chickstherefore appear to avoid pecking the red bead merely by obser-vation of their demonstrator: the index that represents the level ofpreference for pecking the white over the red bead (discriminationratio) did not show a significant difference between observers and

100

*

Dis

crim

inat

ion

rat

io

80

60

40

20

0MeA Dry

Figure 1. Experiment 1. dark columns: Demonstrator chicks; light columns: Observerchicks. MeA group: pairs of chicks in which the demonstrator was trained witha bitter-tasting red bead dipped in MeA; Dry group: pairs in which the demonstratorwas trained with a neutral dry red bead. Bars show the mean value of the discrimi-nation ratio (number of pecks at white bead/(number of pecks at white þ red beads)).Vertical lines represent SEMs. *P < 0.05.



demonstrators, and the value of this index was higher for the MeA-chicks than for Dry-chicks. A more pronounced avoidance of the redbead was evident for observers in Johnston et al.’s (1998) experi-ment than in the present experiment. Nevertheless, this should notbe a matter of concern, since the overall pattern of results showsa remarkable correspondence between our study and that ofJohnston et al. (1998). Moreover, various minor aspects of theprocedure differed between our and Johnston et al.’s study (e.g.familiarization time spent in the cages before testing and breed ofchicks) and could have influenced the level of red bead avoidance.

For both observers and demonstrators the number of avoiders(chicks that avoided pecking the red bead after the training phase)was higher for MeA-chicks than for Dry-chicks. However, whereasfor demonstrators the relative number of avoiders in the MeA-chicks was larger than the relative number of nonavoiders this wasnot true for the observers. Moreover, in terms of the overall numberof avoiders and nonavoiders the performance of Obs- and Dem-MeA-chicks was significantly different. Nevertheless, any differencebetween observers and demonstrators should not be of too greata concern, even though such a difference was not present inJohnston et al.’s (1998) study, since the MeA- and Dry-Obs-chicksstill showed a significant difference in performance.

These results, therefore, gave us the basis with which to testfurther the social-learning abilities of chicks, with emphasis on therole of brain lateralization on observer performance.

EXPERIMENT 2

Having determined that our chick model provides evidence ofsocial learning for the PAL task, we wished to see whether there isan effect of brain lateralization on such learning in the observerchicks. Brain lateralization can affect various aspects of socialcognition (see Introduction). We suggest that since chicks can learnthe PAL task just by observing a conspecific, it is likely that suchlearning would be lateralized: the way in which the demonstrator(and its behaviour) is regarded would depend upon the hemisphereused which, in turn, would therefore determine the characteristicsof the demonstrator’s response selected by the observer.

Methods

SubjectsSubjects were 180 (92 male and 88 female, see the Appendix)

light-incubated domestic chicks. They came from an initial group of312 chicks. Of these, 132 did not complete the experiment or wereexcluded from the analysis because they did not peck at beadsduring the test (see exclusion criteria listed above): (1) 86 chicksdid not complete the experiment because of pretraining failure; (2)21 chicks did not complete the experiment because of trainingfailure; and (3) 25 chicks did not peck either of the two beads attest. The number of chicks available for each dependent variable inthe various experimental groups is shown in the Appendix. As hasbeen suggested by an anonymous reviewer, the increased exclusionrate with respect to experiment 1 may have been caused by thedifferent (with respect to experiment 1) incubation condition thechicks experienced in this experiment (i.e. exposure to light duringincubation).

Monocular occlusion procedureThe experimental apparatus and the general procedure were the

same as described in the General Methods, with the exception thatObs-chicks were tested with vision limited to one eye (monocularvision condition), although they underwent training binocularly, asin the previous experiment. Immediately after training chicks weregiven temporary eye patches (by using a removable, sticky tape), to

obtain two monocular vision conditions: observers that were testedwith vision limited to the left eye (right-eye occluded) are referred toas LE-chicks, whereas the remaining observers, which were testedwith vision limited to the right eye (left-eye occluded), are referredto as RE-chicks. The procedure was relatively minor, requiringhandling for only a few seconds; it consisted of gently placingremovable tape over one eye (the tape was cup-shaped and did notprevent the normal movements of the chick’s eyelid). After receivingthe patches, all chicks were allowed 30 min to become accustomedto the new monocular condition (the same 30 min delay timebetween training and testing was also present in the first experi-ment, as part of the learning procedure). This procedure was aimedat investigating the presence of lateralization in the social-learningtask (in birds with laterally placed eyes a stimulus monitored by oneeye is processed mainly in the contralateral hemisphere; Andrew1991; Rogers 1995; 1996; Gunturkun 1997; Deng & Rogers 1998a,1998b). Dem-chicks always underwent pretraining, training andtesting in the binocular condition (exactly as in experiment 1).

Results

In the ANOVA run on discrimination ratios for the overall sample(including both observers and demonstrators), significant maineffects were obtained for all the factors considered (condition:F1,164 ¼ 30.425, P < 0.001; role: F1,164 ¼ 6.922, P ¼ 0.009; eye in use:F1,164 ¼ 5.206, P ¼ 0.024; sex: F1,164 ¼ 5.301, P ¼ 0.023). Moreover,significant interactions emerged for role*sex (F1,164 ¼ 5.095,P ¼ 0.025) and condition*eye in use*sex (F1,164 ¼ 8.567, P ¼ 0.004).

No other interaction was significant (condition*role: F1,164¼ 0.017,P¼ 0.895; condition*eye in use: F1,164¼ 0.160, P¼ 0.690; role*eye in use: F1,164¼ 0.920, P¼ 0.339; condition*role*eye in use:F1,164¼ 0.728, P¼ 0.395; condition*sex: F1,164¼ 0.295, P¼ 0.588;condition*role*sex: F1,164¼ 0.680, P ¼ 0.411; eye in use*sex:F1,164¼ 1.578, P ¼ 0.211; role*eye in use*sex: F1,164¼ 0.028, P¼ 0.868;condition*role*eye in use*sex: F1,164¼ 1.327, P¼ 0.251).

Overall, the MeA-chicks tested in this experiment showeda stronger avoidance of the red bead than Dry-chicks (as demon-strated by the significant main effect of condition). However, thepresence of a significant main effect of role made it possible to restrictour analysis to Obs-chicks, which are our main focus of interest. InANOVA run on discrimination ratios for Obs-chicks, significant maineffects were obtained for all the factors considered (condition:F1,82 ¼ 13.911, P < 0.000; eye in use: F1,82 ¼ 5.039, P¼ 0.027; sex:F1,82 ¼ 9.975, P¼ 0.002). Moreover, a significant interaction of con-dition*eye in use*sex (F1,82¼ 7.983, P¼ 0.006) emerged from thisanalysis. All other interactions were not significant (condition*eye inuse: F1,82¼ 0.753, P¼ 0.388; condition*sex :F1,82¼ 0.038, P¼ 0.846;eye in use*sex :F1,82¼ 0.970, P¼ 0.327).

The presence of a main effect of the factor condition shows that,overall, Obs-MeA-chicks showed a stronger avoidance of the redbead than did Obs-Dry-chicks (Fig. 2).

The presence of the significant interaction condition*eye inuse*sex gave us the possibility to investigate the presence oflateralization effects in MeA-Ob-chicks by comparing the perfor-mance of LE- and RE-chicks independently for males and females.Only the performance of males was significantly affected by the eyein use (t28 ¼ 2.316, P ¼ 0.028) with LE-Obs male chicks showinga stronger avoidance of the red bead than RE-Obs males, whereasfor females no significant difference could be detected (t22 ¼ 0.990,P ¼ 0.333; Fig. 3). Therefore, the effect of lateralization found in theanalysis seems to be caused mainly by male observers.


2 ¼ 16.726,P < 0.001; MeA: 35 avoiders and 12 nonavoiders; Dry: 11 avoiders



and 26 nonavoiders) and Obs-chicks (c12 ¼ 11.177, P ¼ 0.001; MeA:

27 avoiders and 22 nonavoiders; Dry: six avoiders and 27 non-avoiders). However, there was a significant difference in thedistribution of avoiders and nonavoiders between Dem- and Obs-MeA-chicks (c1

2 ¼ 3.933, P ¼ 0.047). This difference was notsignificant for Dry-chicks, however (c1

2 ¼ 1.256, P ¼ 0.261).This pattern of results is consistent with that obtained in

experiment 1: a higher number of avoiders was present for Dem-and Obs-MeA-chicks than for Dem- and Obs-Dry-chicks (theopposite was true for nonavoiders). Nevertheless, the number ofavoiders was higher in Dem-MeA-chicks than in Obs-MeA-chicks(again, the opposite was true for nonavoiders).

The distribution of avoiders and nonavoiders was not signifi-cantly different between LE- and RE-Obs-MeA-chicks (c1

2 ¼ 0.928,P ¼ 0.336; LE: 16 avoiders and 10 nonavoiders; RE: 11 avoiders and12 nonavoiders) or between LE- and RE-Obs-Dry-chicks (c1

2 ¼ 0.007,P ¼ 0.935; LE: three avoiders and 14 nonavoiders; RE: three avoidersand 13 nonavoiders).

Discussion

The results of the second experiment confirmed what we hadalready demonstrated in the first, that observer chicks are able tolearn the passive avoidance task by social observation.

Moreover, we obtained evidence of the differential involvementof the two cerebral hemispheres in this social-learning task. Thiseffect, however, seemed to be confined to males. A significantdifference between LE- and RE-Obs-chicks in the values of theindex that represents the level of preference for pecking the whiteover the red bead (discrimination ratio) was present only in males(whose performance was higher when tested with their left eye(right hemisphere) in use). The presence of this lateralization effectin only male subjects is not completely surprising; many behav-ioural and structural lateralization effects are known to be limitedto males, or at least more pronounced in males (see Rogers 1995).

As pointed out by an anonymous referee, in this context theexplanation of lateralization effects in terms of ‘recall’ should betreated with caution, since other factors may have influenced theresponses to the two beads of the different groups of chicks. In thisregard, evidence suggests a differential encoding of colour cues inthe two hemispheres (the left hemisphere would be preferentiallyin charge of encoding such information in the standard PAL task,Patterson & Rose 1992). The effects obtained could thus beexplained as differences between the two hemispheres both interms of recall and/or in terms of encoding.

The superior performance of LE-chicks (processing the stimulusmainly with their right hemisphere) is not consistent with theliterature concerning lateralization in PAL. In fact, a major involve-ment of the left hemisphere is usually reported for this task (seeIntroduction). However, the task that observers had to perform wassomewhat different with respect to the standard PAL task (the socialcomponent was predominant in the task: learning could be ach-ieved only by observing a familiar conspecific that interacted withthe relevant stimulus). The right hemisphere is known to playa dominant role in social cognition in chicks (see Introduction). Afirst possible interpretation of this evidence is that during training,observers tended to look at their demonstrator preferentially withtheir left eye (possibly because of a preference for maintainingsocial or familiar objects in the left visual hemifield in chicks;Andrew 1991; Vallortigara & Andrew 1991; Vallortigara 1992;Vallortigara et al. 2001). Owing to this eye use preference, the chickswould have consequently stored information in the right hemi-sphere, and subsequently had a poorer performance in retrievingthe memory trace when tested with their right eye (left hemi-sphere). An alternative explanation is that a spontaneous tendencyto process relevant information in the right hemisphere (when bothhemispheres have access to visual information during training, as inexperiment 2) could exist independently from eye use.

Nevertheless, we cannot conclude that this lateralization effectwas due to a spontaneous preference for looking at conspecificswith the left eye or for processing relevant information in the righthemisphere; instead it may have been caused by an intrinsicallypoorer ability of the left hemisphere to process the informationrequired for achieving social learning in this task. A possible way todisentangle this issue at least in part would be to try to force the useof the left hemisphere when learning the task, which is exactlywhat we did in experiment 3.

100

Dis

crim

inat

ion

rat

io

*

80

60

40

20

0MeA Dry

Figure 2. Experiment 2. MeA group: pairs of chicks in which the demonstrator wastrained with a bitter-tasting red bead dipped in MeA; Dry group: pairs in which thedemonstrator was trained with a neutral dry red bead. Bars show the mean value ofthe discrimination ratio (number of pecks at white bead/(number of pecks atwhite þ red beads)) for Observer chicks. Vertical lines represent SEMs. *P < 0.05.

100

*

Dis

crim

inat

ion

rat

io 80

60

40

20

0M

LE RE

F M F

Figure 3. Experiment 2: observers. Dark columns: MeA-observers (whose demon-strator was trained with a bitter-tasting red bead dipped in MeA) light columns:Dry-observers (whose demonstrator was trained with a neutral dry red bead). M:male; F: female. LE group: pairs of chicks tested with only the left eye in use; REgroup: pairs tested with only the right eye in use. Bars show the mean value of thediscrimination ratio (number of pecks at white bead/(number of pecks atwhite þ red beads)) for Observer chicks. Vertical lines represent SEMs. *P < 0.05.



EXPERIMENT 3

In experiment 2 a superior performance of the right hemisphereemerged for social learning of pecking avoidance. As pointed outabove, we wished to disentangle the role of the hemispheres inobservational learning, investigating whether the right-eye system(left hemisphere) has the potential to learn as well as the right onefrom the behaviour of conspecifics in the social version of the PALtask. To do this, in the present experiment, we forced the observerchick to learn the task with either the left or the right hemisphere,by performing both training and test phases in a monocular visioncondition.

Methods

SubjectsSubjects were 134 (70 male and 64 female, see the Appendix)

light-incubated domestic chicks. They came from an initial groupof 228 chicks. Of these, 94 did not finish the experiment or wereexcluded from analysis because they did not peck at beads duringthe test (see exclusion criteria listed above): (1) 64 chicks did notcomplete the experiment because of pretraining failure; (2) 12chicks did not complete the experiment because of trainingfailure; and (3) 18 chicks did not peck either of the two beads attest. The number of chicks available for each dependent variablein the various experimental groups is shown in the Appendix.Again, the increased exclusion rate with respect to experiment 1may have been the result of the incubation condition (see above).

Monocular occlusion procedureThe experimental apparatus and the general procedure were the

same as that used in experiment 2 (see also General Methods), withthe exception that Obs-chicks were not only tested but also pre-trained and trained with vision limited to one eye (monocular visioncondition). The monocular occlusion procedure was performed30 min before the beginning of pretraining, to give the animals timeto become accustomed to the new monocular condition. Each chickwas both trained and tested with the same eye in use. This procedurewas aimed at investigating the presence of lateralization in Obs-chicks that, during training, were forced to process the relevantinformation in one hemisphere only and were then tested with thesame hemisphere in use. In this way we could check whether thelateralization effects that emerged in experiment 2 were due toa spontaneous preference for the use of the dominant right hemi-sphere during training or to the poorer performance of the otherhemisphere when processing the appropriate information. Dem-chick, were always pretrained, trained and tested in the binocularvision condition (as in experiments 1 and 2).

Results

In the ANOVA run on discrimination ratios for the overallsample (including both observes and demonstrators), the onlysignificant factor was that of role (F1,118 ¼ 14.074, P < 0.001), withDem-chicks showing a stronger avoidance of the red bead thanObs-chicks. No other effects or first-level interactions were signif-icant: condition (F1,118 ¼ 2.165, P ¼ 0.144), sex (F1,118 ¼ 0.914,P ¼ 0.341), eye in use (F1,118 ¼ 0.050, P ¼ 0.824), condition*role(F1,118 ¼ 3.026, P ¼ 0.085), condition*sex (F1,118 ¼ 0.672, P ¼ 0.414),role*sex (F1,118 ¼ 3.723, P ¼ 0.056), condition*role*sex(F1,118 ¼ 1.159, P ¼ 0.284), condition*eye in use (F1,118 ¼ 0.378,P ¼ 0.540), role*eye in use (F1,118 ¼ 0.360, P ¼ 0.550), con-dition*role*eye in use (F1,118 ¼ 0.432, P ¼ 0.512), sex*eye in use(F1,118 ¼ 3.614, P ¼ 0.060), condition*sex*eye in use (F1,118 ¼ 0.192,P ¼ 0.662), role*sex*eye in use (F1,118 ¼ 1.394, P ¼ 0.240).

However, the presence of a significant main effect of role made itpossible to restrict our analysis to Obs-chicks, which are the mainfocus of our interest. In the ANOVA run on discrimination ratios forObs-chicks, the only source of significance was the sex*eye in useinteraction (F1,63 ¼ 4.406, P ¼ 0.040). Overall (regardless ofwhether they were MeA- or Dry-chicks), male and female chicksseemed to differ in their responses to the red bead according to theeye in use (Fig. 4). All other effects and interactions were notsignificant: condition (F1,63 ¼ 0.033, P ¼ 0.856), sex (F1,63 ¼ 0.439,P ¼ 0.510), eye in use (F1,63 ¼ 0.066, P ¼ 0.798), condition*sex(F1,63 ¼ 1.668, P ¼ 0.201), condition*eye in use (F1,63 ¼ 0.001,P ¼ 0.977), condition*sex*eye in use (F1,63 ¼ 1.455, P ¼ 0.232).


2 ¼ 20.318,P < 0.001; MeA: 21 avoiders and four nonavoiders; Dry: eightavoiders and 25 nonavoiders) and Obs-chicks (c1

2 ¼ 10.918,P ¼ 0.001; MeA: 13 avoiders and 12 nonavoiders; Dry: four avoidersand 29 nonavoiders). However, there was a significant difference inthe distribution of avoiders and nonavoiders between Dem- andObs-MeA-chicks (c1

2 ¼ 5.882, P ¼ 0.015). This difference was notsignificant for Dry-chicks (c1

2 ¼ 1.630, P ¼ 0.202). This pattern ofresults is the same as obtained in experiments 1 and 2: a highernumber of avoiders was present for Dem- and Obs-MeA-chicksthan for Dem- and Obs-Dry-chicks (the opposite was true fornonavoiders). Nevertheless, the number of avoiders was higher inDem-MeA-chicks than in Obs-MeA-chicks (again, the opposite wastrue for nonavoiders).

In the chi-square test of independence run to compare the twomonocular vision conditions the distribution of avoiders and non-avoiders was not significantly different between LE- and RE-Obs-MeA-chicks (c1

2 ¼ 0.371, P ¼ 0.543; LE: seven avoiders and fivenonavoiders; RE: six avoiders and seven nonavoiders) or betweenLE- and RE-Obs-Dry-chicks (c1

2 ¼ 0.768, P ¼ 0.381; LE threeavoiders and 15 nonavoiders; RE: one avoider and 14 nonavoiders).

Discussion

The results of the third experiment confirmed what had alreadybeen demonstrated in experiments 1 and 2 in regard to thepossibility for observer chicks to learn the passive avoidance task bysocial observation, as proved by the analysis of the distribution ofavoider and nonavoider chicks. In this third experiment it was notpossible to obtain any evidence of learning of the task from the

100

Dis

crim

inat

ion

rat

io

80

60

40

20

0M F

Figure 4. Experiment 3. Light columns: LE-chicks (tested with only their left eye inuse); dark columns: RE-chicks (tested with only their right eye in use). M: male; F:female. Bars show the mean value of the discrimination ratio (number of pecks atwhite bead/(number of pecks at white þ red beads)) for Observer chicks. Vertical linesrepresent SEMs.



analysis of the discrimination ratios. A likely explanation could bethat the presence of the eye patch from the very beginning of theexperiment could be a source of additional stress. This in turn couldaffect the behaviour of the observer chick in the first place (but alsothat of the demonstrator, because of its interaction with theobserver), and reduce its performance, according to the possiblymore sensitive dependent variable, the discrimination ratio.Nevertheless, this should not be of too great a concern, if weconsider that, as already noted, the distribution of avoiders andnonavoiders in experiment 3 was the same as found in all previousexperiments.

Moreover, in the present experiment it was not possible tofind the same pattern of lateralization effects that was present inexperiment 2 (i.e. a selectively lower performance in male RE-Obs-MeA-chicks). This result suggests that the lateralizationeffect obtained in experiment 2 could have been caused by thespontaneous preference for looking at conspecifics with the lefteye or for processing and encoding the relevant information forthe task with the right hemisphere (see above), and not to anintrinsically poorer ability of the left hemisphere to process theinformation required for achieving social learning in this task.Nevertheless, more direct evidence is required to clarify this issueand inferences from nonsignificant effects should be treated withcaution. In the present experiment, RE-Obs-chicks were forced touse their left hemisphere to encode the relevant informationdisplayed at training by their demonstrator. Nevertheless, theirlearning performance at test was not significantly lower than thatof LE-chicks (which had to use their right hemisphere attraining). In fact, the only significant effect in the analysis run onobservers’ discrimination ratios was a sex*eye in use interaction,as overall (regardless of whether they were MeA- or Dry-chicks),male LE-chicks seemed to show a stronger avoidance of the redbead than male RE-Obs-chicks, whereas the opposite may havebeen true for females. The factor condition was not involved inthis interaction. This means that, even if male LE-Obs-chicksavoided the red bead more than male RE-Obs-chicks, this is truefor both MeA- and Dry-Obs-chicks (which should not avoid thered bead). The same argument is also valid for female LE- and RE-Obs-chicks. It is therefore possible to conclude that this differencebetween LE- and RE-Obs-chicks in the level of avoidance of thered bead does not constitute a difference in the learning perfor-mance between the two groups.

GENERAL DISCUSSION

Our results confirmed that 2-day-old domestic chicks are able tolearn to inhibit pecking at a potentially poisonous item, and did sojust by observing a demonstrator’s behaviour. This result isconsistent with evidence obtained by Johnston et al. (1998), andalso with other evidence of social transmission of food preferencesin gallinaceous birds (see Introduction). Nevertheless, otherexperiments with adult subjects seemed to question whether sociallearning could affect avoidance of items that elicit a disgust reactionin conspecifics (Sherwin et al. 2002). It was thus important toconfirm that this ability is present, at least in young animals (seeIntroduction).

We also obtained evidence for a differential involvement ofthe two cerebral hemispheres in this social-learning task. Thisresult is relevant from various points of view. First, it has beentheorized that complex tasks require the interaction of bothhemispheres and cannot be performed with only one hemisphere(Prior & Wilzeck 2008). In the present study, we obtainedevidence of social learning in chicks tested and/or trained inmonocular visual conditions (e.g. the number of avoiders andnonavoiders differed between MeA- and Dry-chicks in all

observers tested, monocularly, in experiment 2 and both trainedand tested monocularly in experiment 3). Nevertheless, thesocial-learning task used here does not seem to be less complexthan the selective feeding task used by Prior & Wilzeck (2008):chicks had to associate different behavioural cues from theirdemonstrator with the details of the appearance (colour) of thered bead that was present within the cage at that moment. Thisseems to imply that features other than task complexity coulddetermine which tasks require the interaction of both hemi-spheres. Nevertheless, in Prior & Wilzeck’s (2008) study, theyounger chicks (1–6 days old, which are the group morecomparable to the sample of 2-day-old chicks tested here)showed a certain degree of selectivity even when testedmonocularly. Age could thus be another crucial factor to beexplored by further studies.

Another difference from Prior & Wilzeck’s (2008) study is in thedirection of the lateralization effect. Chicks performed better intheir selective feeding task when using the left hemisphere (in linewith results obtained in the literature for other pecking tasks, e.g.pebble floor task, Rogers 1997), whereas in the present study chicksshowed more avoidance of the bitter-tasting red bead over theneutral white bead when using the right hemisphere (results of thediscrimination ratios). Nevertheless, the comparison betweenthese two results is complicated not only by the presence ofobvious differences in the tasks, but also by the fact that Prior &Wilzeck tested only female subjects, whereas the lateralizationeffect obtained in the present study was confined to males. It couldthus be interesting to investigate the same issue studied by Prior &Wilzeck also in males.

The superior performance in the social version of the PAL taskobtained in male chicks using their right hemisphere compared toothers using their left hemisphere is in conflict with the generalconsensus that the memory for the standard nonsocial version ofthis task forms in the left hemisphere (see Introduction). Thisapparent inconsistency can be accounted for by consideringdifferences in the social task the observers had to perform inrespect to the standard PAL task. It is thus possible to hypothesizethat the dominance of the right hemisphere found in the presentstudy may result from its involvement in different aspects of socialcognition (see Introduction).

Moreover, a comparison of the results of experiments 2 and 3suggests that the lateralization pattern observed is due toa spontaneous tendency to use the left-eye system (right hemi-sphere) to observe social partners or to process informationabout social partners (see above), rather than a poorer ability ofthe left hemisphere to encode the information required for thesocial-learning task. As suggested by an anonymous referee,a possible way to disentangle this issue would be to check for thepresence of a preference for observing demonstrators with theleft eye during training in male MeA observers in experiment 2.No evidence of such a preference was found in this subsample ofchicks (Wilcoxon signed-ranks test on the relative amount oftime spent viewing the demonstrator chick with left or right eye:Z ¼ �1.535, N ¼ 35, P ¼ 0.125). Thus, it seems that our resultscould be more easily explained in terms of a spontaneoustendency to encode relevant information in the right hemisphereregardless of eye use.

The interpretation of our results in terms of spontaneousprocessing in one hemisphere is appealing for various reasons.The specific components of the procedure (including thedemonstrator chick’s behaviour, and the colour, shape and size ofthe bead, together with various aspects of movement associatedwith the presentation) may be differentially processed by the twobrain hemispheres. Evidence from other studies suggests that thismay indeed be the case. The right hemisphere encodes



information required for social recognition (Deng & Rogers 2002),whereas information maintained in the left hemisphere is relatedto the properties of the object to be manipulated (Gibbs et al.2003). That information is preferentially maintained according tohemisphere would also suggest that the processing of the infor-mation is also lateralized. Thus, regardless of a lack of laterali-zation of viewing propensity, it is highly likely that the two eyesare seeing different components of the task and that chickspossess a spontaneous preference for encoding specific compo-nents of this task in one of the two hemispheres (the right one),when they are free to do so (as was the case for experiment 2,where both hemispheres had access to information duringtraining, unlike in experiment 3).

Nevertheless, more direct evidence is required to clarify thisissue and inferences from nonsignificant effects should be treatedwith caution. We did not plan the video-recording procedure withthe objective of scoring observers’ behaviour, and thus technicalissues could influence the result reported here. Further researchcould be devoted to the study of preferences for eye use in the socialversion of the PAL task, to clarify this issue.

A last point concerns the time course of the lateralizationeffects. Here we obtained a dominance of the right hemispherewhen testing observers 30 min after training. Evidence in theliterature suggests that at this time point (which lies on theboundary between the first and second phase of intermediate-termmemory, ITM-A and –B), information held by the right hemispherecould be particularly important in the standard nonsocial version ofthe task (information in the right hemisphere may be necessary forthe left hemisphere during this phase, Gibbs et al. 2003; see alsoStewart et al. 1992; Daisley & Rose 2002). This could affect thelateralization pattern obtained in the present experiment. Furtherwork may clarify this point by investigating lateralization in thesocial-learning version of the PAL at other time points. Specifically,time points should be tested at which there is a left-hemisphereretrieval event in the standard nonsocial version of the PAL task(Andrew 1999).

Acknowledgments

This research is part of the project EDCBNL (Evolution andDevelopment of Cognitive, Behavioural and Neural Later-alisationd2006–2009), supported by the Commission of theEuropean Communities within the framework of the specificresearch and technological development programme ‘Integratingand strengthening the European Research Area’ (initiative ‘What itmeans to be Human’), through a financial grant to L.R. and G.V. Wethank Diana Castelli, Tania Mattarello and Fabrizio Spinedi for theirhelp with gathering data and caring for the chicks and MassimoNucci for revising the statistical analyses.

References

Allen, T. & Clarke, J. A. 2005. Social learning of food preferences by white-tailedptarmigan chicks. Animal Behaviour, 70, 305–310.

Andrew, R. J. 1991. The nature of behavioural lateralization in the chick. In: Neuraland Behavioural Plasticity: the Use of the Chick as a Model (Ed. by R. J. Andrew),pp. 536–554. Oxford: Oxford University Press.

Andrew, R. J. 1999. The differential roles of right and left sides of the brain inmemory formation. Behavioural Brain Research, 98, 289–295.

Andrew, R. J., Tommasi, L. & Ford, N. 2000. Motor control by vision and theevolution of cerebral lateralization. Brain and Language, 73, 220–235.

Andrew, R. J., Johnston, A. N., Robins, A. & Rogers, L. J. 2004. Light experience andthe development of behavioural lateralisation in chicks. II. Choice of familiarversus unfamiliar model social partner. Behavioural Brain Research, 155, 67–76.

Bartashunas, C. & Suboski, M. D. 1984. Effects of age of chick on social trans-mission of pecking preferences from hen to chicks. Developmental Psychobi-ology, 17, 121–127.

Caldwell, C. A. & Whiten, A. 2003. Scrounging facilitates social learning in commonmarmosets, Calithrix jacchus. Animal Behaviour, 65, 1085–1092.

Cherkin, A. 1969. Kinetics of memory consolidation: role of amnesic treatmentparameters. Proceedings of the National Academy of Sciences of the U.S.A., 63,1094–1101.

Daisley, J. N. & Rose, S. P. 2002. Amino acid release from the intermediate medialhyperstriatum ventrale (IMHV) of day-old chicks following a one-trial passiveavoidance task. Neurobiology of Learning & Memory, 77, 185–201.

Deng, C. & Rogers, L. J. 1998a. Bilaterally projecting neurons in the two visualpathways of chicks. Brain Research, 794, 281–290.

Deng, C. & Rogers, L. J. 1998b. Organisation of the tectorotundal and SP/IPS-rotundal projections in the chick. Journal of Comparative Neurology, 394,171–185.

Deng, C. & Rogers, L. J. 2002. Social cognition and approach in the chick:lateralization and effect of visual experience. Animal Behaviour, 63,697–706.

Entenmann, C., Lorenz, F. W. & Chaikoff, I. L. 1940. The lipid content of blood, liverand yolk sac of the newly hatched chick and the changes that occur in thesetissues during the first month of life. Journal of Biological Chemistry, 133,231–241.

Freeman, F. M. & Rose, S. P. 1999. Expression of Fos and Jun proteins followingpassive avoidance training in the day-old chick. Learning & Memory, 6,389–397.

Gajdon, G. K., Hungerbuhler, N. & Stauffacher, M. 2001. Social influence on earlyforaging of domestic chicks (Gallus gallus) in a near-to-nature procedure.Ethology, 107, 913–937.

Gibbs, M. E. & Ng, K. T. 1977. Psychobiology of memory: towards a model ofmemory formation. Biobehavioural Reviews, 1, 113–136.

Gibbs, M. E., Andrew, R. J. & Ng, K. T. 2003. Hemispheric lateralization of memorystages for discriminated avoidance learning in the chick. Behavioural BrainResearch, 139, 157–165.

Gunturkun, O. 1997. Avian visual lateralization. A review. NeuroReport, 8,3–11.

Hogan, J. A. 1973. Development of food recognition in young chicks: I. Matu-ration and nutrition. Journal of Comparative and Physiological Psychology, 83,355–366.

Johnston, A. N. B., Burne, T. H. J. & Rose, S. P. 1998. Observation learning in day oldchicks using a one-trial passive avoidance learning paradigm. Animal Behaviour,56, 1347–1353.

Koshiba, M., Nakamura, S., Deng, C. & Rogers, L. J. 2003. Light-dependentdevelopment of asymmetry in the ipsilateral and contralateral thalamofugalvisual projections of the chick. Neuroscience Letters, 336, 81–84.

Lossner, B. & Rose, S. P. 1983. Passive avoidance training increases fucokinaseactivity in right forebrain base of day-old chicks. Journal of Neurochemistry, 41,1357–1363.

McQuoid, L. M. & Galef, B. G. J. 1992. Social influences on feeding site selectionby Burmese fowl (Gallus gallus). Journal of Comparative Psychology, 106,137–141.

McQuoid, L. M. & Galef, B. G. J. 1993. Social stimuli influencing feeding behaviour ofBurmese fowl: a video analysis. Animal Behaviour, 46, 13–22.

McQuoid, L. M. & Galef, B. G. J. 1994. Effects of access to food during trainingon social learning by Burmese red junglefowl. Animal Behaviour, 48,737–739.

Moffat, C. A. & Hogan, J. A. 1992. Ontogeny of chick responses to maternal foodcalls in the Burmese red junglefowl (Gallus gallus spadiceus). Journal ofComparative Psychology, 106, 92–96.

Ng, K. T., Gibbs, M. E., Crowe, S. F., Sedman, G. L., Hua, F., Zhao, W.,O’Dowd, B., Rickard, N., Gibbs, C. L., Sykowa, E., Svoboda, J. & Jendelova, P.1991. Molecular mechanisms of memory formation. Molecular Neurobiology,5, 333–350.

Nicol, C. J. 2004. Development, direction, and damage limitation: social learning indomestic fowl. Learning & Behaviour, 32, 72–81.

Nicol, C. J. 2006. How animals learn from each other. Applied Animal BehaviourScience, 100, 58–63.

Nicol, C. J. & Pope, S. J. 1992. Effects of social learning on the acquisition ofdiscriminatory keypecking in hens. Bulletin of the Psychonomic Society, 30,293–296.

Nicol, C. J. & Pope, S. J. 1993. Food deprivation during observation reduces sociallearning in hens. Animal Behaviour, 45, 193–196.

Nicol, C. J. & Pope, S. J. 1994. Social learning in small flocks of laying hens. AnimalBehaviour, 47, 1289–1296.

Nicol, C. J. & Pope, S. J. 1996. The maternal feeding display of domestic hens issensitive to perceived chick error. Animal Behaviour, 52, 767–774.

Nicol, C. J. & Pope, S. J. 1999. The effects of demonstrator social status and priorforaging success on social learning in laying hens. Animal Behaviour, 57,163–171.

Patterson, T. A. & Rose, S. P. 1992. Memory in the chick: multiple cues, distinctbrain locations. Behavioural Neuroscience, 106, 465–470.

Patterson, T. A., Gilbert, D. B. & Rose, S. P. 1990. Pre- and post-training lesions ofthe intermediate hyperstriatum ventrale and passive avoidance learning in thechick. Experimental Brain Research, 80, 189–195.

Prior, H. & Wilzeck, C. 2008. Selective feeding in birds depends on combinedprocessing in the left and right hemisphere. Neuropsychologia, 46,233–240.

Rickard, N. S. & Gibbs, M. E. 2003a. Effects of nitric oxide inhibition on avoidancelearning in the chick are lateralized and localized. Neurobiology of Learning &Memory, 79, 252–256.



Rickard, N. S. & Gibbs, M. E. 2003b. Hemispheric dissociation of the involvement ofNOS isoforms in memory for discriminated avoidance in the chick. Learning &Memory, 10, 314–318.

Rogers, L. J. 1995. The Development of Brain and Behaviour in the Chicken. Wall-ingford: CAB International.

Rogers, L. J. 1996. Behavioral, structural and neurochemical asymmetries in theavian brain: a model system for studying visual development and processing.Neuroscience & Biobehavioral Reviews, 20, 487–503.

Rogers, L. J. 1997. Early experiential effects on laterality: research on chicks hasrelevance to other species. Laterality, 2, 199–219.

Rogers, L. J. 2008. Development and function of lateralization in the avian brain.Brain Research Bulletin, 76, 235–244.

Rogers, L. J. & Bolden, S. W. 1991. Light-dependent development and asymmetry ofvisual projections. Neuroscience Letters, 121, 63–67.

Rogers, L. J. & Deng, C. 1999. Light experience and lateralization of the two visualpathways in the chick. Behavioural Brain Research, 98, 277–287.

Rogers, L. J. & Sink, H. S. 1988. Transient asymmetry in the projections of therostral thalamus to the visual hyperstriatum of the chicken, and reversal of itsdirection by light exposure. Experimental Brain Research, 70, 378–384.

Rosa Salva, O., Regolin, L. & Vallortigara, G. 2007. Chicks discriminate human gazewith their right hemisphere. Behavioural Brain Research, 177, 15–21.

Rose, S. P. 2000. God’s organism? The chick as a model system for memory studies.Learning & Memory, 7, 1–17.

Rose, S. P. & Stewart, M. G. 1999. Cellular correlates of stages of memory formationin the chick following passive avoidance training. Behavioural Brain Research,98, 237–243.

Sandi, C., Patterson, T. A. & Rose, S. P. 1993. Visual input and lateralization of brainfunction in learning in the chick. Neuroscience, 52, 393–401.

Sherwin, C. M., Heyes, C. M. & Nicol, C. J. 2002. Social learning influences thepreferences of domestic hens for novel food. Animal Behaviour, 63, 933–942.

Stewart, M. G., Bourne, R. C. & Steele, R. J. 1992. Quantitative autoradiographicdemonstration of changes in binding to NMDA-sensitive [3H]glutamate and[3H]MK801, but not [3H]AMPA receptors in chick forebrain 30 min after passiveavoidance training. European Journal of Neuroscience, 4, 936–943.

Suboski, M. D. & Bartashunas, C. 1984. Mechanisms for social transmission ofpecking preferences to neonatal chicks. Journal of Experimental Psychology.Animal Behavior Processes, 10, 182–194.

Tommasi, L. & Vallortigara, G. 2001. Encoding of geometric and landmark infor-mation in the left and right hemispheres of the avian brain. Behavioral Neuro-science, 115, 602–613.

Vallortigara, G. 1992. Right hemisphere advantage for social recognition in thechick. Neuropsychologia, 30, 761–768.

Vallortigara, G. & Andrew, R. J. 1991. Lateralization in response by chicks to changein a model partner. Animal Behaviour, 41, 187–194.

Vallortigara, G. & Andrew, R. J. 1994. Differential involvement of right and leftcerebral hemisphere in individual recognition in the domestic chick. Behav-ioural Processes, 33, 41–58.

Vallortigara, G. & Rogers, L. J. 2005. Survival with an asymmetrical brain:advantages and disadvantages of cerebral lateralization. Behavioural and BrainSciences, 28, 575–633.

Vallortigara, G., Regolin, L., Bortolomiol, G. & Tommasi, L. 1996. Lateral asym-metries due to preferences in eye use during visual discrimination learning inchicks. Behavioral Brain Research, 74, 135–143.

Vallortigara, G., Cozzutti, C., Tommasi, L. & Rogers, L. J. 2001. How birds use theireyes: opposite left-right specialisation for the lateral and frontal visual hemi-field in the domestic chick. Current Biology, 11, 29–33.

APPENDIX

Table A1Number of subjects in experiment 1

Dependent variable

Number ofavoiders

Discriminationratio

MeA-chicks Dem-chicks Males 12 4Females 12 5

Obs-chicks Males 12 4Females 14 6

Dem-chicks Males 12 12Females 12 13

Obs-chicks Males 12 12Females 12 13

Total N ¼ 98 N ¼ 69


Dependent variable

Numberof avoiders

Discriminationratio

MeA-chicks Dem-chicks Males LE 9 10RE 11 11

Females LE 14 15RE 13 13

Obs-chicks Males LE 15 15RE 11 11


Dry-chicks Dem-chicks Males LE 8 9RE 8 9




Total N ¼ 166 N ¼ 180


Dependent variable

Number ofavoiders

Discriminationratio

MeA-chicks Dem-chicks Males LE 5 7RE 5 5




Dry-chicks Dem-chicks Males LE 9 9RE 9 10




Total N¼116 N¼134


Lateralization of social learning in the domestic chick, Gallus gallus domesticus: learning to avoid

Documents

Transcript of Lateralization of social learning in the domestic chick, Gallus gallus domesticus: learning to avoid