The Auk 110(2):247-254, 1993
COMMON
CALLS
RAVENS
OF FOOD
ARE ATTRACTED
DISCOVERERS
BY APPEASEMENT
WHEN
ATTACKED
BERNDHEINRICH•, JOHN M. MARZLUFF
2, AND
COLLEEN S. MARZLUFF 2
•Department
of Zoology,Universityof Vermont,Burlington,
Vermont05405,USA;and
2Greenfalk
Consultants,
8210Gantz,Boise,
Idaho83709,USA
ABSTRACT.--When
adult territorial Common Ravens (Corvuscorax)aggressivelydefend
carcasses
from vagrantimmatures,bothdominantandsubordinate
immaturebirdsrespond
with beggingposturesand vocalizations.We demonstratedthroughexperimentsin an aviary
complexthat thesevocalizations
attractotherravensthat then learn of new feedingopportunities.Nearby free-rangingravensalsoare attractedto beggingvocalizations.Beggingis
most commonwhen six or fewer immaturesinteract with adults. Typically, fewer than five
birdsareattractedlocallyto beggingvocalizations,
a numberthat is not sufficientto account
for the largenumbersof ravensthat eventuallyassemble
at a carcass.
Mostof the recruitment
to carcasses,
therefore,cannotbe explainedby attractionto local vocalizationsalone.Received
21 October1991, accepted19 November1992.
THEEVOLUTIONARY
significanceof bird aggre-
gationshaslongbeenof interest(Pulliam 1973).
Numerous advantagesof flocking have been
suggested,including increasedvigilance (Caraco 1981), swamping of predatorsor competitors
(Wallaceand Temple 1987, Marzluff and Heinrich 1991), and diversion or protection from
predators(Kenward1978).However, despitethe
wealth of information and theory on the potential evolutionary significance or adaptiveness of bird aggregations,the proximate behavioral mechanisms causing birds to form
groupshave been little investigated.
There are numerous potential scenariosfor
how birds may aggregate.Solitary individuals
may incidentally aggregate by using the resource itself as a cue. Numbers
of birds could
then gradually increase simply because no
mechanismexiststo dispersethem. Group size
efits, they should evolve to mute or alter those
signalsthat potential followers could use.
If the actionsassociatedwith preexisting behaviors (flight patterns, vocalizations, etc.)
function as adequatesignals for attracting conspecifics,the gradual evolution of specificrecruitmentsignalswould depend, in part, on the
perceptive limitations of the receivers. Thus,
"passive"recruitment in highly perceptiveanimals would not always be functionally different from "active" recruitment involving such
obvioussignalslike recruitmentdancesand trail
pheromonesof social insects.
We previously defined recruitment on functional grounds as an attraction of others that
increasesthe signaller's fitness;proximate intent (consciousmotivation) of the signallerneed
not be considered (Heinrich and Marzluff 199I).
The loud "yell" vocalization that nonterritorial
could also increase as naive birds follow others
Common Ravens(Corvuscorax)give near food
alreadyat or on their way to the resource.When and which strongly attracts others (Heinrich
the advantagesof individuals that follow also 1988) is a casein point. The call is proximally
accrueto those already in the group, then the releasedby hunger and functionally related to
originally "passive"cuesthat are given by birds expressing high status which is important in
in the group and that the followers decipher establishing feeding rights (Heinrich and
should evolve to become amplified and/or to Marzluff 1991). By expressinghigh status,othbe more frequently given. A still further step ers are attracted to the status signalling and
in the evolution of group formation could in- then all are able to feed more and fight less.
volve the giving of specificcuesor calls(Elgar However, statussignalling at food theoretically
1986,Chapmanand Lefebvre 1990)if any of the could be accomplishedby gestureswithout voabove mechanismsof group formation are in- calizationsthat carry for 1 to 2 km. We here
adequatein generatinggroups.In the sameway, examine through field and aviary experiments
when the signallersincur more coststhan ben- whether or not "begging" or appeasementcalls
247
248
HEINRICH,MARZLUFF,
ANDMARZLUFF
that Common Ravensgive in addition to yells,
especiallywhen they are in the presenceof territorial adults defending food, can function as
attraction signals.
We assuredcomplete independence between subjects and between treatments, as emphasized by
Kroodsma(1989), in three ways: (1) A new group of
10 immatureswasusedfor eachreplicate.(2) Different
subordinateand dominant birds from each group of
10 were
METHODS
Subjects
andapparatus.--ImmatureCommon Ravens
were captured at carcasses
in the field as described
previously(Heinrich 1988)and housedin a largeaviary complex(seeHeinrich and Marzluff 1991).The
complex consistedof a main aviary (ca. 40 x 70 m)
with two sidearms(ca.90 m long) leading to smaller
aviaries (25 m diameter). One of the smaller aviaries
containeda mated territorial pair of adults that had
also been captured in the wild at the site. Unless
otherwise indicated, the immature ravens were kept
in the main aviary during the experiments.Accessto
the peripheral side arms and their associatedaviaries
was controlled by raising or lowering gatesby guy
wires operated from an observation hut. Each side
arm containedtwo gates,one at the arm's interface
with the main aviary and another about 5 m down
the arm. All observationswere made through oneway glassfrom windows in an observationhut locatedto provide visual coverageof mostof the aviary
complex.
Fivegroupsof immatureswerecapturedin the field
for the experiments.They were marked with uniquely numberedand coloredpatagialtags(for detailsof
study area and capture and marking techniques,see
Heinrich 1988). The birds adapted quickly (within
severalhours) to their tags and their new surroundings. They fed togetherand roostedas a communal
groupin a coveredshed.We determineddominance
relationshipsamongthe immaturesfor a week or two
during interactionsat food in the main aviary as previously described(Marzluff and Heinrich 1991).The
two wild-caught adults raised a brood of young to
maturity in their side aviary during the winter and
spring prior to when most of the experimentswere
conducted.
Experimentalprotocol.--The experiment testing
whether
or not ravens were attracted
to the behavior
[Auk, Vol. 110
allowed
to discover
defended
and
unde-
fended food. (3) Different recordingsof immatures
beggingwere usedas playbackstimuli.
We followed the sameprotocolto initiate eachphase
of the experiment.First, prior to any experiment,the
immatureswere not fed for three days.Nevertheless,
we assumethey fed on at least somecachedfood. To
initiate an experiment,the first of two doorsclosing
eachperipheralarm of the aviarycomplexwasopened.
This created a short (5-7 m deep) vestibule in each
arm that the birds from the main aviary could enter.
The experimenter, however, then walked down the
full length of each arm of the aviary to place food
(carriedin bags)near the adult and undefendedaviaries. Neither pile of meat was visible to the birds
anywhere from the main aviary or from the vestibule
then availableto the birds.For playbackexperiments,
the speakerinstead of food was at this time placed
into the undefended aviary, and the experimenter
walked down to the defended aviary.
The
immatures
were
next
watched
for a 30-min
acclimationperiod to determine a pretreatment num-
ber of birdsenteringthe now-openedvestibulesleading to the sideaviaries.After the 30-min period,the
experimenteragain walked down the entire arm of
each aviary to move the food directly into each peripheral aviary and to releasea "discoverer"into each
aviary provided with the food bonanzas.At this time
the last (third) gate near the discovererand the territorial pair also was opened, so that the birds were
free to interact and to move up and down the aviary
arms. The next 30 min constitutedthe experimental
or treatmentperiod (i.e. when the controland stimuli
were applied). The number of immatures(of the 10)
entering the two vestibulesagainwas tabulated.
We summarized the subjects' responsesto each
treatmentby quantifyingtheir preferencefor the right
vestibule (that leading to food defended by the territorial pair) relative to the left vestibule by subtract-
ing the number of birds entering the left vestibule
from the numberenteringthe right. Thus,we created
times with 10 different immatures (nonbreeders of
a scaleof -10 to 10, with -10 indicating exclusive
unspecifiedagebut at leastsix monthspostfledging) preferencefor the left (undefended)side,0 indicating
comprisingeachgroup.Eachreplicatehadfour phases no preferencefor either side, and 10 indicating exthat were employed in random order with the stip- clusivepreferencefor the right (defended)side.The
ulation that eachphaseoccurredfirst and last at least samescaleswere calculatedfor both the pretreatment
once. In one phase of the experiment, no stimulus and treatment phases.We comparedthe preference
was applied. In the secondand third phases,we had for the right sidein treatmentversusacclimationpeeither dominant or subordinate birds discover food
riods for each type of stimulus using a Wilcoxen
(40-60 kg of meat)in eachaviary.In the fourth phase matched-pairstest (Wilkinson 1989)to find out if the
we broadcastedthe appeasementvocalizations("beg- immatureswould show:(1) a preferencefor the right
ging") of the discoverers
from the undefendedaviary or defended side in the experimental phase when
so as to be only weakly audible to us from the main discovererswere present;(2) a preferencefor the left
aviary.
side during playbacks;and (3) no preferencewhen
of discoverersof defended food was replicated five
April1993]
Attraction
ofAppeasement
Calls
no stimuluswaspresented.In all casesone-tailedtests
were done to test the predictedresponse.
We createda secondresponsevariable by subtracting the valuesobtainedfor the pretreatmentfrom the
treatment phasesto create a scale from -20 to 20,
with - 20 indicatingcompletepreferencefor the right
sideduring pretreatmentversusleft sideduring treatment (avoidanceof defendedarm during treatment),
0 indicating no preferencefor either side in pretreatment relative to treatment, and 20 indicating complete preferencefor the left sideduring pretreatment
and the right side during treatment(preferencefor
defendedarmduring treatment).Theseresponses
were
analyzed in a one-way repeated-measuresANOVA
(Wilkinson 1989),where eachof the four phaseswere
repeated measuresof the group's responses.
Observations
offree-ranging
birds.--Weobservedfreeranging ravens at naturally occurring carcassesand
at carcasses
and meat pileswe placedin the field from
1989 through 1991. We watched ravens at these food
bonanzasfrom blinds as previously described(Marzluff and Heinrich 1991) and recorded the number of
ravensat thesebonanzasperiodicallythroughoutthe
day, the presenceor absenceof defensiveadults,and
the repertoire of vocalizationsused by adults and
vagrant immatures.
We also assessed the attractive
nature
of vocaliza-
tions to free-ranging birds by counting the number
249
ging, gave kaws on two occasions
and yells on
one occasion.Begging calls were given more
often in the presenceof adults than in their
absence (X 2 = 20.0, 1 df, P < 0.001).
One-half of the juveniles in the aviary with
the adults defending food were able to feed
briefly despitethe adult defense.Six juveniles
ate alone at undefended
food. The other
four
did not go down to feed during the 0.5-h observationperiods.
The relativepreferencefor the right sideduring the experimental phase differed significantly acrossthe four stimuli (repeated-measuresANOVA, F = 9.5, df = 3 and 12, P = 0.001).
Contrastsof the responsebetween each phase
indicated that this was primarily due to the
strong preferencefor the left side during the
playbackphase.Resultsfrom Tukey'sHSD tests
(Wilkinson 1989) were significant when comparingpreferenceduring playbackphaseto: (a)
preference with no stimulus (F = 5.1, df = 1
and 4, P = 0.05),(b) preferencewith subordinate
discoverers(F = 49.9, df = 1 and 4, P = 0.001),
and (c) preferencewith dominant discoverers
(F = 114.4, df = 1 and 4, P < 0.001). None of
our capturedbirds fed. These"outsiders"were counted during 1- to 10-h watches on 218 days (from 1
October through 30 April of 1989, 1990, 1991). We
analyzedthe variation in the number of outsidersper
day with a three-factor ANOVA (Wilkinson 1989).
the contrastsbetween phaseswith discoverers
and no stimuluswere significant(Tukey'sHSD,
P > 0.20). However, this lack of significanceis
largely due to one replicate, where 10 birds
wandered into the right vestibulealmostat the
very end of the 30-rainacclimationperiod (Ta-
The factorswere: (1) numberof birdsin aviary (10 or
ble 1).
of free-rangingravensthat cameto the aviary while
20);(2)vocalactivityof birdsin aviary(quietandnot
The preferencefor the right vestibule in experimental and control responsesdiffered between the four stimuli as predicted (Table 1).
feedingsiteknown within 2 km, or a groupfeeding When no stimulus was present the number of
within 2 km). Thesewild birdsusuallyfed at carcasses birds entering the right or left vestibuleswas
we placed in the field and regularly monitored. It is variable and not significantly different (Wilunlikely that a group was feeding within 2 km of the coxon T -- 6, P > 0.05). In contrast,the right
aviary unknown to us.
side (with defending adults) was preferred in
everytrial when discoverers
were present(Wileating, occasionallyfighting and calling, often vocalizing and fighting while feeding); and (3) proximity of a feeding site of wild birds to aviary (no
coxon T = 0, P < 0.05 for dominant and subRESULTS
Aviaryexperiments.--Discoverers
behaveddif-
ferently when they encountereddefendedversus undefended
food. Each of the 10 iramatures
ordinate discoverers).We observedthat, assoon
as the discoverersstarted to beg, the birds in
the main aviary startedto move in the direction
of the sounds. Both dominant
and subordinate
that discovereddefended food gave begging juvenilesbeggedin the presenceof the adults.
calls in submissivepostureas the adults at- On the average,subordinatediscoverersattracttacked them. Discoverers of undefended food
ed 3.2 _+SE of 0.97 more birds in the experiwere silent on seven occasions and uttered a
mental versuscontrol period, and dominants
0.86(Fig. 1). In contrast,in every
mix of kaw, yell and trill vocalizationson the attracted5.2 +__
other three occasions. Discoverers of defended
playbacktrial of beggingcalls,the birds preffoodswere never silentand, in addition to beg- erentially entered the left vestibule (the one
250
HEINRICH,MARZLUFF,
ANDMARZLUFF
[Auk, Vol. 110
TABLE
1. Numbersof juvenilesenteringright versusleft vestibuleof aviaryduring acclimation(Acclimation)
and experimental(Experiment)portion of four phasesof experimentin five replicates.Territorial adults
defendedright aviary from discoverers.Speakerplacedin left wing during playbackexperiments.
Replicate
1
Phase of
experiment
2
3
Right
Left
Right
Left
0
0
3
!
!0
0
2
4
Right
4
Left
Right
!
0
0
0
5
Left
Right
Left
0
0
0
0
0
2
0
0
No stimulus
Acclimation
Experiment
Subordinate
discoverers
Acclimation
0
1
2
1
0
0
2
0
1
Experiment
6
0
5
2
6
4
7
2
4
2
Dominant
discoverers
Acclimation
2
0
0
0
0
0
0
!
0
4
Experiment
5
0
5
!
6
0
7
0
3
2
Acclimation
0
0
1
1
1
0
1
2
1
Experiment
2
6
2
6
0
2
!
6
0
Playback
with the speaker);on average,4.0 + 0.30 more
birds entered the left side relative to the right
(Wilcoxon T = 0, P < 0.05). Thus, in all three
experimentaltestswith stimuli, the birds preferentiallyenteredthe sideaviaryassociated
with
soundsravens make when attacked by dominant territory owners.
As demonstrated above, by following vocalizations of discoverers, naive birds located new
6
5
food sourceswithin the right aviary containing
the adultswhere they had never been before.
Field observations.--The
behavior
of discov-
erersin the aviary was confirmedin the field.
One to sixjuvenileswere seenat food defended
by adults on 26 occasionsand at food not defended by adults 24 times. Juvenilesbeggedin
the presenceof adults on 21 of 26 times, but
only beggedin 2 of 24 times when no adults
were in the area. The occurrenceof begging
calls in the field was significantly associated
with the presenceof adults (X2 = 26.4, df = 1,
P < 0.001).
P < o05
i
o
a
asdiscoverersat 21 carcasses
(single birds were
releasedat 16 carcasses,
and groups of five individuals [with one radiotagged]were released
at 5 carcasses).
In one case,the discovererjoined
2
4
5
No
Stimulus
Subordinate
Discoverer
Dominant
Discoverer
Playback
of Begs
Fig. 1. Number of birds entering right vs. left ves-
tibule of aviary within 30 min. Right vestibule contained territorial pair of adults(and food) in a second
aviary at the end. Left vestibuleconnectedto aviary
that containedfoodonly. "Discoverers"
refersto birds
placed in adult aviary and that made appeasement
begsto adults.Playbacksof appeasementbegsmade
from left (unoccupied)aviary. Birds entered vestibules leading to other aviaries in response to appeasementbegsmadeby both subordinateand dominant
birds.
The begging calls discoverersmake in the
presenceof adults also attract others into the
area. We released 21 marked juveniles (with
patagialwing tagsand with radio transmitters)
a nearbycommunalrooston the evening of its
release and recruited
roost mates to the bonanza. In 11 cases the discoverers remained in the
area (within 1 km) and begged when adults
approachedthem.In 10of these11 cases,other
juveniles were attractedduring the subsequent
four days.In nine casesthe discoverersleft the
area (no radio signalswithin at least 10 km) or
did not beg and, in seven of these nine cases,
no juvenile was attractedto the carcassduring
the subsequentfour days. Juveniles were significantly more likely to be attractedwhen dis-
April1993]
Attraction
ofAppeasement
Calls
251
55
A. Aviary
Other ravens feeding ;learby
No ravens nearby
ioo
5O
]
}
•o
,•
3o
40
]
o
10
B.
20
Field
Ioo
80
60
1
Number
40
1
2
6
of immatures
7-9
>
20
feeding with adult
3
4
5
of ravens
6
7
8
9
attracted
l0
it
12 13
to aviary
Fig. 3. Frequencyof different numbersof outsiders attractedto aviary while 10 to 20 juvenileswere
feeding inside. Filled bars refer to instanceswhen
feeding crowdswere in vicinity (< 2 kin) outsideavi-
20
Number
2
pair
ary.
Fig. 2. Occurrenceof begging in (A) aviary and
(B) field by immaturesas function of their numbers
birdswere outsidethe aviary at one time while
when feeding (B) or confined with and/or feeding
with adult pair (in A). Occurrenceof begging reflects our birdswere feedinginside (Fig. 3). The only
significant factor influencing the number of
the aggressionof adults toward newcomers.
birds gathered outsideour aviary was whether
or not a crowd was already at a nearby food
covetersbeggedthan when they did not (X 2 =
source(F = 19.1, df = 1 and 208, P < 0.001). On
7.1 with Yate's correction, df = 1, P < 0.001).
average,3.7 + 0.27 birds were attractedwhen
Resultsduring experiments when discoverers
did not beg and when discoverersleft the area
were lumped in the previous analysis to increasesamplesize. However, both reactionsby
discovererswere equally ineffective at attracting other ravens(ravens were attractedin one
of five caseswhen the discovererstayedbut did
not beg, and in one of four caseswhen the
birds were near, in contrast to 1.7 -+ 0.24 when
no crowd was known to be nearby.On 9 of 10
days during which 10 or more birds came to
our aviary,we alsowereawareof anotherfeeding site nearbywhere a raven crowdwas gathered.
We observeda similar pattern of attractionto
food outsideof the aviary. Rarelydid numbers
discoverer left the release site).
of birds at a carcassin nature increaseby more
Beggingsignalsthe locationof defendedfoods than 10 throughoutthe courseof the day (Fig.
when one or a few immaturestry to feed with 4). For 9 of 11 caseswhere 10 or more birds
adults. Adult defense is intense at this time and
were attractedthroughoutthe day, we knew of
beggingwasobservedmore commonlyin small another feeding site within 2 km. More birds
groupsthan in larger groupsof immatures,both were attracted to carcasseswhen the feeding
in the aviary and in the field (Fig. 2). As group groupsalreadythereweresmallthan when they
size increases,adult defense declines (Marzluff
were larger (groups of 1-5 birds, œ= 4.16 _
and Heinrich 1991) and, accordingly,begging 0.92, n = 50; groupsof 6-10, œ= 2.78 _ 1.33,n
becomeslessfrequent.
= 23; groups > 10, œ= 2.64 + 0.93, n = 11). As
Few birds were attracted to the aviary com- in the aviary, the trend for more birds to be
plex. It was common for only a single or no attractedto small groupsis correlatedwith the
birds to visit the aviary, but up to 13 visitors increasedfrequency of begging within small
were observed.On the average,only 2.6 + 0.18 groupswhen adultsare very defensive(Fig. 3).
252
•
HEINRICH,
MARZLUFF,
ANDMARZLUFF
;
[Auk,Vol. 110
bow, fluff their feathers,spreadand quiver their
wings in a submissivedisplay (Heinrich !988,
!989), and give begging vocalizations.The second major socialcontestsat feeding crowds are
among the immaturesthemselvesfor accessto
oftenlimited feedingspots.The mostdominant
among the immaturesgainsthe choicefeeding
spot,and thesebirds signal their dominanceby
yell vocalizations(Heinrich and Marzluff ! 99!),
which are one of the cuesusedby other hungry
ravens in the vicinity (<2 km) to locate the
B. 6 10 birds at start of day
contested food (Heinrich !988, 1989).
v-
We sought to answer two questions:(!) Do
the beg vocalizationsalsoserveas an attractant
cue as doesthe yell? (2) Can local cues(suchas
yells, begs,and all other potential sortsof local
C. > 10 birds at start of day
Number
of ravens
attracted
during
day
Fig. 4. Increase in numbers of Common Ravens
throughoutday at defendedcarcasses
in field involving (A) 1-5, (B) 6-10, and (C) >10 birds at the start
of day. Generally, numbers were stable throughout
day. Smaller groupstended to attract more additional
birdsthan largerfeeding groups.Filled barsrepresent
instanceswhen feeding crowdswere in vicinity (<2
kin) at other food bonanzas.
Common Ravensin western Maine forage by
flying singly or in pairs over the forestbut in
the winter they commonly aggregate(Mylne
! 96!) in large numbersat animal carcasses,
provided the carcasses
are openedfor them (Heinrich !988, 1989). The immatures consist primarily of vagrant birds that require a crowd of
some8 to !0 individuals before accessis gained
to the food bonanzasthat, otherwise,are highly
defended by the resident territorial adults
!988, !989, Marzluff
of sufficiently large crowds to neutralize the
defensesof the territorial adult pair?
Our experimentsin the aviary indicate that
the begs given during agonistic encountersdo
attract hungry immatures.The encountersoccur primarily at food worth defending. Thus,
the calls then given are a reliable indicator of
the presenceof a food bonanza.In our experiments in the aviary, immatures following such
begswere led to new food. Similarly, our field
results
DISCUSSION
(Heinrich
enhancement) alone account for the attraction
and Heinrich
!99!). Only the immatures attract others, and
the mechanism
of crowd formation
involves
vocalizations
near the food and recruit-
both
ment from communal nocturnal roosts (Hein-
rich !988, !989, Marzluff et al. in prep.).
show
that ravens
are attracted
to new
foodby suchcallsin the wild. When discoverers
feed quietly at a carcassor leave the area other
vagrant ravens rarely find the bonanza. However,when discovererscower,beg and flee from
defensive adults, recruits are quickly assembled.
Both in the aviary and in the field (Fig. 2),
the probabilityof beggingdecreasedgreatlyas
the numbers of immatures feeding with the
adults increased.For example,begging always
occurredwith only one immature present,but
when about 20 individuals were present,then
begging occurredat only 20 to 25% of the observation,which reflectsa very substantialreduction in beggingper individual. The reduction in begging in the presenceof a defending
adult pair was probably due to the reduction in
Feeding crowds at deer or moose carcasses aggressionby the adults, which also decreases
commonlyrange up to 20 and 60 birds, respec- as crowd size increases (Marzluff and Heinrich
1991).
tively, and two major interactionsoccurwithin
suchfeedingcrowdsat defendedcarcasses.
First,
Despitethe great reduction in the beg signal
at least before a crowd has assembled, the adults
at larger group sizes,the number of birds atattempt to repel all other birds in the vicinity tracted to carcassesthroughout any one day
bothby chasingthem in the air and by attacking when the begs (and many other signals) are
them when they are perchednear the bait on given decreasesonly marginally, from a mean
the ground or on trees.The attackedimmatures of 4.2 at one to five birds at the start of the day
April 1993]
to 2.6 when
Attraction
ofAppeasement
Calls
more than 10 birds were at the start
of the day (Fig. 4). That is, a decreaseof about
253
adults, and since they are not given at food
bonanzas in the absence of adults, it seems like-
55% (from about 80 to 35%) in the incidence of
ly that they function proximally in reducing
aggression.Nevertheless,the attractionof even
4.2 to 2.6 birds) in the number of birds attracted. oneothernearbyravencouldbe important,even
In only seven instancesdid we observe 10 or if locally attractedbirds are hardly ever likely
more birds attractedthroughout the day and, to be numerous enough to overpower the terin five of these, we knew of another nearby ritory holders. The initial vagrant discoverer
foodbonanzafrom which they couldhavecome. that is attackedmay or may not have accessto
beggingresultedin a decrease
of only 39%(from
The above results from
the field do not dis-
a roost to recruit others, but by attracting even
tinguishbetweennumbersof birdsattractedby one bird that does have access to a roost, the
begsversusthoseattractedby someother cue(s). local attraction in effect could draw in a distant
The lack of correspondencebetween numbers crowd (Marzluff et al. in prep.) that then can
of birdsattractedversusamountof beggingcalls gain accessto the bonanza(Marzluff and Heincould indicate either that a few begsare alone rich 1991). We observedsingle individuals resufficient to attract the local birds and/or, since
leased near baits (hence, out of contact with
the incidenceof yelling increasesrather than roosts)to remain in the bait vicinity for days.
decreasesas competition at the bait increases, Eventually, all of these baits were eaten by
this secondsignal may compensateas the begs crowdsof birds, despitetheir defenseby terdecrease.
ritorial pairs.
Throughout three winters we seldom saw
Since the begging attractsothers who could
more than three to four ravens attracted near
bring in a crowd, immaturesrespondingwith
the aviary complexcontainingboth captivebirds begs, in effect, exert a penalty on the adults'
and exposedmeatpiles.There were usuallydif- aggression.Therefore, if the adults could be
ferent meat piles in different portions of the certain that an intruder would not recruit then
aviary. Furthermore,in the main portion of the it should be advantageousfor them to tolerate
aviary there was no tree cover,making both the it. However, if any of the attracted birds ultibirdsand the meathighly visible.Birdspoten- matelybringin a crowdfroma roost,theyshould
tially couldhavebeenattractedto the aviaryby be vigorously repelled despite the small risk
callsof the inhabitantsand by any of a variety involved of altering othersduring the eviction
of other possible cues. However, since there process.Since the adults have no assurancethat
were seldom more than 8 to 10 birds around
a stranger will not recruit, perhaps the best
the aviary, the local cuesthat the inhabitants strategy for an adult is to always attempt an
were restrictedto giving were not sufficientto immediate eviction, but to become tolerant if
attract the crowds observed at carcassesand/or
the vagrant shows itself to be very dominant
meat piles located nearby in the forest. Fur- and resistant to eviction. Indeed, we find (unthermore, our results also indicate that when
publ. data) that the subordinate birds are the
birds seemeat,but are unable to reachit, they
first to leave feeding aggregationswhere they
are challenged.
Thereis greaterrelativecostto sharinga small
rather than a large carcasswith a crowd (Marz-
do not recruit. When
more than nine birds were
attractedto the vicinity of the aviary,we always
knew of anothernearbyfeedingbonanzafrom
which thesebirdsprobablycame,because
birds
in any one area move freely between different
locally available food bonanzas(Heinrich 1988,
1989).As indicated for previousstudies(Heinrich 1988,1989),the observationsreportedhere
support the idea that cuesgiven near the bait
act only locally, and that the large numbersof
ravensassembledat carcasses
can only be explained by recruitment from a nocturnal roost
(Marzluff et al. in prep.).
Given that the begs seldom result in the attraction of sufficient numbers of immatures to
overcome the defenses of the resident territorial
luff and Heinrich 1991). Therefore, a territorial
bird
should
have
more
confidence
that
a va-
grant would not recruit to a small carcassand,
hence, it could be more tolerant of it. However,
the greater cost of sharing a small carcassapplies to the adult and the vagrant as well. The
adult should not share a small food item it can
consumeitself. Thus,the oppositeeffectsof carcass size could
cancel out.
ACKNOWLEDGMENTS
The work was supported by NSF Grant BNS8819705.
254
HEINRICH,
M,•mZLUFL
,•,NDM,•,RZLUFF
LITERATURE CITED
KFaqW,•XD,R.E.
[Auk,Vol. 110
1978. Hawks and doves: Factors af-
fectingsuccess
and selectionin Goshawkattacks
CARACO,
T. 1981. Energybudgets,risk and foraging
on Woodpigeons.J. Anim. Ecol. 47:449-460.
preferencesin Dark-eyedJuncos(Juncohyemalis). KROODSMA,
D.E. 1989. SuggestedexperimentaldeBehav. Ecol. Sociobiol. 8:213-217.
signsfor songplaybacks.Anim. Behav.37:600CHAPMAN,
C. A., ANDL. LEFESvRE.
1990. Manipulat609.
ing foraginggroupsize:Spidermonkeyfood calls MARZLUFI%
J., AND B. HEINRICH. 1991. Foraging by
at fruiting trees.Anim. Behav.39:891-896.
Common Ravensin the presenceor absenceof
ELGAR,
M. A. 1986. House Sparrowsestablishforterritoryholders:An experimentalanalysisof soaging flocks by giving chirrup calls if the recial foraging. Anim. Behav.42:755-770.
source is divisible.
Anim.
Behav. 34:169-174.
HEINRICH,B. 1988. Winter foraging at carcasses
by
three sympatric corvids, with emphasison recruitmentby the Raven,Corvuscorax.Behar.Ecol.
Sociobiol.
23:141-156.
HEINRICH, B. 1989. Ravens in winter. Summit Books,
Simon & Schuster, New York.
HEINRICH, B., AND J. MARZLUI•I•. 1991. Do Common
Ravensyell becausethey want to attractothers?
Behar.
Ecol. Sociobiol.
28:13-21.
MYLNE,C.K. 1961. Large flocksof Ravensat food.
Br. Birds 54:206-207.
PULLIAM,H. R. 1973. On the advantageof flocking.
J. Theor. Biol. 38:419-422.
WALLACE,
M.P., AND S. A. TEMPLE.1987. Competitive interactionswithin and between a guild of
avian scavengers.Auk 104:290-295.
WILKINSON,
L. 1989. SYSTAT:The systemfor statistics. SYSTAT, Inc., Evanston, Illinois.