Exp Brain Res (2010) 205:195–204

DOI 10.1007/s00221-010-2352-x

RESEARCH ARTICLE

Local sleep: a spatial learning task enhances sleep in the right hemisphere of domestic chicks (Gallus gallus)

Cristian Nelini · Daniela Bobbo · Gian Gastone Mascetti

Received: 16 April 2010 / Accepted: 27 June 2010 / Published online: 13 July 2010© Springer-Verlag 2010

Abstract During sleep, domestic chicks (Gallus gallus)show brief and transient periods during which one eye isopen while the other remains shut. Electrophysiologicalrecordings showed that the hemisphere contralateral to theopen eye exhibited an EEG with fast waves typical ofwakefulness, whereas the hemisphere contralateral to theclosed eye exhibited an electroencephalogram (EEG) typi-cal of slow-wave sleep. We investigated the time spent insleep and in monocular-unihemispheric sleep (Mo-Unsleep) following the learning of a spatial discriminationtask. A group of experimental chicks from days 8 to 11post-hatching were trained singly to select one containeramong four, having a hole on the top (making food avail-able) and positioned in a corner of a rectangular arena.Chicks of the control group did not learn the task becauseall four containers had a hole on the top and thereforechicks could randomly select any one of them. Experimen-tal and control chicks underwent the same number of trials.Experimental chicks had more total time spent sleepingthan control chicks. Experimental chicks spent more timein left Mo-Un sleep, which would be connected with adominance of the right hemisphere during learning trials.Control chicks showed no eye closure bias at days 8 and 9;however, a slight bias for more right eye closure at days 10and 11 was observed, suggesting that there was an absenceof hemispheric dominance during the Wrst 2 days of controltrials and a dominance of the right hemisphere during thelast 2 days of control trials. Overall, chicks that learned thespatial task slept signiWcantly more than chicks that wereexposed to the experimental paradigm but did not learn the

task. This suggests that the Mo-Un sleep pattern showed byexperimental chicks is a type of local sleep associated witha process of functional recovery and/or with consolidationof memory in the right hemisphere, which would be mainlyengaged during training trials.

Keywords Sleep · Unihemispheric sleep · Spatial learning · Domestic chick · Lateralization

Introduction

Domestic chicks (Gallus gallus) have a remarkable laterali-zation of brain functions (Rogers and Andrew 2002; Vallor-tigara and Rogers 2005). Overall, chicks show a left-eye/right-hemisphere superiority in spatial tasks and a right-eye/left-hemisphere superiority in tasks in which discriminationis based on classifying stimuli (i.e., color or pattern).Another aspect of chicks’ lateralization is monocular-uni-hemispheric sleep (Mo-Un sleep). During sleep, chicksshow brief and transient periods in which one eye is openwhile the other remains shut (Spooner 1964). Electrophysio-logical recordings have shown that the hemisphere contra-lateral to the open eye shows an EEG with fast waves typicalof wakefulness, while the hemisphere contralateral to theclosed eye shows an EEG typical of slow-wave sleep. More-over, numerous studies have shown that bilateral eye closureis associated with bihemispheric (Bin-sleep), slow wave andREM sleep (Ookawa and Takagi 1968; Ookawa 1971; Ballet al. 1988; Rattenborg et al. 1999; Bobbo et al. 2002).

It is known that experience during wakefulnessinXuences subsequent patterns and homeostasis of sleep(i.e., Horne and Walmsley 1976; Horne and Minard 1985;Borbely 2001, Martinez-Gonzalez et al. 2008; Rattenborget al. 2009; Hanlon et al. 2009). Several studies have

C. Nelini · D. Bobbo · G. G. Mascetti (&)Department of General Psychology, University of Padova, Via Venezia 12, Padua, Italye-mail: giangastone.mascetti@unipd.it

123

196 Exp Brain Res (2010) 205:195–204

pointed out that sleep has a regional aspect (local sleep),which is dependent on the speciWc activation of brainregions during wakefulness (Kattler et al. 1994; Huberet al. 2004). The study of Huber et al. (2004) provided alink between local sleep and learning. They showed thatthere was a signiWcant increase of SWS intensity in theright parietal region of subjects submitted to a task of motorrotation adaptation as compared to control subjects submit-ted to a no rotation task but under kinematically identicalcondition. That is, when one part of the brain is involved ina learning process, it would subsequently have more sleepthan other parts less or not at all involved in the learningprocess. The increase of local SWS would be beneWcial tothe local plastic changes in the right parietal regionsactively involved in the learning task (Huber et al. 2004).

Mascetti et al. (2004a, b, 2007) studied the associationbetween Mo-Un sleep patterns and spatial learning using aparadigm of right/left position discrimination of two whiteboxes. Experimental chicks learned the task and showed a sig-niWcant bias toward more left Mo-Un sleep suggesting a pre-dominant engagement of the right hemisphere during trials.This hypothesis was supported by studies indicating a domi-nant role of right hemisphere of domestic chicks in spatialtasks and in topographical orientation (Rashid and Andrew1989; Vallortigara et al. 1988). In contrast, control chicks didnot learn the task and showed no eye closure bias. Two inter-pretations were formulated for control Mo-Un sleep pattern(Mascetti et al. 2007): (1) an absence of hemispheric domi-nance ascribed to a lack of learning; (2) the absence of hemi-spheric dominance as a consequence of a low level of visualstimulation and a lack of suYcient incentive or motivation.An unambiguous interpretation of control data is essential forascribing a functional signiWcance of experimental Wndings;that is to say that the bias for more left Mo-Un sleep could beattributed to a right hemisphere dominance during learningand not to a higher level of stimulation and incentives or moti-vation of experimental chicks. Therefore, the aim of the pres-ent study was to record the patterns of Bin-sleep and Mo-Unsleep of chicks that learned a spatial position task, using theparadigm of geometric modules (Vallortigara et al. 1990),which allows both experimental and control chicks to be sub-mitted to the same amount of visual stimulation and to thesame level of motivation. Thereby, diVerences in sleep pat-terns could only be attributed to learning in experimentalchicks and non-learning in the control group.

Materials and methods

Subjects

The subjects were 12 female Hybro domestic chicks (Gal-lus gallus) hatched from eggs obtained from a commercial

hatchery (“Agricola Berica”, Montegalda, Vicenza, Italia).The eggs were incubated in the laboratory in an automati-cally turning incubator FIEM snc, MG 100H(45 £ 58 £ 43 cm), under constant temperature (37.7°C),humidity (about 50–60%) and darkness as previous studiesshowed that light stimulation in ovo aVects the pattern ofMo-Un sleep (Bobbo et al. 2002). We used 8-day-oldchicks because they show a consistent pattern of sleep andstable feeding and drinking behavior during the secondweek post-hatching. Moreover, at that age, they are alreadyimprinted and their body growth and motor activity aresuitable for the learning–training used in this experiment.

Apparatus

The apparatus (Fig. 1) used for chick rearing and sleeprecording has been described elsewhere (Bobbo et al. 2002,2006; Mascetti et al. 2004a). BrieXy, it consisted of twoglass home cages (30 £ 40 £ 40 cm) with semi-transparentclothes along the wall serving as a one-way screen. Eachcage was continuously lit from above by a 60 W electric-light bulb. Two transparent glass containers provided waterand food for the whole duration of the experiment. Theimprinting object was suspended freely in the middle of thecage at about head height for the chick.

For spatial training, a paradigm of geometric moduleswas used (Vallortigara et al. 1990). The apparatus (Fig. 2)consisted of a rectangular-shaped arena with walls made ofhomogeneously white-painted wood and the Xoor made ofwhite opaque plastic. The walls were 120 cm (longer walls)and 60 cm (shorter walls) in length and 40 cm in height.Transparent glass containers (diameter 5 cm, height 6 cm)were positioned at each of the four corners and Wlled withfood. All containers had a thin plastic net glued over thetop opening. In the experimental learning condition, one

Fig. 1 Schematic representation of the cages used for sleepobservations

123

Exp Brain Res (2010) 205:195–204 197

container also had a small hole (diameter 2 cm) cut on topof the net to allow chicks access to the food. In the controlcondition (see the following paragraph), all four containershad a small hole cut on top so that food was accessible to allof them. The top of the arena was covered by a veil andillumination was provided by three light bulbs (25 W each),positioned at the top that homogeneously illuminated thearena. The paradigm used for the testing of learning was thesame arena, but all containers at the four corners were com-pletely covered with plastic net, making food inaccessible.

Procedure

Immediately after hatching, chicks were placed singly intothe glass home cages and kept there from days 1 to 8 post-hatching. On day 7, chicks were placed singly into thetraining arena for 30–40 min, where they could move freelyexploring the arena and pecking for food at the corners.Thereafter, chicks were returned to the home cage anddeprived of food for 13 h to induce the necessary level ofmotivation. Training sessions were carried out on days 8, 9and 10 post-hatching. Sessions consisted of three blocks often trials, with each trial lasting a maximum of 2 min andeach session separated by a 3-min interval. At the begin-

ning of each trial, the chick was taken from a cardboard boxwith one hand after being disoriented, taking care to coverits eyes with the other hand for the entire duration of dis-placement. The animal was then placed in the middle of thearena in a random orientation with respect to the sagittalaxis of the body. During the inter-trial interval, the chickwas placed back into the cardboard box and slowly rotatedfor 5–10 s to exclude the use of compass or inertial infor-mation. Meanwhile, the Xoor of the arena was cleaned toensure that any trace of food or any other debris was com-pletely removed.

Six chicks (experimental group, Exp-Group) weretrained singly in the spatial task in which only one con-tainer with a hole on top, positioned in corner A of thearena (Fig. 3a), was available for food. The trial endedeither when the animal chose the reinforced container (A)and ate some food (few pecks were allowed), or when thechick failed to choose it. Six chicks (control group, Cont-Group) were trained in the control condition in which foodwas accessible from all the containers (Fig. 3b), with thetrial lasting until the animal ate some food from any one ofthe containers.

The Wrst corner the chick selected and approached wasscored. Approach and visual security of the container (usu-ally associated with pecking behavior) was necessary for avisit to be considered as a score. In the “spatial learning”condition, given the shape of the arena, two pairs of posi-tions could be identiWed: (1) the “correct corner”, whichincluded both the reinforced corner and its rotational equiv-alent (corners A and C, Fig. 3). In this position the longwall of the arena was at the left side of chick’s body and acorrect trial was scored either when the chick Wrstapproached corner A or corner C; (2) the “incorrect corner”,which included the other two corners (corners B and D,Fig. 3). In this position the long wall was at the right side ofthe chick’s body and when these corners were approached,

Fig. 2 The discrimination apparatus (arena) used for the spatial learn-ing task

Fig. 3 Schematic representa-tion of the arena with the corners where food containers were placed. a Arena used for spatial learning of Exp-Group chicks; corner A rewarded. b Arena used for Cont-Group chicks rewarded at any corner. c Arena used for test trials

123

198 Exp Brain Res (2010) 205:195–204

an incorrect trial was scored. The choice of reinforcing onlyone of the two correct corners was made in order to be surethat both were really indistinguishable for the chicks. Thelearning criterion was achieved when the animals chose thecorrect corners in 24 or more trials out of 30. Animals thatdid not reach this criterion on any of the training days werediscarded (25%).

On day 11, both groups of chicks were Wrst submitted to20 training trials and thereafter to two test blocks of 7 trialseach. In the “test condition”, the procedure was the same asfor training except that food was not accessible in any ofthe four containers (Fig. 3c), such that the tests were con-ducted in an extinction paradigm. The criterion wasattained when experimental chicks chose the corners A or Cin at least 80% or more of test trials. Control chicks weresubmitted to the same paradigm and the corner was chosenrandomly from trial to trial.

At the end of each training session and following the testsession, chicks were returned singly to their home cage(Fig. 1) and sleep behavior (eye closure and body position)was scored for 3 h consecutively by direct observation.Small mirrors mounted on rods allowed the experimentersto approach the animal and to check for the eye state (openor closed) when the chick’s position or posture hinderedobservation from outside. The experimenter recorded thenumber and duration of episodes of Bin-sleep (both eyesclosed) and Mo-Un sleep (one eye open while the otherremained shut).

The authors state that they have adhered to the “Princi-ples of Laboratory Animal Care (NIH publication N. 86-23,revised 1985)” and to the legal requirements of Italy(Authorization of Italian Ministry of Health, DM n. 156/2006-C).

Data analysis

For learning, the mean percentages of correct choices (cor-ner A or corner C) on the total choices showed by chicksduring training and test sessions were calculated. Data wereanalyzed by repeated measures ANOVA with CONDI-TION (Exp-Group and Cont-Group) as a between-subjectsfactor and SECTION (training days 8, 9, 10, 11 and “test”day 11) as a within-subjects factor. One-sample two-tailedt tests for the Exp-Group and for the Cont-Group were alsocalculated. SigniWcant departures from chance level (50%)indicated signiWcant choice for the correct spatial position.

For sleep, the time (calculated in seconds) of total sleep(binocular plus monocular), Bin-sleep, Mo-Un sleep (rightplus left) and left and right Mo-Un sleep, were analyzed byrepeated measures ANOVA (after checking for homogene-ity and normality of data distributions) with CONDITION(Exp-Group and Cont-Group) as between-subjects factorsand DAY (days 8, 9, 10 and 11) as within-subjects factors.

To evaluate the Mo-Un sleep biases, we used a propor-tional measure of the time spent sleeping in Mo-Un. This isbecause we wanted to check for left–right diVerences irre-spective of the absolute amount of Mo-Un sleep. The “lat-erality index” was calculated for time spent in right or leftmonocular sleep using the following formula: {[time (num-ber of episodes) spent with the left eye closed] ¡ [time(number of episodes) spent with the right eye closed]/[time (number of episodes) spent with the left eye closed] +[time (number of episodes) spent with the right eyeclosed]} £ 100. Laterality index was analyzed by repeatedmeasures ANOVA (after checking for homogeneity andnormality of data distributions) with CONDITION (Exp-Group and Cont-Group) as a between-subjects factor andDAY (days 8, 9, 10 and 11) as within-subjects factors. Sig-niWcant departures from chance level (0%) in the “lateralityindex”, which indicated signiWcant bias toward more rightor left eye closure, were evaluated with one-sample two-tailed t tests.

Results

Learning

The results about learning are reported in Fig. 4. They rep-resent the percentage of the Wrst choice in each position(correct position: corners A or C; and wrong position: cor-ners B or D) during 3 successive days of training and ashort training session at day 11 before the test. Repeatedmeasures ANOVA showed a main eVect of CONDITION[F(1,10) = 144,623, p < 0.001], a main eVect of DAY[F(4,40) = 6,745, p < 0.001] and an interaction eVect ofCONDITION £ DAY [F(4,40) = 3,955, p = 0.008]. The datashowed a systematic improvement in searching behavior atcorner A or C (correct position) for chicks of the Exp-Group, but the performance of chicks in the Cont-Groupremained at chance level. One-sample two-tailed t tests cal-culated for each single group at each day revealed signiW-cant departures from chance level (50%) only for theexperimental group, which was present during all trainingsessions [day 8, t(5) = 3,180, p = 0.025; day 9, t(5) = 4.722,p = 0.005; day 10, t(5) = 8,675, p < 0.001 day 11,t(5) = 7,980, p < 0.001] and also during the test session [test,t(5) = 15,539, p < 0.001]. Moreover, chicks of the Exp-Group reached the learning criterion (80% of correctchoices) at day 11 and maintained that level of performanceduring the test session.

Time spent sleeping

Repeated measures ANOVA showed a main eVect of CON-DITION [F(1,10) = 7,206, p = 0.023]. Overall, chicks of the

123

Exp Brain Res (2010) 205:195–204 199

Exp-Group spent more time sleeping than chicks of theCont-Group (Fig. 5).

Binocular sleep

Total time spent in Bin-sleep is shown in Fig. 6. Repeatedmeasures ANOVA revealed a signiWcant main eVect ofCONDITION [F(1,10) = 7,273, p = 0.022]. The mean durationof episodes of Bin-sleep is shown in Fig. 7. Repeated mea-sures ANOVA revealed a signiWcant main eVect of CONDI-TION [F(1,10) = 5,319, p = 0.044]. Chicks of Exp-Groupspent signiWcantly more time in Bin-sleep and had longerepisodes of Bin-sleep than chicks of the Cont-Group.

Monocular/unihemispheric sleep

Total time spent in Mo-Un sleep and time spent in right,left and total Mo-Un is shown in Figs. 8, 9 and 10.

Fig. 4 Learning performance of Exp-Group and Cont-Group chicks.Learning criterion and chance level are shown. Abscissa the trial days.Ordinate percentage of correct responses

Fig. 5 Total time spent sleeping of Exp-Group and Cont-Groupchicks. Abscissa trial days. Ordinate time in seconds

Fig. 6 Total time spent in Bin-sleep of Exp-Group and Cont-Groupchicks. Abscissa trial days. Ordinate time in seconds

Fig. 7 Duration of episodes of Bin-sleep of Exp-Group and Cont-Group chicks. Abscissa trial days. Ordinate mean duration in seconds

Fig. 8 Total time spent in Mo-Un sleep of Exp-Group and Cont-Group chicks. Abscissa trial days. Ordinate time in seconds

123

200 Exp Brain Res (2010) 205:195–204

Repeated measures ANOVA on the total time spent inMo-Un sleep (right and left together) and right Mo-Un sleep(right eye closed) revealed no signiWcant eVects (Figs. 8, 9).Repeated measures ANOVA on the time spent in leftMo-Un sleep (left eye closed) revealed a signiWcant eVectof CONDITION [F(1,10) = 6,715, p = 0.027] (Fig. 10). Totaltime spent in Mo-Un sleep and time spent in right Mo-Unsleep was similar in both groups of animals, whist timespent in left Mo-Un sleep was signiWcantly higher in theExp-Group chicks than in the Cont-Group.

Laterality index of Mo-Un sleep

Laterality index of the percentage of time spent in Mo-Unsleep is shown in Fig. 11. Repeated measures ANOVArevealed signiWcant main eVects for both CONDITION[F(1,10) = 125,343, p < 0.001] and DAY [F(3,30) = 3,135,p = 0.04]. Chicks of the Exp-Group showed a preferencefor left Mo-Un sleep, whereas chicks of the Cont-Groupshowed the reverse pattern, with a preference for right Mo-Un sleep. This eVect was consistent during all training andtest sessions. One-sample two-tailed t tests revealed signiW-cant departures from chance level (0%) with a bias towardmore left Mo-Un sleep for chicks of the Exp-Group aftertraining sessions [day 8, t(5) = 10,705, p < 0.001; day 9,t(5) = 9,279, p < 0.001; day 10, t(5) = 6,293, p = 0.001] andafter the test session [t(5) = 6,194, p = 0.002]. The t tests onchicks of the Cont-Group revealed signiWcant departuresfrom chance level (0%) with a bias toward more rightMo-Un sleep only after the 10th day of the training ses-sion [t(5) = ¡2,913, p = 0.033] and the test session[t(5) = ¡2,643, p = 0.046]. Overall, chicks of the Exp-Group showed a signiWcant bias for more left Mo-Un sleepafter all training sessions and the test session, while chicksof the Cont-Group showed a signiWcant bias for more right

Fig. 9 Total time spent in right Mo-Un sleep of Exp-Group and Cont-Group chicks. Abscissa trial days. Ordinate time in seconds

Fig. 10 Total time spent in left Mo-Un sleep of Exp-Group and Cont-Group chicks. Abscissa trial days. Ordinate time in seconds

Fig. 11 Laterality index of the percentage of time (seconds) spent in Mo-Un sleep of Exp-Group and Cont-Group chicks. Asterisks indicate signiWcant departures from chance level (0%)

123

Exp Brain Res (2010) 205:195–204 201

Mo-Un sleep only during the 10th training day and testsession.

Discussion

Chicks of the Exp-Group spent signiWcantly more timesleeping and total time in Bin-sleep than chicks of the Cont-Group, in agreement with current evidence which indicatesthat waking experiences (in this case, learning) inXuencesubsequent duration, pattern and homeostasis of sleep (inhumans: Horne and Walmsley 1976; Horne and Minard 1985;Borbely 2001; in animals: Martinez-Gonzalez et al. 2008;Rattenborg et al. 2009; Hanlon et al. 2009). The increase inthe time spent in binocular eye closure (Bin-sleep) ofExp-Group chicks would reXect an increase of bihemi-spheric SWS and REM sleep.

Overall, it may be suggested that an Mo-Un sleep biaswould be a consequence of dominant activation of onehemisphere during trials, while an absence of bias would beassociated with a lack of hemispheric dominance during tri-als. There was no diVerence between both groups of chicksin the total time spent in Mo-Un sleep and in the right Mo-Un sleep. However, chicks of the Exp-Group showed a sig-niWcant bias for more left Mo-Un sleep and more time spentin left Mo-Un sleep (left eye closure) after each session ofspatial training. Therefore, when the spatial position wasthe proper learning cue for gaining access to reinforcement,there would be a predominant activation of the right hemi-sphere and, subsequently, a signiWcant bias for more leftMo-Un sleep. This Wnding is in agreement with a previousstudy by Mascetti et al. (2004b, 2007) showing a signiWcantbias for more left Mo-Un sleep after learning of a spatialdiscrimination of left/right position. The chick’s preferen-tial involvement of the right hemisphere in spatial tasks hasbeen reported in several studies (Rashid and Andrew 1989;Rogers and Anson 1979; Vallortigara 2000; Vallortigaraand Andrew 1991; Vallorigara et al. 1996; Vallortigara andRogers 2005). In particular, Vallortigara et al. (1988)observed asymmetries in conditions of binocular vision ofsimultaneous visual discrimination learning. In that task,which involved right–left discrimination of the spatial posi-tion of two otherwise identical discriminanda, learning cri-terion was reached faster when the negative stimulus wasplaced on the right side than on the left side of the chick’sbody. Andrew (2002) reported that chicks showed a stand-ing bias for a left hemisphere dominance at day 8, a no-clear bias at day 9 (probably it is a transition day of shiftingdominance) and a right hemisphere dominance at days 10and 11. Therefore, in the present study, the spatial learningreversed the hemisphere dominance bias of Exp-Groupchicks at day 8 and drove the dominance in favor of righthemisphere at day 9, both followed by a subsequent bias for

more left Mo-Un sleep. At days 10 and 11 (Andrew 2002),standing bias was for a right hemisphere dominance, whichfairly coincided with the right hemisphere control of spatiallearning behavior.

Chicks of the Cont-Group at days 8 and 9 showed no eyeclosure bias, that is to say they opened/closed both eyesequally during sleep, which could be associated with anabsence of both, learning (scores were at chance level) andhemispheric dominance during trials. According to thepattern of standing bias reported by Andrew (2002), lefthemisphere dominance would be suppressed in 8-day Cont-Group chicks that was followed by a lack of Mo-Un sleepbias. Likely, in Cont-Group chicks placed in a new environ-ment (an arena), none of the hemispheres prevailed in thecontrol of behavior used for Wnding food. Since the stand-ing was not clearly deWned at day 9 (Andrew 2002), thepattern coincided with the absence of bias of day 9 Cont-Group chicks. However, at days 10 and 11 there was aslight, but signiWcant, bias for more right Mo-Un sleep,suggesting that there would be a higher activation of the lefthemisphere during trials, although the spatial scores were atchance level. This pattern is opposite to the standing biasfor a right hemisphere dominance reported by Andrew(2002) at days 10 and 11. It may be assumed that controlchicks used the right eye when they explored the arena andsubsequently decided which one of the four corners andcontainers to approach for getting food. In other terms,there would be a sort of categorization, which is a behaviorunder a dominant control of the left hemisphere (Gastonand Gaston 1984; Vallortigara 1989, Vallorigara et al.1996).

There are some diVerences between the present studyand a previous one of Mascetti et al. (2007). Those authorsstudied the relationship between spatial learning and sleepusing a paradigm of discrimination of the left–right positionof positive and reward stimuli. They did not Wnd any sig-niWcant diVerences between experimental and controlchicks for total time spent sleeping or Bin-sleep, whereas inthe present study both sleep measures were higher in theExp-Group than in the Cont-Group. The diVerence in theamount of sleep between the two studies could be ascribedto diVerences between the learning paradigms. In the para-digm of geometric modules used in the present study(Vallortigara et al. 1990), chicks chose one reinforced spatialposition among four possible choices, which could be amore demanding task and thereby inducing more time spentsleeping than if they only have to choose the left/right sideposition of one box between two (Mascetti et al. 2007).Another diVerence between the two studies is related to theMo-Un sleep pattern of the Cont-Group chicks. In the studyof Mascetti et al. (2007), control chicks showed an absenceof Mo-Un bias on all of the training days, whereas in thepresent study the bias was absent only on days 8 and 9 and

123

202 Exp Brain Res (2010) 205:195–204

there was a bias for more right Mo-Un sleep on days 10 and11. Two interpretations were formulated by Mascetti et al.(2007) for explaining the lack of Mo-Un bias of their con-trol chick that did not necessarily exclude each other: (1)there was an absence of hemispheric dominance during tri-als because there was no learning of a spatial position; (2)the absence of hemispheric dominance would be ascribedto both a low level of visual stimulation (two white identi-cal boxes) and a low level of incentives or motivation(chicks were reinforced at 50% of trials and randomly atleft and right position) (Mascetti et al. 2004b, 2007). In thepresent study, the levels of both visual stimulation andincentives were similar for chicks of both the Exp-Groupand the Cont-Group and therefore an inXuence of both con-ditions can be ruled out. This suggests that the Mo-Un pat-tern of the Cont-Group chicks can safely be attributed to alack of hemispheric dominance during trials at days 8 and 9and to a left hemispheric dominance at days 10 and 11.

It is known that sleep has a role in memory consolidation(i.e., Ambrosini et al. 1988; Smith and Butler 1982; Win-son 1993; Smith 1996; Graves et al. 2001; Maquet 2001;Walker and Stickgold 2004, 2006; Orban et al. 2006). Onone hand, it may be argued that the functional role of thebias for more left Mo-Un sleep of the Exp-Group chickswould be associated with a process of consolidation of spa-tial memory in the right hemisphere. A relationshipbetween learning and Mo-Un sleep pattern was reported byMascetti et al. (1999) regarding the consolidation ofimprinting memories. They found that imprinted chicksshowed relatively more right Mo-Un sleep than chicks inwhich the imprinting process was not allowed. This biaswas assumed to be related to the consolidation of imprint-ing memories in the left hemisphere (Mascetti et al. 1999).On the other hand, during trials there would be a higheractivation of the right hemisphere and the bias for more leftMo-Un sleep would also reXect a process of recovery in theright hemisphere. Episodes of Mo-Un sleep in chicks areintermingled with periods of binocular sleep. In the presentstudy, chicks of the Exp-Group showed signiWcantly moretotal sleep and Bin-sleep than chicks of the Cont-Group,but an additional diVerence resided in the Mo-Un sleepbias. The right hemisphere, being engaged in spatial learn-ing trials, had relatively more total sleep (binocular plusmore Mo-Un sleep) than the non-dominant left hemisphere.In other terms, the increased left Mo-Un sleep recorded inthe Exp-Group chicks may slightly extend the sleeping timeof the right hemisphere, thus favoring a recovery processand/or consolidation of spatial memory.

It has been proposed that the biological functions of Mo-Un sleep in birds include the possibility for the animal toperiodically and monocularly monitor the environment dur-ing sleep against the risk of predation (Rattenborg et al.1999) or to check for the presence of a social partner

(Mascetti et al. 1999; Bobbo et al. 2006). In a typical epi-sode of Mo-Un sleep, the chick opens one eye and the con-tralateral hemisphere awakens (EEG of fast and lowamplitude waves) (Bobbo et al. 2002). Therefore, the pat-tern of eye opening seems to be determined by the hemi-sphere mainly engaged during trials. In other words, toperiodically monitor the environment, chicks of the Exp-Group opened the eye connected with the hemisphere thatwas not engaged in learning spatial trials and with less needfor functional recovery and memory trace consolidation. Infact, chicks of the Exp-Group showed a bias for more righteye opening, while chicks of the Cont-Group opened botheyes equally at days 8 and 9 (i.e., neither hemisphere waspredominantly activated during trials) and opened the lefteye at days 10 and 11 (i.e., the right hemisphere was notinvolved in the control of behavior during control trials).

Another aspect regards the time spent in Mo-Un sleep ascompared to the total time spent sleeping. In this study andin previous ones (Mascetti et al. 1999; Mascetti and Vallor-tigara 2001; Bobbo et al. 2002; Mascetti et al. 2004a, b,2007), total time spent in Mo-Un sleep was around 1–2% oftotal time sleeping. Therefore, as the percentage of timespent in Mo-Un sleep is quite low, a legitimate issue iswhether Mo-Un sleep should have a general biologicalfunction. Some Wndings can be argued in favor of a role forMo-Un sleep: (A) the unilateral closure/opening of the rightor left eye during sleep does not occur randomly, but isclosely correlated with shifts of hemispheric dominanceduring the chick’s development (Mascetti et al. 1999,2004a, b). (B) Mo-Un sleep patterns are modulated by spe-ciWc experiences during wakefulness, such as the presence,absence or changes of the imprinting object (Mascetti et al.1999, 2004a, b), light stimulation in ovo (Mascetti and Val-lortigara 2001; Bobbo et al. 2002), learning (Mascetti et al.2007 and the present study) and brief monocular occlusion(unpublished data). In addition, if eye opening is connectedto a vigilance function during sleep and/or to a control envi-ronment or the social partner (Rattenborg et al. 1999;Mascetti et al. 1999), it can be argued that 1–2% of Mo-Unsleep should be a suYcient percentage of time to accom-plish those functions. On the other hand, a highly signiW-cant left–right diVerence in eye closure/eye openingoccurring in a small portion of total sleep time would alsobe relevant considering that several important behavioralpatterns occur with reduced frequency and time in animals(i.e., anti-predatory or aggressive responses), but are none-theless of enormous relevance from a biological point ofview.

Several studies have pointed out that sleep has a regionalaspect dependent on the speciWc activation of brain regionsduring wakefulness (Kattler et al. 1994; Vyazovskiy et al.2000; Oleksenko et al. 1992; Huber et al. 2004, Hanlonet al. 2009). This aspect of sleep is named local sleep. It

123

Exp Brain Res (2010) 205:195–204 203

postulates that when one part of the brain is involved in alearning process, it would subsequently have more sleepthan other parts that are less or not at all involved in thelearning process or experience. These studies suggestedthat the increase of local SWS would be beneWcial to thelocal plastic changes in those cortical regions activelyinvolved in the learning task. It has been suggested thatMo-Un sleep may be involved the processes of hemisphericplastic changes and recovery (Horne 1988; Bennington andHeller 1995; Vyazovskiy et al. 2000). Following this line ofthought, the bias for more right hemisphere sleep found inthe Exp-Group chicks would reXect the diVerential involve-ment of the two hemispheres in the learning tasks. Theincrease of Mo-Un sleep would be beneWcial to the hemi-sphere directly involved in learning and further conWrm thatsleep, while being a global process, also shows a regionalaspect, which is also found in birds.

Acknowledgments This study has been supported by Grant-PRIN-COFIN n. 200073JY8HY4_001 of the Italian Ministry of Universityand ScientiWc Research (MIUR). The authors thank Dr. Sarah Lough-ran of University of Zurich for a critical review of the manuscript andher helpful comments. The authors state that they have adhered to the“Principles of Laboratory Animal Care (NIH publication No. 86-23,revised 1985)” and to the legal requirements of our country (Authori-zation of Italian Ministry of Health, DM n. 156/2006-C).

References

Ambrosini MV, Sadile AG, Gironi Carnevale UA, Mattiaccio M,Giuditta A (1988) The sequential hypothesis on sleep function.I. Evidence that the structure of sleep depends on nature of previouswaking experience. Physiol Behav 143:325–337

Andrew RJ (2002) Behavioural development and lateralization. In:Rogers LJ, Andrew RJ (eds) Comparative vertebrate lateraliza-tion. Cambridge University Press, Cambridge, pp 157–205

Ball NJ, Amlaner CD, ShaVery JP, Opp MR (1988) Asynchronouseye-closure and unihemispheric sleep in birds. In: Koella WP,Schultz H, Obal F, Visser P (eds) Sleep’86. Gustav Fischer Verlag,Stuttgart, pp 151–153

Bennington JH, Heller HC (1995) Restoration of brain energy metab-olism as a function of sleep. Prog Neurobiol 45:347–360

Bobbo D, Galvani F, Mascetti GG, Vallortigara G (2002) Light expo-sure of chick embryo inXuences monocular sleep. Behav BrainRes 134:447–466

Bobbo D, Vallortigara G, Mascetti GG (2006) The eVects of early post-hatching changes of imprinting object on the pattern of monocu-lar/unihemispheric sleep of domestic chicks. Behav Brain Res170:23–28

Borbely AA (2001) From slow waves to sleep homeostasis: newperspectives. Arch Ital Biol 139:53–61

Gaston KE, Gaston MG (1984) Unilateral memory after binocular dis-crimination training: eft hemisphere dominance in the chick?Brain Res 303:190–193

Graves L, Pack A, Abel T (2001) Sleep and memory: a molecular per-spective. Trends Neurosci 24:237–243

Hanlon EC, Farraguna U, Vyazovskij V, Tononi G, Cirelli C (2009)EVects of skilled training on sleep slow wave activity and corticalgene expression in the rat. Sleep 32:719–729

Horne JA (1988) Why we sleep? The functions of sleep in humans andother animals. Oxford University Press, New York

Horne JA, Minard A (1985) Sleep and sleepiness following a behavio-urally active day. Ergonomics 28:567–575

Horne JA, Walmsley B (1976) Daytime visual load and the eVectsupon human sleep. Psychophysiol 13:115–120

Huber R, Ghilardi MF, Massimini M, Tononi G (2004) Local sleep andlearning. Nature 430:78–81

Kattler H, Dijk DJ, Boberly A (1994) EVects of unilateral somatosen-sory stimulation prior to sleep to the sleep EEG in humans.J Sleep Res 3:159–164

Maquet P (2001) The role of sleep in learning and memory. Science294:1048

Martinez-Gonzalez D, Lesku JA, Rattenborg NC (2008) IncreasedEEG spectral power density during sleep following short-termsleep in pigeons Columbia livia: evidence for avian sleep homeo-stasis. J Sleep Res 17:140–153

Mascetti GG, Vallortigara G (2001) Why do birds sleep with one eyeopen? Light exposure of chick embryo as a determinant of mon-ocular sleep. Curr Biol 11:971–974

Mascetti GG, Rugger M, Vallortigara G (1999) Visual lateralizationand monocular sleep in the domestic chick. Cogn Brain Res7:451–463

Mascetti GG, Bobbo D, Rugger M, Vallortigara G (2004a) Monocularsleep in male domestic chick. Behav Brain Res 153:447–452

Mascetti GG, Rugger M, Vallortigara G, Menesatti S (2004b) Uni-hemispheric sleep and visual learning in domestic chicks. J SleepResearch 13(suppl. 1):482

Mascetti GG, Rugger M, Vallortigara G, Bobbo D (2007) Monocular-unihemispheric sleep and visual discrimination learning in thedomestic chick. Exp Brain Res 176:70–84

Oleksenko AI, Mukhametov LM, Polyakov IG, Supin AY, KovalzonVM (1992) Unihemispheric sleep deprivation in bottlenose dol-phins. J Sleep Res 1:40–44

Ookawa T (1971) Electroencephalograms recorded from the telen-cephalon of blinded chickens during behavioural sleep and wake-fulness. Poultry Sci 50:731–736

Ookawa T, Takagi K (1968) Electroencephalograms of free behavioralchicks at various developmental ages. Jpn J Physiol 18:87–99

Orban P, Rauchs G, Balteau E, Degueldre C, Luxen A, Maquet P, Peig-neux P (2006) Sleep after spatial learning promotes covert reorga-nization of brain activity. PNAS 103:7124–7129

Rashid N, Andrew RJ (1989) Right hemisphere advantages for topo-graphical orientation in the domestic chick. Neuropsychologia27:937–948

Rattenborg NC, Lima SL, Amlaner CJ (1999) Half-awake to the riskof predation. Nature 397:397–398

Rattenborg NC, Martinez-Gonzalez D, Lesku JA (2009) Avian sleephomeostasis: convergent evolution of complex brains, cognitionand sleep function in mammals and birds. Neurosci Behav Rev33:253–270

Rogers LJ, Andrew RJ (eds) (2002) Comparative vertebrate lateraliza-tion. Cambridge University Press, Cambridge

Rogers LJ, Anson JM (1979) Lateralization of function in the chickenforebrain. Pharmacol Biochem Behav 10:679–686

Smith C (1996) Sleep states, memory processes and synaptic plasticity.Behav Brain Res 78:49–56

Smith C, Butler S (1982) Paradoxical sleep at selective times followingtraining is necessary for learning. Physiol Behav 29:469–473

Spooner C (1964) Observations on the use of the chick in the pharma-cological investigation of the central nervous system. Ph.D. Dis-sertation, UCLA

Vallorigara G, Regolin L, Bortolomiol G, Tommasi L (1996) Lateralasymmetries due to preference in eye use during visual discrimi-nation learning in chicks. Behav Brain Res 74:135–143

123

204 Exp Brain Res (2010) 205:195–204

Vallortigara G (1989) Behavioural asymmetries in visual learning ofyoung chickens. Physiol Behav 45:797–800

Vallortigara G (2000) Comparative neuropsychology of the dual brain:a stroll through left and right animals’ perceptual worlds. BrainLang 73:189–219

Vallortigara G, Andrew RJ (1991) Lateralization of response to changein a model partner by chicks. Anim Behav 41:187–194

Vallortigara G, Rogers LJ (2005) Survival with an asymmetrical brain:advantages and disadvantages of cerebral lateralization. BehavBrain Sci 28:575–589

Vallortigara G, Zanforlin M, Cailotto M (1988) Right–left asymmetryin position learning of male chicks. Behav Brain Res 27:189–191

Vallortigara G, Zanforlin M, Pasti G (1990) Geometric modules in ani-mal spatial representations: a test with chicks (Gallus gallus).J Comp Psychol 104:248–254

Vyazovskiy V, Borbely A, Tobler I (2000) Unilateral vibrissae stimu-lation during waking induces interhemispheric asymmetry duringsubsequent sleep in the rat. J Sleep Res 9:367–371

Walker MP, Stickgold R (2004) Sleep dependent learning and memoryconsolidation. Neuron 44:121–133

Walker MP, Stickgold R (2006) Sleep, memory and plasticity. AnnRev Psychol 57:139–166

Winson J (1993) The biology and function of rapid eye movementsleep. Curr Opin Neurobiol 3:243–248

123