-
THE BEHAVIOR OFTHE LABORATORY RATA Handbook with Tests
Edited by
IAN Q. WHISHAWBRYAN KOlBDepartment of psychology and
NeuroscienceCanadian Centre for Behavioural Neuroscience
.
t
OXFORDUNIVERSITY PRESS
2005
-
Social LearningBENNElTG. GALEF, JR.
Systematic observation of free-living mam-mals and birds often
reveals differences in thebehavior of species members that live in
dif-ferent areas. Such geographic variation in thebehavior of
chimpanzees and orangutans isparticularly well documented (Whiten
et al.,1999; van Schaik et al., 2003) and is widelyknown because of
the attention it has receivedin the popular press. However, before
the re-cent dramatic increase in field studies of thegreat apes, it
was not unreasonable to pro-pose, as did Steiniger (1950, p. 369),
that "the[Norway] rat appears especially able to de-velop local
traditions, more so perhaps thanother more-closely examined
mammals, pos-sibly including the anthropoids."
NORWAY RATS
Norway rats are arguably the most successful,and surely the most
widely distributed, non-human mammals on Earth. Breeding
popula-tions have been reported from Nome, Alaska,at 60 degrees
North latitude, where rats feedon human garbage, to South Georgia
Island,at 55 degrees South latitude, where tussockgrass, beedes,
and ground-nesting birds pro-vide sustenance for colonies of Norway
rats.
As the preceding two examples suggest,much of the success that
rats enjoy resultsfrom the extraordinary range of foods thatthey
are able to exploit, and as in the greatapes, much of the known
variation in behav-ior in free-living Norway rats involves
forag-ing behavior. Rats in West Virginia catch and
34eat fingerling fish in trout hatcheries, whereasthose living
on Norderoog island in the NorthSea stalk and kill ducks and
sparrows. Yetother R. norvegicus living along the banks ofthe Po
River in Italy dive for and feed on mol-lusks living on the bottom
of the river, whiletheir fellow rats in Japan scavenge dead
fishthat wash up on the seashore. Such naturallyoccurring
variability in feeding behavior hasbeen the focus of most
experimental studiesof social learning in the species.
PREVIEW
I begin the present brief review of the litera-ture on social
influences on food choices ofNorway rats with a description of
fieldworkstrongly suggesting that interactions betweenadult
free-living rats and their young can de-termine which foods the
young come to eat.I then describe very briefly several
behavioralprocesses that have been shown in the labo-ratory to be
sufficient to influence food choicein young rats. Last, I describe
in somewhatgreater detail a type of social influence on rats'food
preferences that has already proved to beuseful in studies o~ the
physical substrates oflearning and memory.
FIELD OBSERVATIONSOF NORWAY RATS
Fritz Steiniger, an applied ecologist whoseprofessional interest
lay in enhancing the effi-
363
-
364
ciency with which rodent pests could be ex-terminated, was the
first to report difficultiesin controlling pest populations of
Norway ratsusing the economically desirable method ofplacing
permanent stations containing poi-soned bait in rat-infested areas
(Steiniger,1950). Steiniger found that although rats ateample
amounts of poison bait and died inlarge numbers when a permanent
bait stationwas first introduced into their colony's terri-tory,
later acceptance of the bait by colonymembers was very poor, and
colonies tar-geted for extermination soon returned to theirinitial
sizes.
Steiniger reported that permanent baitstations failed because
young rats, born tocolony members that had survived their ini-tial
contact with the poisoned bait and hadlearned to avoid eating it,
refused to even tastethe bait that the adults of their colony
wereavoiding.
A LABORATORY ANALOGUE
Avoidance by young wild rats of a food thatadults of their
colony have learned to avoideating is a robust phenomenon that is
easilyobserved in rats transferred from their natu-ral habitats to
laboratory enclosures. We cap-tured adult wild rats (R. norvegicus)
on garbagedumps in southern Ontario, transferred themto our
laboratory, and placed them in groupsof five or six in 2 m2
enclosures that each con-tained nesting boxes and nesting materials
andprovided ad libitum access to water. For 3hours each day, we
offered each colony twofoods that differed in taste, smell,
texture, andcolor (Galef and Clark, 1971b).
To begin a typical experiment, we intro-duced a sublethal
concentration of toxin intoone of the two foods that we gave our
captivesto eat daily. The rats soon learned to avoid eat-ing the
poisoned food, and for weeks there-after, they avoided eating the
food that hadbeen noxious, even when we gave them un-contaminated
samples ofit (Garcia et al., 1966).
DEFENSE AND SOCIAL BEHAVIOR
After we had trained our colonies toavoid one of the two foods
that we placed intheir enclosure each day, we waited for
femalecolony members to give birth and for theiryoung to grow to
weaning age. As the youngapproached independence, we started to
ob-serve their colony on closed-circuit televisionthroughout daily
feeding periods. When theyoung started to eat solid food, we
recordedthe frequency with which they ate each of thetwo foods in
their cage: one that adult colonymembers were eating and the other
that theadults had learned to avoid.
We found, without exception, that wean-ing rats ate only the
food that the adults oftheir colony were eating and totally
avoidedthe alternative food that the adults hadlearned to avoid.
Even after we removed pupsfrom their natal enclosures, housed them
in-dividually, and offered them the same twofoods that had been
available when they werein their colony cages, pups continued to
eatonly the food that the adults of their colonyhad eaten (Galef
and Clark, 1971b) (Fig. 34-1).
ANALYSISOF THE PHENOMENON
My students and I have spent much of the past30 years
determining how the food choices ofadult rats might influence those
of the youngthey rear (see reviews in Gale, 1977, 1988,1996a,
1996b). Over those years, those work-ing in my laboratory and in
other laboratoriesas well have discovered many different waysin
which the food choices of young rats areaffected by social
interactions with conspecificadults.
Prenatal EffectsFetal rats exposed to a flavor while still in
theirmother's womb (through injection of a fla-vored solution into
the dam's amniotic fluid)will, when grown, drink more of a
solutioncontaining that flavor than will control ratsthat lack
prenatal exposure to it (Smother-man, 1982). Even feeding a food
with a strongodor to a female rat while she gestates a litter
-
With colony
c:100
G>OJ Colony poisoned on;G> 80 Diet B ___0 40G;a.c:as
20G>
Chapter 34. Social Learning
In isolation
2 4 6 8 10 12 14 16 18 20 22 24 26Day
Figure 34-1. Rat pups born into colonies trained to avoideating
either diet A or diet B are offered a choice for 3 hoursper day
between diet A and diet B. Abscissa shows the dayssince pups
started to eat solid food; ordinate, relative fre-quency with which
pups &om the two types of colony atediet A, Pup diet choice
while still in their natal colonies (lift)and the amount of diet A
eaten, as a percentage of totalamount eaten, by pups afTer transfer
to individual cages andoffer of diet A and diet B for 9 hours per
day (right). (Data&om Galef and Clark [1971].)
suffices to enhance her postnatal preferencesof her young for
the odor of that food (Hep-per, 1988).
Effects during Suckling.
1
"
Flavors of foods that a rat dam eats while lac-tating affect the
flavor of her milk, and expo-sure to milk flavored by the foods
that a lac-tating dam eats while rearing her young
r affects the food preferences of her pups at
\
weaning (e.g., G.Jef and Sherry, 1973).
Effects during WeaningGalef and Clark (1971a) used time-lapse
videorecordings to observe each of nine wild ratpups that had ad
libitum access to solid foodtake its first meal. All nine pups ate
solid foodfor the first time under the same circum-stances. Each
ate at the same time that anadult member of its colony was
feeding,which was highly unlikely given the temporaldistribution of
adult meals, and each ate at thesame place an adult was feeding,
not at an al-ternative feeding site a short distance away.
365
Even an anesthetized adult rat placed near oneof two otherwise
identical feeding sites madethat site far more attractive to pups
than onewithout an adult present (Galef, 1981).
By comparing the circumstances inwhich intact and visually
deprived ratsweaned, we found that intact pups use visualcues to
approach adults trom a distance whenselecting a place to eat solid
food for the firsttime.
Effects of Snatching Food from AdultsYoung rats, like the young
of many othermammalian species, seem to be especially in-terested
in the particular piece of food thatsomeone else is eating.
Juvenile rats will walkacross a cage floor carpeted with food
pelletsand steal an identical pellet tram the mouthor paws of an
adult or a peer that is eating it.Young rats that have stolen a
pellet of unfa-miliar food tram the mouth of a
conspecificsubsequently show a greater preference forthat food than
do young rats that have eatenan identical food pellet taken trom
the floorof their cage (Galef et al., 2001).
Effects of Scent Marks and Scent TrailsWhile feeding, adult rats
deposit olfactorycues both on and around a food they are eat-ing
(Galef and Beck, 1985). Such residualodors attract pups and, like
the physical pres-ence of an adult rat at a feeding site,
causeyoung rats to prefer marked sites. Further,when an adult has
finished eating and travelsback to its burrow, it deposits a scent
trail thatdirects young rats seeking food to the locationat which
the adult ate (Galef and Buckley,1996).
IMPLICATIONS OF REDUNDANCY
Redundancy in the behavioral processes thatsupport social
influences on food choice inrats is in itself important. Such
redundancysuggests that for rats, as for the honeybeesstudied by
Karl van Frisch (1967), socially ac-quired information
substantially increases for-
- cQ)1iiQ)cU 60CDi:5'EQ)e 40Q)a.c
-
Chapter 34. Social Learning
The effects of a single brief exposure torecently fed
demonstrator rats on the foodchoices of their observers are both
surpris-ingly powerful and surprisingly long lasting.Many observer
rats taught to totally avoid in-gesting a food by following its
ingestion withan injection of toxin, and then placed
withdemonstrator rats that have eaten the foodthat their observers
had learned to avoid, to-tally abandoned their aversions.
Similarly,most observer rats that interacted with ademonstrator fed
a diet adulterated withcayenne pepper (an inherently
unpalatabletaste to rats) subsequently preferred pepperedto
unadulterated diet (Galef, 1986b). Such ef-fects of demonstrator
rats on the food choicesof their observers can be seen a month
ormore after a demonstrator and observer in-teract (Galef and
Whiskin, 2003).
ANALYSIS
The behavioral process that produces such so-dal influence on
the food choices of observerrats is now quite well understood.
Olfactorycues passing to observer rats &om demonstra-tors cause
observers to increase their prefer-ences for the foods that their
respective dem-onstrators ate (Galef and Wigmore, 1983).Observers
sniff at the mouths of demonstra-tors, and this sampling of a
demonstrator'sbreath is both necessary and sufficient
fordemonstrators to influence the later foodchoices of observers
(Galef and Stein, 1985).
Both food-related odors escaping fromthe digestive tract of a
demonstrator and thescent of bits offood clinging to a
demonstra-tor's fur and vibrissae allow observers to iden-tify the
food that a demonstrator has recentlyeaten. And after an observer
rat experiencessimultaneously the scent of a food and ratbreath,
the observer shows an enhanced pref-
.
erence for the food the scent of which it ex-: perienced
together with rat breath (Galef and. Stein, 1985).
Gas chromatography performed on sam-ples of rat breath has shown
that it contains
367
two sulfur compounds: carbon disulfide andcarbonyl sulfide. Rats
exposed to a fooddusted onto either the head of an
anesthetizedconspecific or a piece of cloth moistened witha dilute
solution of carbon disulfide subse-quently show an enhanced
preference for thatfood. To the contrary, rats exposed to a
food,that had been dusted onto the head of a deadconspecific, onto
the rear of a live conspecificor onto a piece of cloth moistened
with dis-tilled water do not develop a similar prefer-ence (Galef
et al., 1988) (Fig. 34-3). Thus,experience of carbon disulfide, a
natural con-stituent of rat breath, in conjunction with afood odor,
like experience of rat breath in con-junction with a food odor, is
sufficient to en-hance preference for the food.
SYNTHESIS
The breath of humans, like the breath of rats,contains trace
quantities of carbon disulfide.As would be expected on the
hypothesis thatexperience of food odors together with car-
100'"-~o
~
Uico c 80E,!CDas"0 CD 70
90
emCD-0"0CDa.casCD::E
60
50 - -- -- -------
Demonstrator Surrogate+ CS2
Group
Surrogate+ H20
Figure 34-3. Observer rats interacted with either an
anes-thetized demonstrator rat or a cloth "surrogate"
demon-strator. The demonstrator with which each observer
inter-acted had been powdered with either cinnamon-
orcocoa-flavored diet. Surrogates were moistened with eithera
dilute aqueous solution of CSz or an equal amount of dis-tilled
water. The figure indicates the mean percent of eachobserver's
tOtal intake that was the diet with which itsdemonstrator or
surrogate had been powdered. Error barsshow 1 SEM. (Data from Galef
et al. [1988].)
-
368
bon disulfide induces food preferences in rats,when a human
"demonstrator" eats a flavoredfood and breathes on a rat, the rat's
prefer-ence for the food that its human demonstra-tor ate is
markedly enhanced (Galef, 2001).
LIMITATIONS
Surprisingly, rats do not learn to avoid a foodby interacting
with a sick or an unconsciousdemonstrator that has eaten it. To the
con-trary, rats show an increased preference for afood that was
eaten by an ill conspecific withwhich they interacted (Galef et
a!., 1990).
Further, exposure to an odor in conjunc-tion with a conspecific
does not enhance thegeneral affinity of a rat for that odor;
exposureto an odor in a social context that profoundlyaffects food
preference has no effect on theodor preferences of rats in other
contexts. Forexample, rats that have interacted with a con-specific
that has eaten a cinnamon-flavoreddiet prefer cinnamon-flavored
food but showno enhancement of their preference for
cin-namon-scented nest materials or cinnamon-scented nest sites
(Galef and Iliffe, 1994). Suchfindings suggest that social
induction of foodpreference is a learning process evolved
specif-ically to facilitate foraging rather than otheractivities of
rats.
EXTENTIONS
Rats can use information concerning foodsthat other rats have
eaten in some interestingways. For example, after "observer" rats
hadan opportunity to learn where in a three-armmaze each of three
distinctively flavored foodswere to be found, we let each observer
rat in-teract briefly with a demonstrator rat that hadeaten one of
those three foods. Without anyspecific training, the observers went
directlyto the arm of the maze where they hadlearned that the food
that their demonstratorhad eaten was usually located (Galef and
Wig-more, 1983). Obviously, rats can integratetheir cognitive map
of food distribution with
DEFENSE AND SOCIAL BEHAVIOR
socially acquired information about the Cur-rent availability of
foods to increase the effi-ciency with which they forage.
APPLICATION TO STUDIES OFNERVOUS SYSTEM FUNCTION
Socially induced enhanced diet preferenceprovides an efficient
and reliable way to in-duce a learned appetitive behavior in rats
(ormice, gerbils, hamsters, voles or bats) that,like other types
oflearned behavior, can serveas a dependent variable in studies of
brainfunction. Neuroscientists have used the so-cially induced
change in food preference de-scribed here to study the effects of
manipula-tions of the neural substrate on learning andmemory
(Burton et al., 2000; Winocur et al.,2001; see Galef, 2002, for
further references).As one might expect, both direct and
geneticmanipulations of the nervous system affectsocial learning of
food preferences.
There are several advantages in using so-cially learned food
preference as a dependentmeasure in studies of brain function: (1)
learn-ing occurs in a single trial, (2) little or no skillis needed
to train subjects, (3) no specialequipment is needed to train
subjects, and (4)subjects need never be deprived or stressed.The
procedure for inducing social enhance-ment of food preference
consists of threestraightforward steps. First, a demonstratorrat is
placed on a feeding schedule and givenone of two distinctively
flavored foods to eat.Second, each demonstrator is placed
togetherwith an observer, and demonstrator and ob-syrver rats are
allowed to interact for 15 min-utes or longer. During this period
of interac-tion, observers have the opportunity to smellthe scented
food on the breath of their re-spective demonstrators. Last, each
observer isgiven a choice between the two distinctivelyflavored
foods that were offered to demon-strators in the first step (Galef,
2002) (Fig.34-4). In the third step, observers invariablyshow an
enhanced preference for whichever
-
Chapter 34. Social Learning
demonslr.tor
Stage I demonstrator
.~Stage 2
observer demonstrator
Stage 3 observer
Figure 34-4. Schematic of the three stages of an
experimentdemonstrating social influence on the diet preferences
ofobserver rats'. In stage I, each demonstrator rat ate one oftwo
distinctively flavored foods. (From Galef [2002].Reprinted with
pennission of John Wiley & Sons, Inc.)
flavored food was eaten by their respectivedemonstrators.
The effect is robust. Demonstrator andobserver can be male or
female, young or old,previously familiar or unfamiliar with one
an-other, and genetically related or unrelated toone another (Galef
et al., 1984). Demonstra-tors can ingest almost any scented liquid
orsolid before interacting with their observers.There can be a
delay of several hours betWeenwhen a demonstrator is fed and when
it in-teracts with its observer. Demonstrators canbe separated from
their observers by ahardware-cloth screen while they interact,and
interaction can take place in the homecage of demonstrator or
observer or in a neu-tral arena. There can be a delay of weeks
be-tween when demonstrator and observerinteract and when the
observer is tested. In-variably, if demonstrators that have
recentlyingested a distinctively flavored substance areplaced for a
few minutes together with ob-servers that are otherwise unfamiliar
with theflavor of the food that was eaten by their re-spective
demonstrators, the observers subse-quently show significant
enhancement oftheir relative intake of that food.
369
REFERENCES
Barnett SA (1958) Experiments on "neophobia" in wildand
laboratOry rats. British Journal of Psychology49: 195-201.
Beck M and Galef BG Jr (1989) Social influences on theselection
of protein-sufficient diet by Norway rats.Journal of Comparative
Psychology 103:132-139.
Burton s, Murphy D, Qureshi U, Suton P, O'Keefe J(2000) Combined
lesions of hippocampus andsubiculum do not produce deficits in
nonspatial so-ciallearning. Journal of Neuroscience
20:5468-5475.
Galef BG Jr (1977) Mechanisms for the social transmis-sion of
food preferences from adult to weanling rats.In: Learning
mechanisms in food selection (BarkerLM, Best M, Domjan M, eds.),
pp. 123-150. Waco,TX: Baylor University Press.
Galef BG Jr (1981) The development of olfactory controlof
feeding site selection in rat pups. Journal of Com-parative and
Physiological Psychology 95:615-{i22.
Galef BG Jr (1986) Social interaction modifies learnedaversions,
sodium appetite, and both palatabilityand handling-time induced
dietary preference in rats(Rattusnorvegicus). Journal of
Comparative Psy-chology 100:432-439.
Galef BG Jr (1988) Communication of information con-cerning
distant diets in a social, central-place forag-ing species (Rattus
norvegicus). In: Social learning:psychological and biological
perspectives (ZentallTR and GalefBGJr, eds.) pp. 119-140.
Hillsdale, NJ:Erlbaum.
Galef BG Jr (1992) The question of animal culture. Hu-man Nature
3:157-178.
GalefBG Jr (1996a) Social enhancement offood prefer-ences in
Norway rats. In: Social learning and imita-tion: the roots of
culture (Heyes CM and Galef BGJr, eds.) pp. 49-{i4. New York:
Academic Press.
GalefBG Jr (1996b) Social influences on food preferencesand
feeding behaviors of vertebrates. In: Why weeat what we eat
(Capaldi E, ed.) pp. 207-232. Wash-ington, D.C.: American
Psychological Association.
GalefBG Jr (2001) Analyses of social learning processesaffecting
animals' choices of foods and mates. Mex-icanJournal of Behavior
Analysis 27:145-164.
Galef BG Jr and Whiskin BE (2003) Socially transmittedfood
preferences can be used to study long-termmemory in rats. Learning
and Behavior 31:160-164.
Galef BG Jr and Allen C (1995) A new model systemfor studying
animal traditions. Animal Behaviour50:705-717.
Galef BG Jr and Beck M (1985) Aversive and attractivemarking of
tOxic and safe foods by Norway rats. Be-havioral and Neural Biology
43:298-310.
GalefBGJr and Buckley LL (1996) Use offoraging trailsby Norway
rats. Animal Behaviour 51:765-771.
-
370
Galef BG Jr and Clark MM (1971a) Parent-offspring in-teractions
determine time and place of first inges-tion of solid food by wild
rat pups. psychonomicScience 25:15-16.
GalefBG Jr and Clark MM (1971b) Social factors in thepoison
avoidance and feeding behavior of wild anddomesticated rat pups.
Journal of Comparative andPhysiological Psychology 25:341-357.
Galef BG Jr and Iliffe CP (1994) Social enhancement ofodor
preference in rats: is there something specialabout odors
associated with foods? Journal of Com-parative Psychology
108:266-273.
GalefBG Jr (2002) Social learning of food preferences inrodents:
rapid appetitive learning. Current Proto-cols in Neuroscience.
8.5DI-8.5D8.
Galef BG Jr, Kennett DJ, Wigmore SW (1984) Transferof
information concerning distant foods in rats: a ro-bust phenomenon.
Animal Learning and Behavior12:292-296.
Galef BG Jr, Marczinski CA, Murray KA, Whiskin EE(2001) Studies
of food stealing by young Norwayrats. Journal of Comparative
Psychology 115:16-21.
Galef BG Jr, Mason JR, Pretti G, Bean, NJ (1988) Car-bon
disulfide: a semiochemical mediating socially-induced diet choice
in rats. Physiology and Behav-iour 42:119-124.
Galef BG Jr, McQuoid LM, Whiskin EE (1990) Furtherevidence that
Norway rats do not socially transmitlearned aversions to toxic
baits. Animal Learningand Behavior 18:199-205.
Galef BG Jr and Sherry DF (1973) Mother's milk: amedium for
transmission of information aboutmother's diet. Journal of
Comparative and Physio-logical Psychology 83:374-378.
Galef BG Jr and Stein M (1985) Demonstrator influenceon observer
diet preference: analyses of critical so-
DEFENSE AND SOCIAL BEHAVIOR
cial interactions and olfactory signals. Animal Learn-ing and
Behavior 13:131-138.
GalefBG Jr, Whiskin EE, Bielavska E (1997) Interactionwith
demonstrator rats changes their observers' af-fective responses to
flavors. Journal of ComparativePsychology 111:393-398.
Galef BG Jr and Wigmore SW (1983) Transfer of infor-mation
concerning distant foods: a laboratory in-vestigation of the
information-centte" hypothesis.Animal Behaviour 31:748-758.
Garcia J, Ervin FR, Koelling RA (1966) Learning withprolonged
delay of reinforcement. PsychonomicScience 5:121-122.
Hepper PG (1988) Adaptive fetal learning: prenatal ex-posure to
garlic affects posmatal preference. AnimalBehaviour 36:935-936.
Smotherman WP (1982) Odor aversion learning by therat fetus.
Physiology and Behavior 29:769-77l.
Steiniger von F (1950) Beitrage zur Soziologie und son-stigen
Biologie der Wandetratte. Zeitschrift furTierpsychologie
7:356-379.
Strupp BJ and Levitsky DA (1984) Social transmission offood
preferences in adult hooded rats (RtIttusnorvegicus). Journal of
Comparative Psychology98:257-266.
Van Schaik CP, Ancrenaz M, Borgen G, Galdikas B, Sin-gleton I,
Suzuki A, Utami SS, Merill M (2003) Orang-utan cultures and the
evolution of material cultureimplications. Science 299:102-105.
Von Frisch K (1967) The dance language and orienta-tion of bees.
Cambridge, Mass.: Belknap Press.
Whiten A, GoodallJ, McGrew WC, Nishida T, ReynoldsV, Sugiyama Y,
Tutin CEG, Wrangham RW, BoeschC (1999) Culture in chimpanzees.
Nature 399:682~5.
Winocur G, McDona!d RM, Moscovitch M (2001). An-terograde and
retrograde amnesia in rats with largehippocampal lesions.
Hippocampus 11:18-26.
