BEHAVIOR AND MAINTENANCE OF CAPTIVE WHITE-WINGED VAMPIRE BATS, DIAEMUS YOUNGI

Transcription

1 BEHAVIOR AND MAINTENANCE OF CAPTIVE WHITE-WINGED VAMPIRE BATS, DIAEMUS YOUNGI W. A. SCHurr, JR., F. MURADALI, N. MONDOL, K. JOSEPH, AND K. BROCKMANN Natural Science Division, Southampton College, Long Island University, Southampton, NY (WAS) 17 Winsor Road, Valsayn Park, Trinidad (FM) National Animal Disease Center, Caroni, North Bank Road, Trinidad (NM, KJ) Department of Ecology and Systematics, Corson Hall, Cornell University, Ithaca, NY (KB) Procedures for transportation and maintenance of the white-winged vampire bat, Diaemus youngi, are documented for the first time. Contrary to previous reports, D. youngi has been maintained on a diet of defibrinated bovine blood supplemented weekly with fresh chicken blood. Gregarious by nature, D. youngi exhibits dominance-hierarchy behavior, and behavioral patterns unreported in other species of bats. Although primarily arboreal with regard to feeding behavior, D. youngi demonstrates the ability to feed terrestrially-a behavior documented for the first time. Observations of our captive colony contradict anecdotal and previously published information on feeding and behavior. Differences in feeding behavior between D. youngi and the common vampire bat, Desmodus rotundus, appear to be related to selection of prey (arboreal versus terrestrial prey, respectively). In places where these vampire bats coexist, resource partitioning may serve to reduce competition. Key words: Diaemus, Desmodus, vampire bats, captive maintenance, behavior Reports on the maintenance and behavior of captive vampire bats (Phyllostomidae, Desmodontinae) generally are restricted to the common vampire bat, Desmodus rotundus (Dickson and Green, 1970; Ditmars and Greenhall, 1935; Greenhall, 1965; Trapido, 1946; Wimsatt, 1962a, 1962b; Wimsatt and Guerriere, 1961). D. rotundus is common in the wild (Ditmars and Greenhall, 1935; Greenhall, 1965), easily maintained, and long-lived in captivity (Dickson and Green, 1970; Trapido, 1946; Wimsatt, 1962a, 1962b; Wimsatt and Guerriere, 1961). Wimsatt (1978) maintained a colony of D. rotundus at Cornell University for >20 years and reported that at least one individual survived> 19 years. The hairy-legged vampire bat, Diphylla ecaudata, is less abundant compared with D. rotundus and reportedly difficult to maintain in captivity (Greenhall, 1976; Villa-R., 1967) although Hoyt and Altenbach (1981) maintained them successfully on a diet of blood from live chickens. D. ecaudata remains largely unstudied. Similarly, relatively little is known about the biology of the whitewinged vampire bat, Diaemus youngi (Greenhall and Schutt, 1996). As with D. ecaudata, there is consensus that D. youngi prefers avian blood (Gardner, 1977; Goodwin and Greenhall, 1961; Sazima and Uieda, 1980; Uieda et al., 1992). There are few reports on the maintenance of D. youngi in captivity, and bats in these studies were not transported out of the country of origin. Two adult females, captured and maintained in Brazil, were allowed to feed on a diet of live chickens (Uieda and de Araujo, 1987), and four adults fed on chickens, Gallus gallus, domestic pigeons, Columba livia, and collared doves, Streptopelia decaocto (Uieda et al., 1992). Lack of observational data on D. youngi probably can be attributed to two major factors: the relative rarity of D. youngi in the wild and reports of difficulty maintaining D. youngi in captivity. Goodwin and Greenhall (1961) reported that D. youngi Journal of Mammalogy, 80(1):71-81,

2 72 JOURNAL OF MAMMALOGY Vol. 80, No.1 would not accept bovine blood in captivity, whereas Pye (1967) reported that D. youngi took small quantities of bovine blood but did not survive without avian blood. Contrary to Goodwin and Greenhall (1961), we found that captive D. youngi readily accepts bovine blood. Bovine blood facilitates maintenance of captive colonies, as it is relatively easy to obtain in large quantities, can be stored frozen for extended periods of time, and eliminates the requirement for a daily supply of live blood donors. It has become a dietary staple of captive colonies of D. rotundus (Dickson and Green, 1970; Greenhall, 1965; Wimsatt and Guerriere, 1961). We report here observations on the behavior and maintenance in captivity of 12 D. youngi. Many behavioral observations are reported here for the first time, and this is the first detailed account of a successful maintenance program for captive D. youngi. It should be noted that the daily maintenance program we describe was developed over the past 25 years by workers at the National Animal Disease Center, Caroni, Trinidad. Although this work primarily is concerned with D. youngi, much of the maintenance program described here (e.g., transportation and housing) can be applied to D. rotundus. We speculate that captive colonies of D. ecaudata require a diet composed solely of avian blood. Several references provided valuable information on the care of captive D. rotundus (Dickson and Green, 1970; Greenhall, 1976; Greenhall and Schmidt, 1988; Wimsatt and Guerriere, 1961). MATERIALS AND METHODS Twelve adult D. youngi (10 males, 2 females) were collected 17 Februrary-26 July 1993 by mist netting from the following locations in Trinidad (Guayaguayare, Mayaro; Ortoire, Mayaro; Matura, St. Andrew; Raghunan Road, Caroni, Tabaquite, Caroni). All bats were maintained at the National Animal Disease Center, Caroni, Trinidad, before transport to Cornell University, Ithaca, New York, on 27 July Presence or absence of a cartilagenous zone in the epiphyseal region of the fourth metacarpalphalangeal joint was used for determination of relative age (Kunz and Anthony, 1982). For purposes of identification and record keeping, forearms of bats were fitted with individually numbered metal bands soon after capture. Before and during transport, we complied with all rules and regulations concerning collection, export, and import of live animals. Regulations differ from country to country and special care should be taken to insure that all required permits are completed and available for presentation during transit. Before our trip, we contacted the United States Fish and Wildlife Service's Office of Management Authority (4401 North Fairfax Drive, Room 432, Arlington, VA 22203), and we were instructed that upon entering the United States with the bats, we were to declare them and present United States Customs Form Specimens of D. youngi and D. rotundus, including several newly captured individuals, were transported via commercial airline from Port of Spain, Trinidad, to New York City, and then by car to Ithaca, New York. A wooden travel crate was constructed (90 by 90 by 90 cm), which securely held a pair of wire-mesh, stainless steel, cages. Airholes (0.5 cm diameter) were drilled into the sides of the crate, which were conspicuously labeled (e.g., Live Animals, Do Not Open, This End Up). On several evenings before the trip, bats were placed into the wire cages and the cages fitted into the crate to acclimate the animals to the transport container. The bottom of the cage contained an aluminum tray onto which absorbant material (newspaper) was placed. Before sealing the crate for transport, bats were fed and placed together into one of the wire cages (11 D. rotundus were transported in the second wire cage), and cages were fitted into the crate. The crate lid was sealed with wood screws. All 23 of the bats survived transportation from their country of origin; the crate was sealed for 18 h. Captive D. rotundus and D. youngi were maintained at Cornell University from July 1993 to September In September 1996, bats were transported to the Burnett Park Zoo, Syracuse, New York. At Cornell University, access to the bat colony was restricted to research and animal-care personnel. Specimens, separated by genus, were housed in two, stainless-steel, small-animal cages (56 by 51 by 43 cm). News-

3 1999 SCHurr ET AL.-DlAEMUS YOUNG1 IN CAPTIVITY 73 paper was used as an absorbent on the cage floor and changed daily. A warm damp sponge was used to remove blood and excreta from cage walls, and cages were scrubbed with cage cleaning detergent on a weekly basis. All research and animal-care personnel that had any contact with our vampire bats had first undergone a pre-exposure prophylaxis for rabies virus. Although our D. youngi became tame and did not attempt to bite us after their 1 st week of captivity, we continued to use thick, loose-fitting, leather gloves when handling them. Room temperature in our laboratory was C; relative humidity was variable at 20-S0%. Bats were maintained on a cycle of 12 h of indirect flourescent ceiling light ( h) and 12 h of darkness. This schedule approximated the day-night light conditions in Trinidad. We had few problems with illness in our colonies. Occasionally, however, bats presented some or all of the following symptoms: lethargy, dehydration, bloated abdomen, sneezing, and vomiting of a previous blood meal. We treated our bats in the following manner. At the first sign of illness, the bat was isolated in a wiremesh travel cage draped in a clean towel to reduce drafts. Ambient temperature was increased (ca. SoC) with an electric space heater. If dehydration was observed, the bat was rehydrated daily with a 1.0-ml subcutaneous injection of 0.9% saline solution. Animals were treated as dehydrated if the loose skin covering the region between their shoulder blades did not spring back into place immediately after being gently pinched by the handler. An antibiotic (10 mg Baytril/kg-Bayer Corporation, Shawnee Mission, KS) was administered subcutaneously, once per day. Subcutaneous injections were made beneath the loose skin of the upper back. In the evening, the bat was allowed to feed on a live chicken. This treatment regime proved extremely successful, and all bats treated in this manner recovered within S days. At that time, treatment was discontinued, and the bat was reintroduced into the colony. Our feeding regime for D. youngi was identical to one that had been developed and successfully employed to maintain bats in Trinidad (Muradali et ai., 1993). Specifically, bats were fed bovine blood six times per week and offered live chickens once per week. Contrary to previous reports (Goodwin and Greenhall, 1961; Pye, 1967), our D. youngi readily accepted defibrinated bovine blood. That blood was obtained monthly at a local slaughterhouse from freshly killed domestic cattle. A clean, plastic, 7.S-1 container was used to collect blood. To prevent clots during storage or feeding, blood was fibrinated immediately by manual agitation. A metal kitchen colander was used to skim off clots, and the newly defibrinated blood was stored for transport in plastic jars. In the laboratory, the blood was transferred to individual 473-ml glass bottles. We stored enough in each bottle to feed each bat 30 mi. That volume was more than sufficient because it was about twice the amount normally consumed by our bats. Bottles of blood were frozen at -20 C until use-keeping the bottle's lid loose to prevent glass bottles from cracking when frozen. On the day that the blood was used, a bottle was thawed slowly, and blood was poured into a plastic icecube tray and served at room temperature. We filled one ice-cube tray depression (ca. 30 ml) with blood for each bat. Because D. youngi feeds arboreally in the wild (Gardner, 1977; Goodwin and Greenhall, 1961; Sazima and Uieda, 1980; Uieda et ai., 1992), we elevated the feeding tray ca. 30 cm off the cage floor with a wooden block. Bats were fed in the late afternoon or evening, and the tray was removed and cleaned the next morning. In our study of feeding capacity, the average daily intake of blood was determined, over a 30- day period, for isolated D. youngi. Eight individuals were used (four times each) in 32 trials. Thirty milliliters of blood was offered in a spillproof feeding container. The next day, the amount of blood remaining was subtracted from the starting amount to yield the amount consumed. A correction factor of 2 ml was applied to account for volume loss due to evaporation and blood that clung to the bat during feeding. Although our bats were placed into a stainless-steel feeding cage with the chickens, >SO arboreal feeding sessions were carried out in a 1.8 by 3.0-m screened enclosure, with chickens roosting on a branch. During preparation for those sessions, a chicken was placed onto the upper section of a branch (inverted L-shape, ca. 1.3 m in length, 4 cm in diameter), the lower section of which was attached to a stand constructed with two sections of S- by lo-cm pine studs. With the room lights lowered or turned off, a single D. youngi was placed on the lower

4 74 JOURNAL OF MAMMALOGY Vol. 80. No.1 part of the branch (ca. 15 cm off the floor) in a head-up position. One evening per week, two specimens of D. youngi were placed in a stainless-steel feeding cage (55 by 50 by 42.5 cm). A single subadult chicken (9-10 weeks old) was placed in the cage. Bats were permitted to feed on the chicken overnight. For observations of terrestrial feeding behavior a large wire-mesh cage (51 by 84 by 37 cm) was used. Twenty terrestrial feeding sessions were observed, during which a single bat was placed into the cage with one adult or subadult chicken. Because there were no perches in the cage, the chickens stood on the cage floor. Room lights were turned off and a small flashlight fitted with a red-gel filter was used to observe terrestrial feeding behavior. Sex and age of chickens and the identification number of the bat used were noted. Behavior during those sessions was recorded in a notebook. RESULTS AND DISCUSSION Wimsatt et al. (1973) designed a cage for vampire bats with a novel and time-saving sanitation system, but satisfactory housing can be provided with less elaboration. Basic requirements for cages are that they prevent bats from escaping, allow bats to move about and roost comfortably, and be easy to clean. Because excreta is tar-like and corrosive, housing constructed of stainless steel or plastic is advisable. In a previous study (Uieda and de Araujo, 1987), D. youngi was maintained at temperatures similar to those in our laboratory. In Trinidad, our bats were caged in a open-air facility where temperatures generally were C. Greenhall (1965) maintained D. rotundus and D. youngi in Trinidad at similar temperatures. Although it has been reported that vampire bats require high relative humidity (60% minimum) in captivity (Joermann, 1988), D. youngi does not appear sensitive to low humidity, and we have maintained D. youngi and D. rotundus at relative humidities of 20-50%. This agrees with previous data on maintenance of captive specimens of D. rotundus (Wimsatt and Guerriere, 1961) and indicates that captive vampire bats may not be as sensitive to relative humidity as other species of bats kept in captivity (e.g., vespertilionids). Racey (1972) reported that low humidities can be injurious to wing membranes. Although wing injuries have not been noted in our laboratory, humidity should be given careful consideration in studies where bats are required to fly extensively. Ectoparasite infestation was observed in newly captured vampire bats, but this problem disappeared within a month. As reported in previous studies (Dickson and Green, 1970; Wimsatt and Guerriere, 1961), this problem disappeared gradually over the course of several weeks, probably due to the disruption of ectoparasite lifecycles in captive conditions (Wimsatt and Guerriere, 1961). Information on ectoparasites and endoparasites of D. rotundus exists (Mendez, 1988), but data are fragmentary with regard to D. youngi (Mendez, 1988). Cultures of the bovine blood were tested (Cornell University, Department of Pathology) and indicated presence of bacteria that may have caused occasional gastrointestinal distress observed in several of our bats. Tests to determine presence of Salmonella were negative. Data on the demeanor of vampire bats in captivity is variable. Some authors have found, as we have, that D. rotundus does not tame in captivity (Dickson and Green, 1970), while others report that D. rotundus tames readily (J. S. Altenbach, pers. comm.; Novick, 1963; Wimsatt and Guerriere, 1961). D. youngi has been described as aggressive in captivity (Uieda and de Araujo, 1987), while we have found the opposite to be true. Like D. rotundus, D. youngi appears to be gregarious by nature and can be housed in a group dwelling. There have been no problems maintaining mixed-sex populations, although D. youngi does appear to exhibit dominance hierarchies. The following behaviors were noted in Trinidad, when a new animal was introduced into an established group and at Cornell University,

5 1999 SCHUTT ET AL.-D1AEMUS YOUNG1 IN CAPTIVITY 75 the day after their arrival. In both instances, a male that had been captured recently was approached by another male. These bats are hereafter referred to as the newcomer and the dominant male, respectively. The bats faced each other and both displayed a repertoire of threat behavior. After rising up into an elevated stance, a distinctive hissing was accompanied by other audible vocalizations varying in frequency and intensity. Forelimbs, held in a closed-wing position, were used for parrying. That behavior, which was accompanied by feints, lunges, and head bobbing, lasted ca. 20 s. In the Trinidad episode, the newcomer then crouched quietly, partially opened his mouth and exposing a pair of enlarged oral glands. The dominant male then inserted his tongue into the mouth of the newcomer for several seconds. At Cornell University, after a similar face-off (in this instance there was no tongue insertion), the dominant male approached the newcomer from the rear. With both bats vocalizing loudly, the dominant male partially extended a wing, wrapped it around the back of the newcomer, and climbed over it. This took several seconds, after which time the bats ceased their vocalizations and moved away from each other. That stylized fighting behavior was observed on a number of occasions in our lab especially during sessions with pairs of bats feeding on live chickens. Unlike the violent and bloody clashes reported in studies on D. rotundus during roost-defense encounters (Wilkinson, 1985, 1988), those between D. youngi have not resulted in bloodshed. Diaemus youngi possesses two large oral glands whose function is unknown (Goodwin and Greenhall, 1961; Greenhall, 1988). When disturbed (such as we have seen during dominance-hierarchy behavior or at times during handling), D. youngi opens its mouth, and these large glands can be clearly observed filling the caudolabial part of the oral cavity. A peculiar hissing vocalization, described by Goodwin and Greenhall (1961) as a "pssst," is accompanied by emission of a fine spray of liquid, which has a strong musky odor. We have observed musk spraying only during threat response. We suggest that this aerosol may be used to discourage predators and may serve a role in territory marking and individual recognition. Enlarged oral glands and musk spraying have not been reported in D. rotundus or D. ecaudata. There have been studies on anatomy and histochemistry of the salivary glands of D. rotundus (Di Santo, 1960) and the salivary antihemostatic factors of D. rotundus and D. youngi (Hawkey, 1966, 1967). However, there have been no studies on the anatomy of salivary glands in D. youngi or D. ecaudata. Our vampire bats spent considerable time roosting. During that time, they clung to each other with claws and thumbs of their hind limbs. Hind limbs not in contact with conspecifics generally were used to grasp a branch or the cage ceiling. Number of bats roosting together varied but most often, they clung together in one dense mass. During roosting, we frequently observed an individual apparently biting the bat hanging adjacent and ventral to it. Upon examination, however, we learned that those were not actually bites but rather one bat took a loose fold of skin from a cage mate (usually from the upper back or dorsal surface of the neck, the nape) into its mouth (Fig 1a). There, the skin fold was held securely for several seconds. Bats on the receiving end of that mouth grasp did not appear disturbed. Upon examination of those animals, neither wounds nor hair loss were observed. This behavior has not been reported in other studies on D. youngi in the wild (Sazima and Uieda, 1980; Uieda et ai., 1992) or in captivity (Uieda and de Araujo, 1987; Uieda et al., 1992). Mouth grasping is most reminiscent of behavior seen in non-fiying neonatal bats as they cling onto their mother's body with their mouths, thumbs, and hindlimbs. We suggest that mouth grasping is a component of dominance-hierarchy behavior. A. Stem (pers. comm.) reported similar behavior in

6 76 JOURNAL OF MAMMALOGY Vol. 80, No.1 FIG. I.-a) Mouth-grasping behavior by Diaemus youngi. While roosting or during dominance-hierarchy behavior, individuals often use their mouths to grasp cage mates by the loose skin between the shoulders or on the nape of the neck. b) D. youngi mounting a young female chicken dorsally prior to feeding. Rather than becoming agitated, female chickens mounted in this fashion assume a crouched posture similar to that observed during copulatory behavior; we suggest that the position and body weight of the bat trigger submissive behavior in female chickens. Pteropus vampyrus during dominance fights and copulations in several species of Pteropus. Additionally, mouth grasping appears to assist D. youngi in stabilizing its body during roosting, especially during shifts in position of hind limbs. Mouthgrasping behavior was not observed in our D. rotundus nor has it been reported in other microchiropterans (K. E Koopman, pers. comm.). In previous studies on D. youngi in captivity (Uieda and de Araujo, 1987; Uieda et al., 1992), live birds were used as the sole source of blood. Although our bats readily accepted defibrinated bovine blood, we believe that citrated blood (Wimsatt and Guerriere, 1961), also would be acceptable. Because these bats feed primarily on avain blood in the wild, we chose to provide them with a weekly supplement of blood from a live chicken. While feeding from an elevated ice-cube tray, bats hung from the cage ceiling by their hind limbs and extended their upper bodies toward the tray (until their mouths were just above the level of the blood). As the bat fed, the tongue was extended into the blood and withdrawn at a rate of ca. 4 licks/so Feeding usually lasted ca. 15 min and fighting was never observed during those sessions. Observations on feeding capacity of D. rotundus in captivity (Wimsatt and Guerriere, 1961) indicated that average daily intake of blood by two groups was 15.3 and 15.6 mi. In our study, mean daily intake for D. youngi was 16.4 ± 3.6 (SD) mliday, but there were differences in the method of those two studies (e.g., the study of D. rotundus was carried out over a full year but animals were only fed 6 days/week). Wilkinson (1984) noted that D. rotundus did not necessarily feed every night and some bats received blood regurgitated by other individuals. Blood sharing has not been reported in D. youngi. Sazima and Uieda (1980) reported on the feeding behavior of D. youngi on free-ranging poultry (chickens, Guinea fowl, and turkeys), but were unable to observe the actual biting behavior and the prey's reaction. Uieda and de Araujo (1987) documented some aspects of maintenance of D. youngi and D. ecaudata in captivity but did not go into great detail concerning feeding behavior. Uieda et al. (1992) studied arboreal feeding on chickens and smaller birds (pigeons and doves). They described two feeding postures used by bats, hanging and quadrupedal, and stressed the importance of diameter of branch as a limiting factor in attacks by D. youngi on small birds.

7 1999 SCHUTT ET AL.-DlAEMUS YOUNG1 IN CAPTIVITY 77 The following description characterizes a typical arboreal feeding session in our laboratory and in Trinidad (Muradali et ai., 1993). Holding on with its thumbs and hind feet, D. youngi clings quietly with its ventral body surface held tightly to the branch. During this period, which lasts :::; 10 min, the bat raises its head occasionally, sniffs at the air with its mouth partially open, and observes the chicken perched above it. At this time, the bat moves to the underside of the branch and slowly climbs closer to the perching chicken. D. youngi advances one limb at a time (with the grip on the branch maintained by the three non-moving limbs) always keeping the branch between itself and the underside of the chicken. After positioned beneath the chicken's digits (ca. 1 min), the bat situates itself so that its head is close to a single digit of the roosting bird. On most occasions, the bat positioned its head next to the posteriorly directed hallux (digit 1). It appears that feeding from this digit keeps the bat concealed under its prey to a greater degree than would occur if an anteriorly directed digit (digits 2-4) was fed upon. Moving its mouth to the digit, the bat's tongue is extended and withdrawn (in 'a piston-like manner), and the potential bite area is licked at a rate of ca. 2 licks/so It is not known if saliva of vampire bats contains anesthetic or enzymatic components to aid in preparation of bites or if licking softens the keratinized reticulate scales that cover the surface of the digit. The bitepreparation period lasts from 10 s to 2 min. Bites were never inflicted without this preparation period. In all of our observations, immediately prior to the bite, D. youngi turns its head so that the midline of the head is perpendicular to the long axis of the digit being bitten. The actual bite is never initiated quickly or violently but occurs when licking stops and the bat applies its partially opened mouth to the site. At this time, it appears that the jaws close, driving the razor-sharp upper incisors through the epidermis and into the dermis. This position is maintained for 1-2 s until one of the following events occurs to complete the bite. Either a series of quick backward steps are made by the bat or the chicken shifts its position slightly on the branch and the digit is pulled away from the bat's mouth. In the majority of bites, chickens show little, if any indication that they have been disturbed by the bite. Bites appear as small crater-shaped indentations that extend into the dermis-a layer richly supplied by blood vessels. Bites inflicted by D. rotundus appear similar and have been well characterized (Greenhall, 1988). In D. rotundus, canine teeth are used to shave the fur around a wound site. The lobed lower incisors secure a firm hold on the skin while the concave upper incisors remove a divot of flesh. D. youngi may use its canines in a similar manner to shave feathers, but this has not been reported and we did not observe this behavior. After the bite is inflicted, D. youngi returns to the wound and resumes the licking behavior at a rate of ca. 4 licks/so Within 30 s of the bite, blood flows freely from the wound. Blood is conducted into the bat's mouth via two grooves on the ventral surface of the tongue (Uieda, 1986) through a gap between the two pairs of lower incisors, and above a V-shaped groove in the lower lip. Most feeding sessions last ca. 15 min and may stop temporarily if the bird changes positions or becomes alarmed. When this occurs, the bat may back off, concealing itself under the branch, until the bird calms. On other occasions, the bat remains in position and appears to hide under the body of the bird. After the bird settles down, the bat re-establishes its position and begins feeding. At no time (during either arboreal or terrestrial feeding behavior) did we observe a feeding bat make a second bite on the same chicken. Interruptions in feeding always were followed by a resumption of feeding from the same wound. We have regularly observed periods of urination, lasting several seconds, commencing at ca. 5 min into the feeding session. While urinating, one hind limb usually is extended

8 78 JOURNAL OF MAMMALOGY Vol. 80, No.1 laterally and slightly downward, preventing the bat from soiling itself. On several occasions, bites made on the terminal digital pad of digit 1 (ventral to the proximal portion the claw) were not licked during feeding. Instead, the protruding digital pad was taken into the bat's mouth as it fed. On two occasions, we attempted to dislodge the bat from the bird's digit and found that it took a considerable amount of pulling for the bat to release its hold. Our attempts to dislodge the bat did not disturb the chicken, and no further wounds were apparent on them after the bat was pulled free. We suggest that suction (rather then the use of teeth) was used during this peculiar feeding behavior. This observation lends support to the hypothesis that vampire bats may use their mouths to form a seal around the wound when feeding (Greenhall, 1988). Suction and wound licking appear to be two methods for obtaining blood from a wound. Because suction was observed only when D. youngi fed on the distal digital pad, it appears that these feeding methods may be dependent on location of wounds. After ca. 15 min, D. youngi concluded its feeding session by releasing its thumbs from their hold on the branch. After hanging briefly by its hind limbs, the bat released its grip on the branch. In many instances, flight was initiated by this method, but on a few occasions D. youngi dropped directly to the ground before scrambling off to hide. Because of the design of cages, cages used during weekly supplements of diet with chicken blood were not used to observe bat-bird interactions but solely for feeding purposes. In all cases, subadult chickens placed into the feeding cage were removed and euthanized after 4 h. In a previous study (Uieda and de Araujo, 1987), bats and chickens were kept together continuously. Those chickens exhibited decreased activity and weight loss, and a number of them died within 7-10 days. We examined locations of bites on carcasses of chickens and observed that most bites were located on digits or the posterior side of the intertarsal joint. Some bites, however, were made on the comb, neck, or abdomen, and a single bite was found on the proximal prepatagium of the wing. Descriptions of terrestrial feeding behavior by D. youngi are lacking. Uieda et al. (1992) mentioned a single observation of quadrupedal posture employed by D. youngi during feeding on a domestic pigeon, but it is unclear if that occurred during arboreal or terrestrial feeding. The following is a description of a typical terrestrial feeding session in our laboratory. After the bat is placed into the cage, it uses a hopping gait to position itself under the posterior part of the bird's body behind the feet. If the bird moves about the cage, the bat hops after it, trying to maintain its position under and behind the bird. After the bird is stationary, the bat approaches the posterior side of the metatarsus. Raising itself into an elevated quadrupedal stance, similar to that described for D. rotundus (Altenbach, 1979, 1988; Sazima, 1978), a bite-preparation phase is initiated. In most instances, bites are made on the posterior or lateral surface just distal to the intertarsal joint. Larger scales (scutes or scutella) are partially or fully removed during bites to the metatarsus. Bite infliction is similar in terrestrial and arboreal feeding in that the bite is completed if the chicken moves away from the stationary bat or the bat steps backward away from the bird's leg. After blood begins to flow from the wouqd, the bat approaches the wound and begins to feed. If the bird moves around the cage, the bat disengages feeding and follows the bird by hopping behind and underneath it. Feeding resumes when the bird assumes a stationary pose. During a number of terrestrial feeding sessions, we noted agitated and aggressive behavior by male chickens toward a bat. That behavior consisted of rapid pacing back and forth in the feeding cage by the bird accompanied by pecks directed toward the bat. Inevitably, that behavior resulted in

9 1999 SCHUTT ET AL.-D1AEMUS YOUNG1 IN CAPTIVITY 79 the bat retreating to the cage ceiling where it remained until removed. D. youngi were observed feeding on male chickens on many occasions, but they seem to prefer female chickens. In ca. 10% of terrestrial feeding sessions observed, the bat did not feed from the chicken's leg or digit but leaped or climbed onto the chicken's back to feed on the head region. In those instances, the bat positioned itself with its body facing cranially with regard to the bird. That position was maintained by partially spreading (abducting) forelimbs while grasping onto dorsal feathers with both thumbs and digit claws of the hind limb. Male chickens that were mounted dorsally by D. youngi quickly became agitated, often dislodging bats from their back. When young female chickens were mounted dorsally, however, rather than becoming agitated, they assumed a crouched posture (Fig. 1 b) similar to that observed during copulatory behavior. They remained in this position even when moving around the cage. We suggest that the forces applied to the female bird's back by the bat triggers this innate submissive response behavior. Attacks on poultry roosting in trees are frequent in Trinidad, but there have been no reports of attacks on birds confined in cages on the ground although the mesh of these cages would allow entrance by D. youngi. Our feeding experiments, however, showed that D. youngi are quite capable of efficient terrestrial feeding. Examination of stomach contents of D. youngi has, in fact, shown them to feed on mammalian blood (e.g., cattle and pigs-greenhall, 1988). Our observations suggest that although D. youngi is capable of feeding terrestrially, it will not do so as long as arboreal prey is available. One possibility is that arboreal feeding evolved to reduce competition between D. youngi and the terrestrially feeding D. rotundus where their ranges overlap, although Schutt and Altenbach (1997) suggest that arboreal feeding is a primitive behavior in vampire bats. According to Koopman (1988), there is disagreement with regard to the recognition of Desmodus and Diaemus as seperate genera. Handley (1976) for example, included Diaemus within the genus Desmodus. Enumerating 18 morphological characters, Koopman (1988:12) was "inclined to recognize Diaemus as a valid genus along with Diphylla and Desmodus." In our studies, we observed a number of behavioral characters, related to feeding and roosting behavior (e.g., mouth grasping), which support that view. The locomotor morphology of D. rotundus was studied in detail by Altenbach (1979, 1988) and Schutt et al. (1997). Altenbach (1988) described the quadrupedal stance of this genus as elevated. This stance apparently allows "immediate movement in any axis parallel to the surface or movement away from it without repositioning of either pectoral or pelvic limbs" (Altenbach, 1988:72). The elevated posture appears to be an adaptation for rapid terrestrial locomotion in D. rotundus, allowing it, for example, to initiate flight by jumping vertically into the air. Jumping functions to get the heavily loaded bat airborne. It also aids in predator avoidance (e.g., snakes) and may prevent the bat from being stepped on by ungulate prey while feeding. In contrast, D. youngi exhibits a more crouched stance than D. rotundus. The body of D. youngi is held so that its center of mass is closer to the support (the ground or branch). D. youngi has never been reported to perform flight-initiating jumps in the wild or in captivity. Additionally, we never observed this behavior during forceplatform experiments analyzing terrestrial locomotion (Schutt et al., 1997). We suggest that D. youngi, with its arboreal feeding habits, is under less threat from terrestrial predators than D. rotundus. Unlike the terrestrially feeding D. rotundus, D. youngi is in no danger of being crushed by largeungulate prey. During terrestrial feeding sessions with D. youngi, chickens were observed to suddenly squat down on top of a

10 80 JOURNAL OF MAMMALOGY Vol. 80, No.1 feeding bat. No ill effects have been observed in bats in those situations, and on most occasions there was no interruption in feeding. We believe that the stance of D. youngi is less upright than that of D. rotundus for two additional reasons: climbing efficiency and crypsis. As a climber and arboreal hunter, D. youngi has decreased the tendency to topple from its support (the branch) by several means: 1) keeping its center of mass close to the support (this is especially valauble during vertical climbing behavior); 2) climbing underneath the branch; 3) using its thumbs and hind feet to exert a torque that resists the toppling moment (for a review of the biomechanics of climbing see Cartmill, 1985). Additionally, by clinging with its body under and close to the branch, D. youngi is better hidden from the sight of its avian prey. D. youngi, initiates flight by simply dropping from its arboreal feeding position and has no need for the flight-initiating jumps that are characteristic of the terrestrially feeding D. rotundus. In summary, specimens of D. youngi and D. rotundus captured in Trinidad have been maintained successfully in captivity. The only difference in maintenance programs for these two genera is a once per week supplement of live chicken blood for D. youngi. Many differences in behavior, however, have been noted between D. youngi and D. rotundus. We believe that these differences are related to their mode of prey selection. D. rotundus exhibits a number of unique adaptations for terrestrial feeding (e.g., speed, flight-initiating jumps). Although D. youngi is quite capable of terrestrial feeding, there have been no reports of this mode of feeding in the wild, and it appears that D. youngi is better adapted for arboreal feeding. Blood feeding in mammals is restricted to three genera of vampire bats, and these exhibit many unique adaptations that allow them to maintain a sanguivorous lifestyle (e.g., highly modified digestive and excretory systems, quadrupedal locomotion, altruistic behavior). In Trinidad where D. rotundus and D. youngi coexist, competition between these genera may be reduced through resource partitioning. ACKNOWLEDGMENTS This study was made possible by funding from the American Museum of Natural History (Theodore Roosevelt Memorial Fund. Coleman Postdoctoral Fellowship); the American Society of Mammalogists (grant-in-aid). Sigma Xi. and Cornell University. Special thanks to the Ministry of Agriculture, Lands and Marine Resources. Trinidad. The authors thank the following individuals in Trinidad: the vampire-bat crew (A. Johnson and P. Wallace). D. Booda. C. James, S. Johnstone. N. Gyan. H. Nelson, S. Ramdass. J. Latchman, Ali family. and especially G. Ramsawak and O. Ramsawak (our hosts at Mount Saint Benedict's). We thank individuals at Cornell University: J. W. Hermanson. J. Bertram, D. McCleam. H. Evans. D. Cullinane, Y. H. Chang. C. Coen. P. Faure. R Quimby. L. Carbone. H. Bartlett. D. Woodin. C. Grant, and L. L. Lafrance. We thank the Department of Mammalogy at the American Museum of Natural History, especially N. B. Simmons, R. MacPhee, D. Lunde. T. Conway. and K. R Koopman. Thanks also to J. S. Altenbach. R. Dugal. R. Adamo. R. "Tuna" Sinclair. W. Uieda. A. Stem. T. Kunz, M. Schutt. J. Schutt. W. R. Schutt. and Bloomfield College (especially J. Noonan. I. Anderson. M. LaBare, and P. Russo). This paper is dedicated to our friend. A. M. Greenhall, who pioneered the modem study of vampire bats. Without his encouragement. this work would not have been undertaken. LITERATURE CITED ALTENBACH. J. S Locomotor morphology of the vampire bat. Desmodus rotundus. Special Publication, The American Society of Mammalogists, 6: Locomotion. pp , in Natural history of vampire bats (A. M. Greenhall and U. Schmidt, eds.). CRC Press, Boca Raton, Florida. CARTMILL, M Climbing. Pp , in Functional vertebrate morphology (M. Hildebrand, ed.). Harvard University Press, Cambridge, Massachusetts. DICKSON, J. M., AND D. G. GREEN The vampire bat (Desmodus rotundus): improved methods of laboratory care and handling. Laboratory Animals, 4: DISANTO, p Anatomy and histochemistry of the salivary glands of the vampire bat, Desmodus rotundus murinus. Journal of Morphology, 106: DITMARS, R. L., AND A. M. GREENHALL The

11 1999 SCHUTT ET AL.-DlAEMUS YOUNG1 IN CAPTIVITY 81 vampire bat: a presentation of undescribed habits and review of its history. Zoologica, 19: GARDNER, A. L Feeding habits. Pp , in Biology of bats of the New World family Phyllostomatidae, Part II (R. J. Baker, J. K. Jones, Jr., and D. C. Carter, eds.). Special Publications, The Museum, Texas Tech University, 13: GOODWIN, G. G., AND A M. GREENHALL A review of the bats of Trinidad and Tobago: descriptions, rabies infection, and ecology. Bulletin of the American Museum of Natural History, 122: GREENHALL, A M Notes on behavior of captive vampire bats. Mammalia, 29: Care in captivity. pp , in Biology of bats of the New World family Phyllostomatidae, Part I (R. J. Baker, J. K. Jones, Jr., and D. C. Carter, eds.). Special Publications, The Museum, Texas Tech University, 10: Feeding behavior. pp , in Natural history of vampire bats (A M. Greenhall and U. Schmidt, eds.). CRC Press, Boca Raton, Florida. GREENHALL, A M., AND U. SCHMIDT Natural history of vampire bats (A M. Greenhall and U. Schmidt, eds.). CRC Press, Boca Raton, Florida. GREENHALL, A M., AND W. A. SCHUrr, JR Diaemus youngi. Mammalian Species, 533:1-7. HANDLEY, C. 0., JR Mammals of the Smithsonian Venezuelan Project. Brigham Young University Science Bulletin, Biological Series, 20(5):1-91. HAWKEY, C. M Plasminogen activator in saliva of the vampire bat Desmodus rotundus. Nature, 211: Inhibitor of platelet aggregation present in saliva of the vampire bat Desmodus rotundus. British Journal of Haematology, 13: HoYT, R. A, AND J. S. ALTENBACH Observations on Diphylla ecaudata. Journal of Mammalogy, 62: JOERMANN, G Care of vampire bats in captivity. pp , in Natural history of vampire bats (A M. Greenhall and U. Schmidt, eds.). CRC Press, Boca Raton, Florida. KOOPMAN, K. F Systematics and distribution. pp. 7-17, in Natural history of vampire bats (A M. Greenhall and U. Schmidt, eds.). CRC Press, Boca Raton, Florida. KUNZ, T. H., AND E. ANTHONY Age estimation and post-natal growth in the bat, Myotis lucifugus. Journal of Mammalogy, 63: MENDEZ, E Parasites of vampire bats. pp , in Natural history of vampire bats (A M. Greenhall and U. Schmidt, eds.). CRC Press, Boca Raton, Florida. MURADAU, F., N. MONOOL, AND W. A SCHUrr, JR Observations on feeding behavior in the whitewinged vampire bat, Diaemus youngi. Bat Research News, 34:121A NOVICK, A Orientation in Neotropical bats. II. Phyllostomatidae and Desmodontidae. Journal of Mammalogy, 44: PYE, J. D Bats. pp , in The Universities Federation for Animal Welfare handbook on care and management of laboratory animals. Third ed. Livingstone, Edinburgh, United Kingdom. RACEY, P. A Bats. pp , in The Universities Federation for Animal Welfare handbook on the care and management of laboratory animals. Fourth ed. Williams and Wilkins Company, Baltimore, Maryland. SAZIMA, I Aspectos do comportamento alimentar do morcego hematofago, Desmodus rotundus. Boletim de Zoologia, Sao Paulo, 3: SAZIMA, I., AND W. UIEDA Feeding behavior of the white-winged vampire bat, Diaemus youngi, on poultry. Journal of Mammalogy, 61: SCHUrr, W. A, JR., AND J. S. ALTENBACH A sixth digit in Diphylla ecaudata, the hairy legged vampire bat. Mammalia, 61: SCHUrr, W. A, JR., ET AL Functional morphology of the common vampire bat, Desmodus rotundus. The Journal of Experimental Biology, 200: TRAPIDO, H Observations on the vampire bat with special reference to longevity in captivity. Journal of Mammalogy, 27: UIEDA, W Aspectos da morfologia lingual das tres especies de morcegos hematofagos (Chiroptera, Phyllostomidae). Revista Brasileira de Biologia, 46: UIEDA, W., AND V. F. DE ARAUJO Mantencao dos morcegos hematofagos Diaemus youngi e Diphylla ecaudata (Chiroptera, Phyllostomidae), em cativeiroo Anais Seminario Ciencias da Fiube, Uberaba, 1: UIEDA, W., S. BUCK, AND I. SAZIMA Feeding behavior of the vampire bats, Diaemus youngi and Diphylla ecaudata, on smaller birds in captivity. Ciencia e Cultura, 44: VILLA-R., B Los murcielagos de Mexico: su importancia en la economfa y la salubridad-su clasificaci6n sistematica. Instituto de Biologfa, Universidad Nacional Aut6noma de Mexico, Mexico, Distrito Federal, Mexico. WILKINSON, G. S The social organization of the common vampire bat. II. Mating system, genetic structure and relatedness. Behavioral Ecology and Sociobiology, 17: Social organization and behavior. pp 85-97, in Natural history of vampire bats (A M. Greenhall and U. Schmidt, eds.). CRC Press, Boca Raton, Florida. WIMSATT, W. A I 962a. Observations on the feeding capacities and excretory functions of captive vampire bats. Journal of Mammalogy, 43: b. Responses of captive common vampires to cold and warm environments. Journal of Mammalogy, 43: Vampire bats. Pp , in Zoo and wild animal medicine (M. E. Fowler, ed.). Saunders Press, Philadelphia, Pennsylvania. WIMSATT, W. A., AND A. GUERRIERE The maintenance of the common vampire Desmodus rotundus murinus in captivity. Journal of Mammalogy, 42: WIMSATT, W. A, A GUERRIERE, AND R. HORST An improved cage design for maintaining vampires (Desmodus) and other bats for experimental purposes. Journal of Mammalogy, 54: Submitted 19 August Accepted 16 June Associate Editor was Troy L. Best.