BEHAVIOUR OF AUSTRALIAN PTEROPODIDAE (MEGACHIROPTERA)

BEHAVIOUR OF AUSTRALIAN PTEROPODIDAE (MEGACHIROPTERA) BY JOHN E . NELSON* Zoology Department, University of Queensland . Introduction The genus Pteropus (Megachiroptera) is a very successful one, its members forming camps containing many thousands of individuals . Although_ some observations have been made on the behaviour within the camps (see Eisentraut, 1945, for review), no detailed study has been made of the social organization within the camp . Kulzer {1961) studied the behaviour in captivity , of Rousettus aegyptiacus, and described a large variety of behavioural and sensory capabilities that contrasted sharply with the highly specialized behaviour and sensory abilities of the Microchiroptera (Griffin, 1958) . A study of the social organization of Pteropus not only allows a comparison of the two Chiropteran suborders, but it also allows comparison of a highly mobile mammal with less mobile forms such as the primates and elephant seals . Materials and Methods The three species of Pteropus which occur in south-eastern Queensland were studied, but the majority of observations were made on the most common species-P . poliocephalus Temminck 1825 . Of the other two species, P . gouldi Peters 1867 was present at all times in small numbers within the poliocephalus camps, and P . scapulatus Peters 1862 was found irregularly in either poliocephalus camps, or in camps containing only scapulatus . All general statements refer to poliocephalus . Seven camps within 100 miles of Brisbane were studied regularly, and irregular observations were made in other camps both inside and outside the study area . All camps in the study area were visited monthly during 1961 and 1962 . Some camps were visited weekly for a month or two, and Murwillumbah was visited daily from March 20th to April 8th, 1961 . One camp (Eagle Heights) was in a National Park, and so was relatively undisturbed by the activities of man . Careful observation of the undisturbed animals and detailed recording of their behaviour at the time of observation were made with the aid of a telescope and a tape recorder . The flying *Present address : Department of Zoology and Comparative Physiology, Monach University, Victoria . 544 foxes became disturbed and took to the air as the observer entered the camp, but soon settled down and showed little response to the presence of the observer as long as he remained quiet and did not move quickly . Some individuals could be recognized during March and April (when territories were being defended) by the patterns of bullet holes in the wing membranes . Outside of this period, individual recognition was possible only during the short periods of continuous observation (3 to 5 hours) . Some observations of mother-young behaviour were made on flying foxes held in wire mesh cages 50 x 50 x x 100 cm ., in an open-field shelter on the roof on the Biological Sciences Building at the University of Queensland, Brisbane . These animals were maintained on a diet of cultivated fruit such as bananas and pawpaws . 1 . Locomotion Since the social organization of an animal species is greatly influenced by its mobility and sensory capabilities, it is necessary to describe briefly the locomotion and senses of flying foxes . The chiroptera are unique among mammals in their capacity for sustained flight . As well as enabling flying foxes to explore large areas while feeding at night, the ability to fly allows them to escape rapidly from an approaching predator or aggressive member of the species . When disturbed, flying foxes move quickly from the lower (often shaded) branches, where they usually hang, to the higher and more open areas of the canopy where they quickly flap the wings until their bodies are horizontal, and then after releasing their hold on the branch they fly away . Slow movements along the branch are made bipedally . (The lower leg bends forward from the knee .) Fast movements are made by a quadrupedal "walk" on the underside of the branch . In the latter, the thumb claw affords purchase for the forelimb, the fingers and outer membrane of which are folded along the forearm . Flying foxes usually climb large vertical limbs and trunks of trees by an inch-worm motion in NELSON : BEHAVIOUR OF AUSTRALIAN PTEROPODIDAE which both thumbs are moved up together and grip around the limb and then the legs are drawn up . Smaller limbs, and occasionally large limbs, are climbed by alternatively extending thumb claws and then alternatively or simultaneously drawing up the legs . Landings on exposed branches are made either by passing over the branch, then gripping it with the feet claws so that the body swings down on opposite side to approach or by half rolling on approach side and then gripping with claws as the body passes under branch . In more confined branches, hovering may occur briefly while the feet stretch towards the branch . The thumb claws sometimes assist in landings . On the foliated edges of the canopy, landings may occur with the extending wings providing the traction until the foot claws grasp the branch . 2 . Senses Moehres (1953) and Moehres & Kulzer (1956) found that the pure rod eye of Pteropus is well suited for twilight vision and depth perception, and that Pleropus orientates visually even in very dim light, but refuses to fly in complete darkness . It cannot navigate acoustically as the microchiroptera do ultrasonically and the megachiropteran Rousettus does audibly (Griffin et al ., 1958 ; Novick, 1958) . Kolmer (1911) and Rochon-Duvigneaud (1943) have described the unique folding of the retina and choroid of Pteropus into 20,000 to 30,000 conical papillae so that the rods sit "like the trees in a range of mountains . . . . Presumably the bat sees the visual field at a given distance as a (relatively!) sharp reticulum, the lacunae of the lace-work being much more badly blurred ." (Walls, 1942) . In camps, flying foxes attempt to locate ground disturbances visually by swaying the body from side to side about the hips with the head flexed ventrally or dorsally . Used in this way, a "lace-work" retina would probably be very efficient for visual location among the vegetation of a rain forest . The senses of vision and hearing normally complement each other . Once the source of a sound made by the observer has been located acoustically, flying foxes fix the source visually and continue to follow the movements of the observer around the cage . While "staring" at the source, the pinnae do not oscillate violently with each new sound . In contrast, the pinnae of an animal whose eyes have been treated with atropine, so that the pupil is dilated causing temporary blindness, do oscillate violently with each 54 5 sound . Such treated animals are very nervous, shivering and urinating continually and hanging with their muscles contracted, nor does the animal strike with the thumb claws at objects held close to its head, or "stare" at the observer . However, the treated animal is able to locate food placed in the cage . Untreated P . poliocephalus were able to locate food placed in open paper sacks which were placed under boxes on the floor of an 8 foot high cage . The flying foxes did not normally come down to the floor of the cage for food and water were usually placed in trays suspended from the roof . Rousettus, a genus related to Pteropus, is able to locate 100 milligrams of banana substance and to distinguish between the banana substance and the banana essence (Moehres & Kulzer, 1956). Kulzer (1961) showed that Rousettus used its olfactory sense for . locating food, for identifying the home cages ; and for recognizing individuals . Although Moehres & Kulzer (1956) found that Pteropus were very sensitive to crackling noises such as the breaking of sticks, no quantitative measurements of their auditory ability have been published . Some neural filtering of the noises in the camp must occur, for, although sleeping flying foxes respond to noises by quivering movements of the pinnae, they open their eyes and search for the source only when unusual noises are heard . P . poliocephalus 1 . Camps There are two types of camps-summer camps and winter camps . The summer camps are occupied from September to April or June, and the winter camps are occupied from April or June to September . Most summer camps are occupied regularly each year and some have been for at least 80 years . Each winter camp is occupied continuously or seasonally for only 2 or 3 years . In the summer camps the young are born and raised, the selection of a mate is made, territories are established and defended and conception occurs . The peak population in these camps occurs during December and January when selection of a mate is being made, and before territories are set up . The flying foxes in the winter camps are sexually segregated and contain a large percentage of immature animals . The ultimate factor controlling the population in both summer and winter camps is the food supply (Nelson, 1%5) . 5 46 ~ ANIMAL BEHAVIOUR, XIII, 4 During the day the flying foxes often groom (by combing the fur with the claws of the feet and by licking the fur and membranes), scratch (with the foot claws), fan their wings, a thermoregulatory response (Bartholomew et al ., 1964), and sleep with their wing membranes folded about the body . They are most active in the camps about dawn and dusk, but there are always some active at any time of the day . The flying foxes are least active during rainy periods when they wrap their "waterproof" membranes about themslves and sleep . Females are pregnant from March to October and lactating from October to January or later . Females are receptive to males from about January to March . Males have sperms in the epididymis all the year and will attempt to mate with females at any time . Testis activity is greatest in the month preceding conception and least when the species is dispersed . 2 . Reproductive Behaviour (i) Sexual Segregation before Conception From early September until early December the sexes tend to be segregated, but the extent of the segregation varies . Sometimes a tree contains only one sex, or predominantly one sex, and at other times both sexes are present in similar numbers, but with one sex predominantly higher in the tree and the other lower down . Thus, most of the social contacts at this time are between individuals of the same sex . Aggressive behaviour is not pronounced, being expressed in the brief mild contests when the activity of one animal disturbs another . Each flying fox shows interest in its neighbour, and may touch one with a thumb claw (and receive a cry of protest) or lean towards one and run the nose from the ear to the wing insertion . The scapular glands, from which the peculiar odour of flying foxes originates, lie under this area . Pregnant females and especially females with new born young are nervous, and fly away at the least disturbance . The males occasionally approach females and attempt to lick the genital area . The females object to this by vocalizing and/or biting the neck of the male . If the male persists in his attempts, the female flies away . During this period, both sexes sometimes engage -in homosexual behaviour, which is more common in the males . Each partner wraps the wing membranes around the other, gently bites and licks the chest and membranes of the other, pulls the claws of the leg through the other's fur, and rubs the neck and head along the chest and neck of the other . This behaviour is very similar to the mutual grooming which has been observed in such animals as the primates (Washburn, 1962), and rodents (Eibl-Eibesfeldt, 1958) . That grooming may have a sexual function, as well as a cleaning one, is further indicated by the observation that a male occasionally erects its penis when cleaning its own membranes . Erection also occurs in males which spasmodically rub the underside of the chin and side of the neck along a branch . This behaviour is a low intensity form of "marking" which is fully developed later in the breeding season (February to April) . (ii) Parturition Females give birth to a single young each year between late September and late October . During parturition the female hangs by the feet and sometimes by a thumb claw too . For some time before the appearance of the foetus at the opening of the vagina, the females periodically licks the genital area cleaning it of amniotic fluid . In all of the observed parturition, the young were born head first . The foetus is positioned so that its wings are folded around the head . When the head appears at the vaginal opening the female begins to lick vigorously, interrupting the licking only to investigate local disturbances . The expulsion of the foetus continues smoothly until all of the head is out of the vaginal opening, and then there is a brief pause of 10 minutes or longer . At this stage, the "foetus" has its eyes open, twitches its ears slowly but with no apparent orientation to sound, turns its head about, and licks its lips and nose . The shoulders appear slowly . Then, with a violent contraction, the shoulders and the rest of the foetus are forced out of the vagina . The female uses her mouth and the claws of one leg to hold the young and to guide it over her body to the axillary nipple . The newborn attempts to suckle any protuberances resembling the teat, and can vocalize immediately after birth . Some minutes after birth the female ceases to clean the fur and membranes of the young, and begins to chew through the umbilical cord close to the vaginal opening . Apparently the female eats the placenta, as no traces of it were found on the floor of the cages, or on the ground under the camps, and as puerperal females shot at this time had blood in the lumen of the undamaged gut, NELSON : BEHAVIOUR OF AUSTRALIAN PTEROPODIDAE ~ 547 (iii) Motber-Yereg Rehtionships aid the Develcpiwit of the Youg The newborn are naked on the chest, abdomen and underside of the chin, and sparsely haired on the snout . They are dependent upon the female to maintain their body temperature within the normal adult range (Bartholomew et al ., 1964) . The milk teeth are posteriorly curved, and the claws of the feet and thumbs have hook-like growths on the inside curve of the claw facilitating the attachment to the fur of the female . These juvenile tips (which later fall off) have been described in Rousettus leschenaulti by McCann (1941) . By catching these claws on the fur and biting on the nipple, the young remains firmly attached to the females even in flight . All movements of the newborn are poorly coordinated . By 3 to 4 weeks the young are well co-ordinated. They can hang separately with their wings folded around the body, scratch with one leg while they hang by the other, clean themselves, threaten with outstretched wings when violently disturbed (though they do not strike with the thumb claw), climb along branches, and locate disturbances visually and acoustically . The fur has grown on the ventral surface and the young are able to regulate their own body temperature . However, they are unable to fly . From birth until they are weaned at about 4 months the young become progressively independent of the female . For several days after birth they are rarely seen away from the nipple . They gradually become more active, first hanging from the female with one leg holding the branch (Plate VIII, Fig . 3) and later hanging beside the female . By 3 weeks, although they are unable to fly, the young are left in the camps at dusk when the females fly out to feed . All of the young are left in the same area which contains well foliated trees, and which may be on the edge of the day camp, where the leaves are often stripped from the branches by the activity of the adults (Plate VII, Fig . 2) . The females fly with their young to this area at dusk, and then begin to lick their offspring, which place first one leg and then the other on the the branch . In the dim light it is not possible to see if the female is gently forcing the young away from her body with her snout . Both hang together for a while before the female flies away . The females return before dawn, and circle this area calling to their young . The young reply as the female flies about hovering, changing direction, and finally landing near one of them . The female smells the young's chest, and either accepts it by drawing back her wings to expose her chest, or rejects it by pushing it away with her thumbs and wings, and then flying away . The female licks the accepted young vigorously as it climbs on to her chest, and then folds her wings around it . Laboratory experiments indicated that the 3 week old young does not recognize its parent, and attempts to climb on to any female that approaches it, but the females are able to recognize their offspring, apparently by olfaction . Females whose young had been removed became very agitated when presented with cloth sacks containing crying young, but, once they had smelled the sacks, they remained agitated only near their own young . Kulzer (1961) reported that Rousettus females similarly identified sacked young . He also reported that the young were silent when a strange female was near the sack . Infant mortality is probably highest during the period when the young are left in the camps before they can fly . Young that fall below the canopy usually die because, although they can crawl easily over thin branches, it is difficult for them to climb the tree-trunk to the females which are always reluctant to come below the canopy of the camp . The young are present in the night camps from early November to late March . At first their numbers are small (e.g . 20) but they gradually increase to a peak ( .5,000) in DecemberJanuary, and then gradually decrease . The young fly at 3 months (January) . Their permanent teeth are developed, and they leave the camps at night to feed . Two 3-month old young, held since birth in cages too small to allow flight, were unable to fly when released . When the females first leave their offspring in the night camps, the young flap their wings as the female flies away, but do not release their hold on the branch . As they grow older the young half-fall, half-fly from one branch to another as the female flies away, or as she lands on a neighbouring branch on her return at dawn . Later the young fly short distances from tree to tree, and then fly short distances from camp at dusk, before they return to settle in the night camp . They gradually fly greater distances from the camp, until they fly out with the adults to feed . For some time after they can fly, the young in the day camps move to the females when alarmed, and are carried away by them . By 3 5 4 8 ~ ANIMAL BEHAVIOUR, XIII, 4 months, the offspring fly off independently when alarmed, though the females usually leave first . There is much variation in the period of lactation . Some young leave the female at 4 months and join the juvenile packs (see below), while others suckle for at least 6 months and at this age (weight 400 g .) may even be carried by the female (wt . 700 g .) in flight . Even after the young can clean themselves, the females regularly clean the fur, membranes and genital area of their offspring . Laboratory young were first seen inverting to defaecate at 6 weeks . One female and her 4-month old young were seen mutually "grooming" as described above in homosexual behaviour . (iv) Selection of Mate During December and January, when the peak population occurs in the summer camps, most of the social contacts are those made between the sexes, and in this period the selection of the mate is made . Females, apparently with their mobility reduced by nursing young, tend not to move about the camp unless disturbed . Thus the initial contact is made by the approach of a male . The female with her offspring at the nipple turns away from the male, flaps her folded wings against the sides of her body, and . calls two or three times . Except that the wings are folded, the flapping of the female's wings is similar to the flapping of her wings before she flies away . The female wing-flap may be analogous to some of the ritualized behaviour patterns described in birds, such as appeasement behaviour (Tinbergen & Moynihan, 1955) . The wing flap and accompanying vocalization inhibits the aggressive and sexual behaviour of the male which turns away from the female and either fans or grooms himself or else wraps his wing membranes about himself and goes to sleep . If the female remains near the male, he may stretch towards her and smell the scapular area or attempt to lick her genital area . The female objects at first, but gradually tolerates it for longer periods . In late January mutual grooming between males and females is very common, and may last as long as half an hour . It was not possible during this period to determine the extent of the association between males and females . During continuous observations, of 3 or 4 hours, a male periodically landed near a female, but it could not be determined whether it was the same or a different male that was observed each time, (v) Establishment of Territories From late January onwards, the males each with one or more females gradually become more aggressive as they begin to establish their ter- . ritories . Fighting between the males is very intense in late February, but is less intense in March when the boundaries of the territories have been learned . (If the camp is disturbed at this time, and the flying foxes move to the higher unoccupied trees, fighting is very intense .) The males mark the territories by rubbing the scapular glands, which are maximally developed at this time, along the branches . The male is much more alert and aggressive than the female and so does almost all of the fighting . If the male is asleep the female may begin a territory fight, but the activity then wakes the male who quickly finishes the fight . (vi) Composition of Breeding Camp During March and April the territories are established, and the numbers within the camp are roughly constant . The density of flying foxes in the camp increases from the outside edge (where there are' only I to 5 animals in a tree, and where not all trees are occupied) to the centre (where 400 may be found in each tree) (Plate VIII, Fig . 4) . Four different social groupings are present in these camps : (a) Guard groups, extending around the perimeter of the camp . (b) Family groups comprising the major proportion of the camp . These groups consist of a male, a female and a young conceived by the female in the previous breeding season . (c) Adult groups differing from the family groups in that (1) there are no young present and (2) the male may be either monogamous or polygamous . (d) Juvenile packs . (vii) Guard Group Since activity on the perimeter is low, the trees are well foliated and not stripped as in the centre of the camp . The small numbers of flying foxes on the edge hang at varying heights in the trees . Both sexes are found in this area but males are more common than females, each of which is associated with a male . The males presumably have been unsuccessful in territory fights in the centre of the camp and, in being forced to the edge, act as guards against ground predators, When the observer entered NELSON : BEHAVIOUR OF AUSTRALIAN PTEROPODIDAE the camp, at least one of these guards was disturbed . It gave an alarm cry and, first acoustically and then visually located the disturbance . The nearby flying foxes first located the one that gave the alarm and then by observing the orientation of its eyes and pinnae located the disturbance . (viii) Family Group and Adult Group The males in the family group are always monogamous while those of the adult group may be either monogamous or polygamous . Males are dominant to females, and stop fights between females of the polygamous groups by moving between the fighting pair . The males in the adult group are more aggressive and mark their territories more often than those of the family group . The territories in the former group are usually smaller than those in the latter . The territory of a family male was often a frond of a palm tree and in a large rain-forest tree it occupied about 3 feet of a branch and a 4 to 6 feet diameter cylindrical volume about it . In both groups the males mark the branches in two places-one where the male hangs and the other where the female hangs . The marked area and the surrounding branches within reach are defended by both the male and the female, but the male usually finishes fights started by the female . Although the male is 100 to 200 g . heavier than the female, sexual dimorphism is not pronounced, and both sexes have large canines . The members of a unit appear to recognize each other by olfaction . They often smell each other's scapular area (Plate IX, Fig. 5), and especially so when an intruder has been driven away from the territory and when there are disturbances in the camp. Whenever the female returns to the territory, the male moves from his position on the branch to the female, smells the scapular region, and quickly moves back to his branch . There is also an exchange of short soft calls . When a male lands among a group of adult units, the nearest male gives a territory warning cry . The other males of the group repeat the cry and move to the edges of their territories, threatening the intruder who is quickly smelt and then driven from one territory to another until he flies away . These threats and fights are less intense when a male belonging to their group lands and passes through their territories to his own . Since the territories have been learnt by this time, this suggests that "as the egocentric 549 intragroup aggression declines, co-operation in intragroup aggression increases" (Collias 1944) . Sometimes after a male has driven an intruder away, he utters a territory cry while flapping one or both partly-opened wings over his chest . his male wing-flapping is very common at a when the flying foxes return to their territories . At this time males are very alert, active, and aggressive . The males in the family group occasionally fight with the young . These fights are of short duration and low intensity . No males were seen driving young away, but the general aggressiveness of the male towards the young, even though slight, may drive the young to the juvenile packs . (ix) Juvenile Packs In late January, 2 or 3 young may be found near one adult-usually a male . These groups gradually become larger (as the number of family units become smaller) until late March, when they may contain 50 juveniles and 15 adults, two-thirds of which are male . These males show no territoriality or aggressive behaviour towards each other . They will, however, drive off adult intruders and break up intense fights among the juveniles . These adults regularly smell each other about the scapular gland . The adult male does not strike the juvenile with the thumb claws, as happens in many of the flying fox fights . The claw is thrust towards the young but stops just before its chest or head . Often, only the insides of the partly folded wings are used in these fights . The juveniles are very active in these packs, mutually grooming and "playfully" fighting and smelling each other . (x) Copulation Conception occurs in late March . Although the males attempt to copulate with females at any time, the females are receptive only during this period when the territories have been established . In a territory pair, copulation, though not necessarily ejaculation, occurs several times a day throughout this period . When the male approaches the female to initiate copulation, his pinnae are directed backwards and downwards (Plate VII, Fig . 2), as are the pinnae of frightened animals . (The pinnae of young flying foxes tend to be limp and so directed down and back .) Since the ears of an alert or aggressive animal are directed forwards and upwards, the pinnae of flying foxes are used in visual communications, as are those of wolves and dogs . 55 0 The male initiates precopulatory behaviour by mutual grooming or by licking the vaginal opening . The female may gently rebuff the male's advances for a while . Eventually the male licks the vagina for long periods, rocking backwards and forwards on the branch so that his nose is thrust well into the genital opening (Plate VII, Fig . 2) . Occasionally the female initiates precopulatory behaviour by wrapping her wings and legs around the male, so that he supports her completely, and then moving back to the branch . She repeats this until the male begins to lick her genital area or begins the mutual grooming . Copulation is performed dorso-ventrally with the male holding the female's wings between her upper-arm and his forearm, and gripping the back of her neck with his teeth . The male may also hold the female against his body with the claws of one leg which grip the female's abdomen . The penis is very flexible and can be moved independently of pelvic movements . After copulation each partner grooms the genital area and then the general body surfaces . (xi) Sexual Segregation after Conception After conception the females leave the territories and form groups of pregnant females . The males continue to defend the territories for a few days before they join male groups . Although this segregation is more complete than that seen in the summer camps before and during the birth of the young, odd males are found among the female groups and vice versa . The sexually segregated groups eventually leave the camp, either separately in small numbers or together in large numbers . Thus 60 males passed through Fisherman Island in early April, 1961, and 1,000 or so segregated males and females re-occupied Murwillumbah in May, 1960 . Groups of old males, occasionally containing some immature females, are the last flying foxes to leave the camps in April . If food is abundant as in late summer 1960, the summer camps, which are normally deserted in April, may be re-occupied during May and June by segregated flying foxes . The onset of the non-reproductive phase varies throughout the population so that some flying foxes are defending territories while others have left their territories to join the segregated groups leaving the camps . The result of this is that large deserted areas appear within the camps, fragmenting it into a number of smaller ANIMAL BEHAVIOUR, XIII . 4 camps, each containing about 200 flying foxes . The individuals of the segregated groups hang almost touching each other, yet exhibit no territoriality or aggressive behaviour, although they show great interest in each other, smelling regularly . Mutual grooming is common, and some males spasmodically rub their necks along the branch once or twice and then smell or lick the area . The immature of the juvenile packs are also sexually segregated at this time, immature males showing similar decrease in size of the accessory sex organs as adult males . The juvenile packs leave the summer camps in April, but unlike the adults, which normally disperse throughout the species range (Nelson, 1965), the juveniles form winter camps . 2 . Non-Reproductive Social Behaviour (i) Winter Camps The winter camps, first occupied in mid-April while the summer camps are breaking up, normally contain very large numbers of young in approximately the same proportion to adults as is found in the juvenile packs . In winters when there is an unusually good supply of blossom, adults form a larger proportion of the camp (Nelson, 1965) . The numbers of juveniles in these camps are such that the young must come from a number of summer camps . In early winter, the behaviour of the juveniles is similar to their behaviour in the juvenile packs of the summer camps, where mutual grooming and playful fighting are common, and where the intense fights are broken up by the adult males in the pack . By late winter, however, the sexual segregation is less apparent as the juveniles move about the camp, and the adult males no longer interfere in juvenile disputes . Mutual grooming is common between the same and opposite sexes, and juvenile males lick the vaginal openings of juvenile females, and attempt copulation . Of the two winter camps observed in the study area, one was deserted in early summer, the individuals presumably moving to a summer camp, and the other was occupied through the summer . Both were occupied in late summer when territoriality and conception occur . Most individuals shot at this time were first year adults, and so the juveniles had apparently returned to the camps for their first breeding season . The young are mature at 18 months (Nelson, 1965) . NELSON : BEHAVIOUR OF AUSTRALIAN PTER ~ AE ~ 551 (ii) Small Groups During the winter months the adults disperse throughout the range of the species either singly or in small groups . These small groups, and those in the summer camps at the beginning and end of their occupancy, are very difficult to locate as the flying foxes hang quiet and motionless among the leaves . Individual flying foxes released during the day in small bushes in an open clearing were very reluctant to fly, and when forced to do so, they were attacked by birds as they flew to, and hung in, the nearest tree . Small groups or single flying foxes normally avoid these attacks by birds by remaining quiet and well hidden in trees during the day and by leaving these shelters to feed only during the night . (iii) Nocturnal Feeding Flights The flying foxes leave their camps at dusk in two or three long wide columns each moving in a different direction and each taking up to half an hour to fly over a particular point . Although these columns move in the same direction each night, the numbers in each column vary throughout the summer . One flying fox may lead the column for a while, but will then turn and fly back along it, so that the leaders are continually changing . The size of the column gradually decreases as the flying foxes leave it to drink on the wing from the rivers, or to feed on the blossom and fruit . A number of flying foxes feed in the same tree . Each flying fox announces the possession of a feeding territory by a feeding cry . These cries are more common when the flying foxes first arrive in their feeding territories, which extend for 3 yards or more in each direction . The behaviour of feeding flying foxes cannot be clearly seen in the dim light . However, the analysis of vocal communications of feeding animals (Nelson, 1964) indicates that as well as the fighting behaviour associated with the defence of these feeding territories, flying foxes also engage in sexual behaviour and motheryoung behaviour . Since some of the sexual calls were recorded at the time when males and females form fixed associations, the groups in the day camps may extend into the feeding areas . The flying foxes return to the camp either singly or in small groups in the hour or so before dawn . P . scsPrf W n Although the seasonal behaviour of this species is out of synchrony with the other species, the general pattern is essentially the same . The species was seen only in the period November to April, and it is assumed that it is nomadic from April to November . The territories are defended during November . After conception in late November, the females leave the territories and form female groups . The males continue their territoriality and intragroup co-operation for a few days before they join the male groups . Usually the species establishes territories in large camps containing up to 100,000 individuals, and then after conception it leaves these camps, and either forms small scapulatus camps or joins poliocephalus camps . The young are born in May when the species is dispersed throughout its range, and are apparently raised without the protection of a large population as in the poliocephalus camps . The young are found in juvenile packs within the camps during November and December . The individuals of all species maintain a certain distance from one another at all times . but this distance is smallest just after conception, i .e . when segregated groups form . P . scapulatus maintains a smaller distance than the other species, and after conception the individuals lie against each other, even forming clusters of 20 or so with each member of the cluster facing inwards . This may be a thermo-regulatory behaviour pattern (Bartholomew et al ., 1964) . Interspecific Relations Within the poliocephalus camps, Pleropus gouldi is always present in small numbers, while P . scapulatus is occasionally found in either small or large numbers . The various species in mixed camps usually form separate groups, but if the numbers of each species are small (e.g . 5 to 20), and if all are in the non-reproductive phase, the various species hang together . The seasonal patterns of behaviour of gouldi are similar to and synchronous with that of poliocephalus, so that within the camps there is competition for space when territoriality is established . P . gouldi is larger and more aggressive than poliocephalus, and this species, as suggested by Ratcliffe (1931) and indicated in this study by a slight increase in their numbers within the camps over the 3-year period, may be replacing poliocephalus in the northern areas . 55 2 ~ ANIMAL BEHAVIOUR, XIII, 4 However, gouldi favours fruit more than poliocephalus, and this increase of gouldi may by the result of the exploitation of the relatively untouched food supply . Predators-Wedge-tailed eagles (Aquila audax) and white-breasted sea eagles (Haliacestus leucogaster) were observed attempting to catch flying foxes by swooping into the camps . If a flying fox sees an eagle approaching, it gives an aerial predator alarm cry, alerting the other flying foxes, which begin to utter the general alarm cry . If the eagle lands in a tree, the flying foxes move away from that area, but remain in the tree, threatening the eagle with outstretched wings and with vocalizations . No eagle was successful in catching a flying fox after failing to catch one on its initial swoop into the camp . Goannas (Varanus sp .) and pythons (Morelia sp .) shot in the camps had flying foxes in their intestines . Discussion 1 . Grooming Females begin to groom their young as soon as they are born, and by the time they leave their mother, the young are spending, and will spend for the rest of their lives, several hours a day cleaning the fur and membranes . It is especially important to keep the wing membranes clean, as uncleaned membranes become less supple due to interference with the production of the oily secretion from the epithelial glands . In addition, grooming helps to establish social bonds (as it does in baboons (Washburn, 1961), which also groom for long periods), and probably aids in the development of sexual behaviour patterns . Most of the early sensations of the young, including grooming, are associated with the female who is never far from her offspring . The female regularly cleans the genital area, as well as the fur and membranes of her offspring, while it lives literally on its mother . The young also spends much time enfolded within the female's wings . Later in its life, the young engages in mutual grooming with its mother, and this behaviour is identical with the adult homosexual and heterosexual mutual grooming . The precopulatory licking of the female vaginal opening by the adult male may originate from the cleaning of the excretory products from this area of the offspring by its mother . Collias (1956) drew a similar correlation between the smelling and licking of the female's external genitalia by the male goat and smelling and licking under the tail of a young kid by its mother . While the elements of their sexual behaviour appear to be derived from their associations with the female, the details of the aggressive behaviour are learnt firstly from the male in the family group and later by play in the juvenile pack . In these juvenile packs the young associates with animals of its own size and own level of experience and is protected, by the adult males present, from the dangers of lack of experience in fighting . In these packs the social bonds are greatly expanded, and the adult behaviour patterns, including the conventions of aggressive behaviour, are learnt . 2 . Aggressive Behaviour The conventions of aggressive behaviour, which are acquired in the protected atmosphere of the juvenile packs, include vocalizations, baring of the teeth, and the avoidance of actual body contact . As occurs in several other vertebrate species (Eibl-Eibesfeldt, 1961), fighting in flying foxes is ritualized, and rarely ends in death or serious injury -to either combatant . In the summer camps, aggressive behaviour and the associated territoriality were most marked in the polygamous males, less so in the monogamous, and least in the guard males . The adult males in the juvenile packs showed no aggressive behaviour . The order of dominance is probably polygamous males, monogamous males and guard males . In the monogamous and polygamous groups, the dominance of the males stabilizes the groups, and in the polygamous groups assures that the female fights will be stopped without injury to the females . Although the male of the family group never fiercely attacks the young of the female in his territory, the general aggressiveness of the male probably drives the young to the juvenile packs . 3. Olfactory Stimuli The odour of a mammal is derived from the volatile substances produced by, or present in, faeces, urine, cutaneous glands and substances, such as soil, which adhere to the body, Thus the odour can be affected by food, metabolism, reproductive state, locality and other factors which make for high variability in odour within the species . Cutaneous glands are common among mammals where they are found in all parts of the body (see Bouli6re, Grasses') . Some species have several glands which produce different substances