ORNITOLOGIA NEOTROPICAL 8: 133-14@ The Neotropical Ornithological Society

BEHAVIOR OF INDIVIDUALS AND SOCIAL INTERACTIONSOF THE RED-FRONTED MACAW ARA RUBROGENYS

IN THE WILD DURING THE MIDDAY REST

Elin Pitter & Mette Bohn Christiansen

Zoological Museum, University of Copenhagen, Universitetsparken 15,DK-2100 Copenhagen 0, Denmark.

Abstract. Behavior of the Red-fronted Macaw Ara rubrogenys was studied from September 1991 to March 1992in south-central Bolivia. The macaws spent many hours at midday in groups of 2-30 individuals in quiet gorgesand valleys. The birds always sat in pairs and spent most of the time sitting/resting or autopreening. Social inter-actions such as allopreening and playing/fighting took up 11-18% of their time. Social activities appeared toincrease up to the breeding period. Certain interactions such as allopreening, copulation and courtship feedingpresumably serve to maintain the pair-bond. AII types of interactions between pairs were rare. Individuals in theflock normally communicated by vocalizations and immediate copying of behavior, vocalization as well as move-ment, of one individual by another (social facilitation). Aggressive interaction between the pairs was exceptionallyrare, and always sitting in pairs as well as performing allopreening seemed to keep the aggression at a low level.Accepted 24 Apri11997.

Key words: Redfronted MacaU\ Ara rubrogenys, behavio1; Bolivia, non-social interactions, social interactions.

INTRODUCTION

The Red-fronted Macaw is endemic to the valleysand montane basins within the eastern Cordil-lera in Bolivia (Lanning 1982, 1991; Forshaw1989; Fjeldsa & Krabbe 1990; Pitter & Christian-sen 1995). It is threatened by habitat destruction,trapping for the pet trade and persecution as apest on crops. The species is highly social (Bous-sekey et al. 1991) and during the day it was seenin flocks of 2-90 individuals. The flock sizediffered depending on time of year, area and typeof activity (Pitter & Christiansen 1995).

According to previous studies parrots showfew stereotyped sequences or distinct posturescompared to many other bird groups aonFjeldsa, pers. comm.). However, extremely fewpublished studies have been based on obser-vations of the behavior of wild parrots.

The present paper focuses on the behaviorduring the long midday rest where the macawsgather in clusters of trees. The most commonlyobserved types of behavior performed duringthese hours are described as well as the socialinteractions within pairs and within the group.

STUDY AREA

The two main study areas were Camos, located15 km south of Puente Arce along Río Chico(tributary to Río Grande) (between 18 o 37~45'S

and 65 °08~10'W) and Sucusuma, located 8 kmnorth of Torotoro along Río Caine (between18 °02~08'S and 65 ° 38~46'W). The numerous

valleys in the area are arid to semi-arid, mesother-mic valleys occupied by xerophytic thorn wood-land (Hueck 1966, Solomon 1989). The wood-land vegetation is composed of genera such as

Schinus, Prosopis, Tipuana, Aspidosperma, Loxop-terygium, Schinopsis and several species of cacti.The observations were made in a valley in Sucu-suma and in a gorge on Río Chico.

The valley in Sucusuma was located in thevicinity of a large agricultural area, which wasthe main feeding area for the macaws at that timeof the year. A stream ran through the valley, andthe bottom of the valley was quite narrow(25-30 m) between 10 m tall slopes. The vegeta-tion in the valley and on the hillsides was scruband scattered 7-10 m tall trees. Most of the treeshad a sparse foliage, but provided sufficient shade

33

PITTER & CHRISnANSEJ

for the macaws during the hottest hours of theday. Humans passed through the valley only

rarely.The gorge in Río Chico was 30 m long, 4 m

wide at the bottom and with 10-12 m highslanting cliffs on both sides. The upper 3 m ofthe cliffs were vertical. The vegetation in thegorge was mostly grasses and bushes, but therewere also a few 5-8 m tall trees on the hillsides,most of them on the edge of the gorge. Themacaws prospected three holes in the verticalcliffs.

same time) and playing/fighting. Using the scan-ning method a behavior is recorded at a certainmoment and therefore no distinction has beenmade between sitting and resting or betweenplaying and fighting as these behaviors oftenpassed into each other.

The number of birds covered by each scann-ing varied from 2 to 20 individuals. Duringscanning the behavior of the individuals wasnoted in the same order for as long as possible,but it was difficult to identify individuals froma distance, and as they moved around the orderchanged. Therefore the results reflect the beha-vior of a group rather than of the individual birds.Behavioral types with a duration of less than fiveminutes are classified in the group "other". Thisgroup included locomotion (moving on the samebranch, moving to another branch in the sametree, flying from one tree to another), acrobatics,chewing on different objects, copulation, court-ship feeding. The frequencies of these shorteractivities, noted during the scanning, are com-

pared.

METHODS

The study was conducted from September 1991to March 1992.

Birds always sat together in pairs during themidday rest. Judging from the behavior weassume that the pairs could comprise matedpairs, homosexual pairs and temporary pairs ofimmatures, either siblings or unrelated birds.The behavior of juveniles is treated in Christian-sen & Pitter, 1993b.

Observations were made in Sucusuma duringSeptember and October. The groups observedincluded many juveniles as well as many imma-tures. The fieldwork in Río Chico took placefrom December to March and included the startof the breeding period. In Rio Chico three pairsshowed interest in nest holes, and one pair laideggs during the last days of the study. Onlyonemember of the flock was juvenile. The rest hadadult plumage, but judging from behavioralcharacteristics, many were immatures.

The macaws were studied from a distance of20-30 m with telescope, and the Scan Samplingmethod (Altmann 1974) was employed. In totalwe spent 68 hs scanning (4151 scan observations)in Sucusuma and 69 hs (2666 scan observations)in Rio Chico. When Scan Sampling we noted thecurrent behavior of each individual at five-minute intervals. Behavioral types with a dura-tion of five or more minutes included sitting/resting, autopreening, interactions between indi-viduals concerning allopreening the generalplumage, cloacal region and beak (nonsimultane-ous allopreening refers to one bird preeninganother. This mayor may not be reciprocated bythe preened individual. Simultaneous allopreen-ing refers to two birds preening each other at the

RESULTS

The midday rest of the macaws during the hotmidday hours in Sucusuma normally lasted for5-7 hs, from the time the macaws arrived in thevalley after the morning feeding until they leftfor the afternoon feeding. In Río Chico themacaws flew in and out several times during themidday rest and were often seen in the gorgebefore the first feeding session in the morning(Pitter & Christiansen 1995). The compositionof the flock changed during the day: some birdsstayed only a few hours, while others stayedmost of the day.

Usually the macaws gathered in groups. InSucusuma we recorded up to 18 macaws and inRío Chico 30 macaws. Many birds usually restedin the same tre~ In Sucusuma the mean groupsize in one tree was seven birds (n = 58, range

2-18) (Pitter & Christiansen 1995).All adult and immature individuals in a

resting flock usually remained together in pairsalmost constantly. They changed between sittingtogether touching each other, sitting a few centi-meters apart and more rarely, 20-50 cm apart.Once in a while, though, they separated, movingto other branches in the same tree or another

134

BEHAVIOR OF ARA RUBROGENYS

tree, but after a short while they rejoined.Separating was mostly seen after social inter-actions such as allopreening, playing/fighting or

copulating.During the midday rest, behavior varied de-

pending on number of birds and composition ofthe group (number of juveniles, immatures andadults) resting in the same tree. The larger thegroup, the more active it was, and especially theactivity of a group of noisy immatures couldspread through the entire group.

TIME SPENT ON THE MAIN TYPES

OF BEHAVIOR

The macaws were active in 23-42% of the timeduring the midday rest, performing auto- andallopreening and playing/fighting (Fig. 1). Theyspent 11-18% of the time on social interactions.However, the predominant behavior during themidday rest was sitting/resting. The birds in RíoChico were much more active than in Sucusumaand spent more time on both kinds of socialinteractions.

The group "other" was larger in Río Chicothan in Sucusuma and indicated that the birds inRío Chico were more restless flying from tree to

tree or circling in the gorge.

Autopreening. In Río Chico the macaws oftenpreened before the first feeding session but themain preening period was as in Sucusuma,during the midday rest. Autopreening was thesecond most commonly observed behavior (Fig.1) and was seen throughout the entire middayrest. The macaw preened itself on most parts ofthe plumage, changing frequently between thedifferent parts. The bird nibbled the feathersunder the raised wing by putting the head belowthe wing from either the front or from behind.While preening the tail feathers the bird bent thehead 180 o to the tail and took one feather at a

time in the beak and pulled it between themandibles in a nibbling way. Sometimes the birdfanned its tail feathers, facilitating the preening

NON-SOCIAL INTERACTIONS,BEHAVIOR OF INDIVmUAL BIRDS

Individual birds spent most of the time sitting/resting and autopreening (Fig. 1), but activities ofshorter duration, such as different kinds of loco-motion, chewing on objects, and acrobatics wereoften seen, as well.

Sitting/resting. Sitting/resting was the predomi-nant behavior (Fig. 1). The macaws sat quietlyin a relaxed and sometimes crouched posture.The behavior could last up to several hoursor only a few seconds, interspersed with otherbehavior. Particularly in the middle of themidday rest, during the hottest hours, theyrested for many hours with only few and shortintervals of other behavior. During this periodthe eyes often were closed and they put the headbackwards over the shoulder tucking the beakinto the plumage of the back and often fluffingthe plumage.

135

PITTER & CHRISTIANSEN

Within the tree there was a clear tendency totravel short distances by climbing, sidling orwalking rather than flying. When the macawsmoved along branches they usually moved side-ways one foot after the other. They normallywalked slowly but while playing/fighting, theyoften moved fast. During a fast move throughthe canopy they sometimes kept their balance bygrasping the branch with the beak.

Acrobatics. The macaws were several times ob-served hanging upside down. They sometimeshung for quite a long time, swinging slightlyback and forth or bending the head backwards tolook around. Sometimes they hung by only oneleg. From this position at times they grabbedhold of a lower branch with the beak and swungaround to hang underneath that branch beforeregaining foothold.

Chewing on dijferent objects. The macaws in Su-cusuma and Río Chico were often observedchewing on objects (3 % of the activities in Sucu-suma and 2% in Río Chico). In Sucusuma theychewed on bark, branches, twigs and leaves,while in Río Chico they also chewed on piecesof rock.

of individual feathers. When preening breast orback feathers the beak performed nibbling move-ments.

During a preening session the birds scratchedthe plumage, shook the plumage, stretched dif-ferent parts of the body, and preened the feet orthe beak.

The macaw scratched 'directly' the head andneck region as well as the breast region witheither foot. Scratching and shaking of the plum-age seldom lasted more than a few seconds.

The macaws stretched wings, tail, legs orjaws. They spread one wing at a time down-wards, in a wing-to-foot movement combinedwith fanning that side of the tail, or showed aboth wings stretch, with the tips of the wingsfolded. The jaw muscles were stretched by open-ing the beak as wide as possible either once ormany times in succession.

The foot was nibbled by raising it to thebeak, and the macaw cleaned each toe in a bitingmanner. This activity usually lasted less than aminute. The beak was cleaned by rubbing it withthe foot or against a branch or cliff. Biting orchewing on branches and twigs probably servedto keep the beak clean and fit. On a few occa-sions the macaws were seen rubbing the regionbehind the eye on a branch.

Locomotion. Sitting/resting and autopreeningwere often interspersed with locomotion of shortduration. Locomotion was most frequently ob-served at the beginning or end of a rest period,when the birds were more restless. Most oftenthe macaws only moved a short distance on thesame branch (6 % of the activities noted on scanrecords in Sucusuma and 3% in Río Chico). Themacaws only moved rarely between branches inthe same tree or between trees during the rest,and they often sat for hours on the same branch.Moving to another branch in the same treeaccounted for 2 % of the activities in Sucusumaand 1% in Río Chico. Flying from one tree toanother accounted for 2 % of the activities inSucusuma and 7 % in Río Chico. Circling in thevalley or gorge was seen very few times in Sucu-suma, while it was a common activity in RíoChico (8% of the activities). In Río Chico thenest prospectiIi.g pairs also flew to the cliff neartheir nest holes many times a day (2 % of the

activities).

SOCIAL INTERACTIONS,

BEHAVIOR WITHIN PAIRS

Social interactions made up 11-18% of themacaws' activities (Fig. 1). Both agonistic andnon-agonistic behavior occurred, but most ofthe time was spent on non-agonistic behavior.Besides allopreening and playing/fighting, se-veral times we also observed social interactions ofshorter duration within pairs, such as caressing,copulating and courtship feeding.

Allopreening. Allopreening was only seen be-tween members of pairs and between juvenilesand parents (Christiansen & Pitter 1993b), andnot among groups of three or more.

The far most commonly observed type ofallopreening was simultaneous allopreening (Fig.2). More rarely one bird preened anuther whichwas sitting passively or autopreening. Usually,after some time the roles were reversed or theystarted simultaneous allopreening. Allopreeningoften lasted for 15-30 minutes interspersed withvarious other behavior types such as autopreen-ing, sitting, etc.

136

BEHAVIOR OF ARA RUBROGENYS

FIG. 2. Time spent on simultaneous a!lopreening and non-simultaneous a!lopreening on different parts of the

body: beak, cloaca! and plumage in Sucusuma (a) and Río Chico (b).

137

PITTER & CHRISTIANSEI

The birds faced in the same direction oropposite each other while allopreening. Oftenone of the birds crouched on the branch andraised the feathers on its head and neck, similarto begging behavior of juveniles (Christiansen &Pitter 1993b). The macaws allopreened all partsof the plumage.

During simultaneous allopreening they eitherpreened each other in exactly correspondingparts of the body or they preened different partsdependent on where it was convenient in rela-tion to their position.

The macaws were often seen sitting oppositeeach other on the branch preening each other'sundertail coverts and around the cloaca. In be-tween they grabbed the tail feathers and pulledthem through the beak.

When the macaws alternated between allo-preening and autopreening, the autopreening wasoften carried out as if the macaws imitated eachother and preened exactly the same part of the

body.The mated pairs used allopreening as a greet-

ing ceremony as well, e. g., when the femalereturned after a long stay in the nest hole or aftera long rest without any interaction between thetwo.

Caressing. Pairs were observed nibbling each othervery carefully in the plumage around the beak,on the chin, whiskers, lores and forehead. Pairswere also seen grasping each other's beak andthen rubbing their tongues together.

Playing/fighting. The macaws spent more timeon playing/fighting in Río Chico than in Sucu-suma (Fig. 1). The intensity of aggression varieda lot and in most cases it was not serious fights,but play.

Four main types of agonistic displays wereobserved, three of them are shown in Fig. 3.These partly followed the pattern found bySerpell ( 1981) in captive birds belonging to thegenus Trichoglossus. The macaws often alternatedbetween these displays as well as other behavior.Intensity and type of display depended on thecontext in which it was performed.

they often continued pecking alternately over along period. Attack was avoided by turning thehead aside or moving away or the bird defendeditself by pushing a leg toward the attacker to keepit at a distance. The attacked party could escapeby swinging down below the branch, often hang-ing by one leg, fluttering the wings and defend-ing itself from this position. The attacker oftenpursued the escaping bird gaping while sidlingrapidly along the branch, or while hanging up-side down the attacker bent to peck the other inthe head or continued pecking at the leg of thebird, which often had to release its grip. On oneoccasion two birds were seen pecking at eachother, both of them hanging upside down. Thepecking either stopped suddenly or it continueduntil one bird lost its balance. However, the birdsnever injured each othe!.

(b) Beak wrestling and beak fencing (Fig. 3 b ): Thebirds wrestled by taking hold of each other'sbeak and twisting their heads from side to side orthey fenced with the beaks.

(c) Alternating jerks (Fig. 3c): In this display thebirds demonstrated agitation by very rapid andsudden jerky movements. The partners nearlyalways coordinated and synchronized thesemovements. They both turned the head to theright, to the left, against each other or away fromeach other at the same time or bent the bodiesforwards over the branch with an exaggeratedpeering at a possible disturber or out in the air.During the display they often cackled loudly and

intensely.

( d) 1earing up and tossing objects: When agitated,the macaws were seen tearing up grass violentlyor picking up pieces of rock, only to toss themdown the cliff.

Pecking was the most commonly seen dis-play. The degree of pecking varied much. Beakwrestling and beak fencing, alternating jerks,tearing up and tossing of objects were seen moreseldom and only when the birds were agitated.

Vigorous pecking and the other playing/fighting displays (b-d) were often seen beforeseparation and immediately after very close in-teractions such as copulation and courtship feed-ing. They were also seen as a response to distur-bance from outside. The effect of the cooperationwithin the pair was that the aggression directed

(a) Pecking (Fig. 3a): This behavior was mostlycomposed of gaping and pecking vigorously atthe partner's head or legs. The attacked birdwould often rapidly reciprocate the pecking and

138

BEHAVIOR OF ARA RUBROGENYS

b

c

FIG. 3. Types of display of playing/fighting. 3a: pecking, 3b: beak wrestling and beak fencing, 3c: alternating jerks.

139

PITTER & CHRISTIANSEN

was seen before they left for the feeding area oras a respohse to disturbance and predators orduring the rest. Very often we observed singlebirds or pairs initiating an activity which there-after spread rapidly in mimetic fashion through-out the group. Many individuals could suddenlyperform the same non-agonistic activity, such aspreening, scratching, circling in the gorge, or thesame agonistic activity, such as pecking or alter-

nating jerks.The vocal repertoire varied. Two types of

vocalization commonly spread throughout the

group.

Quiet twitter-vocalization: Partners were oftenseen flying to a particular branch or vertical cliff,crouching very close together for a long timewhile duetting. The loud jolly cackling andtwittering passed into a quieter cooing andchuckle.

Alert-vocalization: A pair often simultaneouslysat upright and vocalized loudly and intensely.At intervals they screeched stridently.

Alert-vocalization was performed in responseto disturbances. Particularly in Río Chico thebirds performed quiet twitter shortly after theyhad arrived in the morning. The nest-prospect-ing pairs flew to the cliffs close to their nestholes, and the other pairs sat in the trees and theentire gorge reverberated with vocalizations.Quiet twitter also followed arrival of conspecificsto the area or to the same tree.

At the end of the midday rest the macawsbecame restless and moved around on thebranches or flew between the trees while vo-calizing loudly. This restlessness coincided inSucusuma with a daily change in the weather,which became windy and overcast and resultedin the whole group leaving for the same feedingarea.

toward the disturber appeared more vigorous and

lmpresslve.Pecking was the type of display that could

have the longest duration. Beak wrestling andbeak fencing, alternating jerks, tearing up andtossing of objects were all of shorter duration andwere very often followed or seen alternating withother displays. After alternating jerks the ma-caws often suddenly turned their aggressiontoward each other and pecked at each other withquick and violent movements. Other displays,su¿h as tossing of rocks, tearing up grass orviolently rubbing the beak against branchesoften accompanied the jerks and pecking dis-plays as well.

The macaws very often changed betweenagonistic and non-agonistic behavior. In stressfulsituations, where the birds became restless, thechanges occurred more often, resulting in eachdisplay only lasting a few seconds.

In particular, alternating allopreening andpecking was frequently seen and could often lasthalf an hour. While peacefully allopreening, twopartners could suddenly start fighting withoutany obvious warning, but often allopreening wasmore intense and violent before it turned intoaggression. After a short fight the birds usuallyquietly resumed auto- or allopreening, or some-times they moved half a meter apart. Partnerseven alternated between caressing around thebeak and fighting several times in succession.

Copulation. This behavior is described in Chri-stiansen & Pitter (1993a). Few copulations wereseen outside the breeding period. During thebreeding period copulations were often seen ( on16 occasions in Río Chico).

Courtship feeding. This behavior is described inChristiansen & Pitter (1993a). Courtship feedingwas seen both during the non-breeding and thebreeding period (on 4 occasions in Sucusuma andon 3 in Río Chico).

Agonistic behavior was usually seen withinpairs. Only occasionally were the aggressionsdirected toward other pairs or individuals.During the breeding period, on nine occasionsthe nest-prospecting pairs were seen to defendtheir territory immediately around the nest holeagainst conspecifics. The defence was characte-rized by a fairly low level of aggression (Christian-sen & Pitter 1993a).

SOCIAL INTERACTION, BEHAVIOR

WITHIN THE GROUP

Only very few interactions took place outsidethe pairs, except that the group as a whole oftencommunicated by immediate copying of beha-vior, vocalization as well as movement, of oneindividual by another (social facilitation). This

140

response to disturbances. However, duetting per-formed by mated pairs close to their nest holesmay be an activity that serve to strengthen thepair-bond in the early breeding season.

Communication in the jlock. Despite the fact thatthe macaws rested in flocks, any interactionsbetween members of different pairs were rare.Social interactions mostly took place withinpairs, between parents and juveniles (Christian-sen & Pitter 1993b), or between juveniles.

The entire flock often communicated eitherby vocalizing or performing the same move-ments synchronously. In general, movementimitation is highly developed among Psittacifor-mes (Moore 1992). Social facilitation as a meansof flock coordination was also observed in Oran-ge-fronted Parakeets Aratinga canicularis (Hardy1965), African L:>vebirds Agapomis spp. (Dilger1960) and Budgerigars Melopsittacus undulatus(Brockway 1964). Levinson (1981) found periodsof high activity in White-fronted Amazonsduring the day where all individuals in the flockwere in a highly excited state. This appeared tofunction in increasing group cohesion.

Harrison ( 1964) suggested, that allopreeningwas often seen between neighbors in socialgroups of birds and reduced aggression betweenmembers of a group. Allopreening never exten-ded outside the mated pair in Canary-wingedParakeet (Arrowood 1990), or in Budgerigars(Trillmich 1976), and pairs in Red-frontedMacaws. The spreading of allopreening probablyreduced aggression to a low level between allpairs of a flock.

Aggressions between neighbors in the flockswere exceptionally few, and chasing of otherindividuals was seen on very few occasions. Inthe feeding grounds we only observed veryfew aggressions between conspecifics (Pitter &Christiansen 1995). This was also the case ina group of Hyacinth Macaws Anodorhynchushyacinthinus (pers. obs.). Aggression betweenindividuals in a group appears to be more com-mon among smaller species of parrots. Serpell(1981) suggested that the genus Trichoglossus isexceptionally aggressive both in the wild and incaptivity toward conspecifics and other species.The same was found for Orange-chinned Para-keets Brotogeris jugularis, where pairs were oftenseen fighting with other pairs or two flocks

DISCUSSION

Intrapair behavio1: Cohesion within the pairs ofRed-fronted Macaws is very strong as they act asa unit throughout the day doing almost every-thing together, while maintaining close proxi-mity. This was also observed in mated pairs ofCanary-winged Parakeets Brotogeris v. versicolu-rus in captivity (Arrowood 1988).

The Red-fronted Macaw must be assumed toinaintain sustained pair-bonds, as also with mostother parrot species (Forshaw 1989). The in-creased amount of time that adult birds spentallopreening in the early breeding period pro-bably indicated the importance of maintainingthe relationship between the pair members inthe beginning of and during the breeding period.Copulations and courtship feeding seen outsidethe breeding period supported the view that pair-bonds were maintained throughout the year.

The macaws allopreened almost any part ofthe body. Harrison (1964) stated that most allo-preening in birds is restricted to head and throat,while especially preening under the wing isexceptional. Hardy (1963) suggested that allo-preening directed to the head, wings and tailareas in pairs of Orange-fronted Parakeets (Ara-tinga canicularis) is the strongest behavioraldevice for maintenance of the pair-bond through-out the year. This may also be the case with theRed-fronted Macaw.

Simultaneous allopreening wa~ by far themost common type of allopreening among themacaws and according to the interpretations ofHarrison (1964), this indicates that there is norank difference between the members of a pair.Allopreening among the Red-fronted Macawsvery often passed into playing/fighting. This isvery common among birds (Harrison 1964) andallopreening serves as an appeasement (Hardy

1963). Presumably allopreening replaces aggres-sive fighting in mated pairs. Social play has alsobeen found in White-fronted Amazons Amazonaalbifrons (Levinson 1981) and in Green-wingedMacaws Ara chloroptera (Deckert & Deckert

1982).Arrowood (1988) stated that duetting be-

tween pair members does not promote bondingin Canary-winged Parakeets, but functions as acoordinated pair threat against conspecifics. Inthe Red-fronted Macaw we observed duetting in

141

fully acknowledge financial support from theWWF (World Wide Fund for Nature) Denmark.

against each other (Power 1966). Agonistic be.havior was frequent in Budgerigars in captivity.However, in the wild there were few agonisticencounters between the birds and no peckorderor dominance hierarchy (Wyndham 1980). Weobserved no evidence of peckorder in flocks ofRed-fronted Macaw, except for a breeding paitthat seemed to be socially superior to non-bree.ding birds (Christiansen & Pitter 1993a).

Aggression between pairs of Red-fronted~acaw was probably reduced by the spreadingout of the pairs when resting. Galahs Eolophusroseicapillus also space themselves out whenroosting and feeding (Rowley 1990).

Furthermore, aggression was diminished bymost individuals in a resting flock sitting ~npairs, most aggression being kept within thepairs and often alternating with appeasingallopreening and caressing. A low degree ofaggressive interaction between pairs contributedto strengthen the group.

Ward & Zahavi (1973) suggested that jointroosting and breeding areas serve as informationcentres where knowledge of good feeding placesis passed on to conspecifics rather than as an anti-predator mechanism. The grouping of the Red-fronted Macaw during the midday rest probablyserves both to avoid predators and exchange in-formation on good feeding areas. The food of themacaws was in certain periods of the year rathersparse (Pitter & Christiansen 1995) and informa-tion on good feeding areas could be of great im-portance for the survival of the macaws. Pere.grine Falcons (Falco peregrinus) occasionally wereseen attacking the macaws and overflying raptorscould cause a large group of macaws to take off(Pitter & Christiansen 1995) so the grouping ofthe macaws also may serve as an antipredatormechanism.

ACKNOWLEDGMENTS

We appreciate greatly assistance during variousaspects of this study from A.B:P.A (AsociacíonBoliviana Para la Proteccíon de las Aves). Wethank Jon Fjeldsa, Zoological Museum, Uni-versity of Copenhagen, for very helpful com-ments on the draft and for his drawings of theRed-fronted Macaw. Thanks are alsodue to MaryE. Petersen, Zoological Museum, University ofCopenhagen for linguistic corrections. We grate-

REFERENCES

Altmann, J. 1974. Observational study of behavior:Sampling methods. Behaviour 49: 227-267.

Arrowood, P. C. 1988. Duetting, pair bonding andagonistic display in parakeet pairs. Behaviour 106:129-157.

Arrowood, P. C. 1990. Male-male, female-female andmale-female interactions within captive Canary-winged Parakeet Brotogeris v. versicolurus f1ocks.Acta XX Congressus Internationalis Ornithologi-ci. Pp. 666-672 in Bell, B. D., Cossee, R. O., Flux,J. E. C., Heather, B. D., Hitchmough, R. A.,Robertson, C. J. R., & M. J. Williams (eds.). NewZealand Ornithological Congress Trust Board,Wellington, New Zealand.

Boussekey, M., Saint-Pie, J., & 0. Morvan. 1991. Ob-servations on a population of Red-fronted MacawsAra rubrogenys in the R¡o Caine valley, centralBolivia. Bird Conserv. Intern. 1: 335-350.

Brockway, B. F. 1964. Ethological studies of the Budge-rigar Melopsittacus undulatus: Non-reproductivebehavior. Behaviour 22: 193-222.

Christiansen, M. B., & E. Pitter. (1993a). Aspects ofbreeding behaviour of Red-fronted Macaws, Ararubrogenys, in the wild. Gerfaut 82-83: 51-61.

Christiansen, M. B., & E. Pitter. (1993b). Aspects ofbehaviour of juvenile Red-fronted Macaws, Ararubrogenys, in the wild. Gerfaut 82-83: 63-69.

Deckert, G., & K. Deckert. 1982. Spielverhalten undKomfortbewegungen beim Grünf1ügelara Ara chlo-roptera G. R. Gray. Bonn. zool. Beitr. 34: 269-281.

Dilger, W. C. 1960. The comparative ethology of theAfrican parrot genus Agapornis. Z. Tierpsychol.17: 649-685.

Fjeldsa, J ., & N. Krabbe. 1990. Birds of the high Andes.

Copenhagen.Forshaw, J. 1989. Parrots of the world. Third revised

edition. Willoughby, Australia.Hardy,J. W. 1963. Epigamic and reproductive behavior

of the Orange-fronted Parakeet. Condor 65:169-199.

Hardy, J. W. 1965. Flock social behavior of the Orange-fronted Parakeet. Condor 67: 140-156.

Harrison, C. J. 0. 1964. Allopreening as agonisticbehaviour. Behaviour 24: 161-209.

Hueck, K. 1966. Die Vegetation Südamerikas. Die Wal-

der Südamerikas. Stuttgart.Lanning, D. 1982. Survey of the Red-fronted Macaw

Ara rubrogenys and Caninde Macaw Ara caninde inBolivia, December 1981 -March 1982. Unpub-lished report to the New York Zoological Societyand International Council for Bird Preservation.

l42

BEHAVIOR OF ARA RUBROGENYS

Lanning, D. 1991. Distribution and breeding biology ofthe Red-fronted Macaw. Wilson Bull. 103: 357-365.

Levinson, s. T. 1981. The social behavior of the White-fronted Amazon Amazona albiJrons. Conservationof New World parrots (R. F. Pasquier, ed.).Washington, District of Colombia, SmithsonianInstitution Press/International Council for BirdPreservation Tech. Publ. No.1: 403-417.

Moore, B. R. 1992. Avian movement imitation anda new form of mimicry: tracing the evolutionof a complex form of learning. Behaviour 122:231-263.

Pitter, E" & M. B. Christiansen. 1995. Ecology, statusand conservation of the Red-fronted Macaw Ararubrogenys. Bird Conserv. Intern. 5: 61-78.

Power, D. M. 1966. Agonistic behavior and vocalizationsof Orange-chinned Parakeets in captivity. Condor68: 562-581.

Rowley, I. 1990. The Galah. Behavioural ecology ofGalahs. NSW

Serpell, J. A. 1981. Duets, greetings and triumph cere-monies: Analogous displays in the parrot genusTrichoglossus. Z. Tierpsychol. 55: 268-283.

Solomon, J. C. 1989. Floristic inventory of tropicalcountries. Pp. 455-463 in Campbell, D. G., &H. D. Hammond, (eds.). Bolivia. New York.

Trillmich, F. 1976. Spatial proximity and mate-specificbehaviour in a flock of Budgerigars Melopsittacusundulatus: Aves, Psittacidae. Z. Tierpsychol. 41:307-331.

Ward, P., & A. Zahavi. 1973. The importance of cer-tain assemblages of birds as "information-centres.for food-finding. Ibis 115: 517-534.

Wyndham, E. 1980. Diurnal cycle, behaviour and socialorganization of the Budgerigar Melopsittacus undula-tus. Emu 80: 25-33.

143