Vocal communication in the white-winged vampire bat (Diaemus youngi) (Thesis format: Monograph)

by Gerald Gunnawa Carter

Graduate Program in Biology A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Faculty of Graduate Studies The University of Western Ontario

London, Ontario, Canada December, 2007 © Gerald Gunnawa Carter, 2007

THE UNIVERSITY OF WESTERN ONTARIO FACULTY OF GRADUATE STUDIES

CERTIFICATE OF EXAMINATION Supervisor

Examiners

______________________________ Dr. Brock Fenton

______________________________ Dr. Paul Handford

Supervisory Committee

______________________________ Dr. Beth MacDougall-Shackleton

______________________________ Dr. Paul Handford

______________________________ Dr. Robert Dean

______________________________ Dr. Scott MacDougall-Shackleton

The thesis by Gerald Gunnawa Carter entitled:

Vocal communication in the white-winged vampire bat (Diaemus youngi) is accepted in partial fulfillment of the requirements for the degree of Master of Science

Date__________________________

_______________________________ Chair of the Thesis Examination Board

ii

Abstract Mother bats use pup isolation calls for recognition and reunion. The functions of social calls emitted by adults, however, are poorly understood. I studied vocal communication in 17 captive white-winged vampire bats (Diaemus youngi). In this species, adults emit social calls that are structurally similar to the pup isolation calls found in many other species. In addition, wild-caught bats seem to exchange these calls in a duet-like fashion. To determine the function of these social calls and this antiphonal calling pattern, I employed isolation and playback experiments, simultaneous recording with a 4-microphone array, and permuted discriminate function analyses (pDFA) of individual variation in call structure. I then used a habituation-discrimination playback experiment to determine if bats can discriminate individuals based on social call alone. I found that bats called within 500 milliseconds of a conspecific significantly more than expected based on a random chance model. Several results suggest that antiphonal calling allows recognition and localization of specific individuals. Bats call when isolated, and their calls attract conspecifics. I found significant individual variation in call structure, and showed that bats can discriminate individuals based on playback of calls alone. This result is the first evidence of individual vocal recognition among adult bats. In D. youngi, I suggest that antiphonal contact calls are used to mediate social interactions between individuals outside the roost.

Key words: bat, Diaemus youngi, white-winged vampire bat, social call, vocal communication, antiphonal calling, individual discrimination, recognition, playback

iii

For Michelle

iv

Acknowledgements To me, bats are the most dazzling, magical creatures. I am first grateful to them for bringing me so much awe and utter joy throughout the years. My other acknowledgements are full of fascinating people who have provided resources, and given me support. Many are the kind of scientists who make me think science is awesome; the rest are the kind of friends who make me think the same about people. The person who has helped me the most with this project is Daniel Abram, who has the captive colony of vampire bats. None of this work would be possible without him and Talking Talons Youth Leadership. I could never thank him enough for all the generosity or all the things I learned from him. He and Laurie Wearne have been terrific friends and are truly inspiring people. They demonstrate the epitome of commitment to a worthy cause. Talking Talons is an impressive nonprofit that forever changes the lives of people and animals in profound ways. My first co-advisor, Brock Fenton, conveniently recognizes that bats are the most fascinating things on the planet. I feel honoured to join the long and impressive line of his former students, all impacted by his infectious enthusiasm. Paul Faure, my other co-advisor, has been a great friend and inspiration. Thanks to Dan Riskin, Daniela Rambaldini, and Claudia Coen for having a large positive influence on me. Mark Skowronski has been my ‘go-to guy’ at the Bat Lab. He and Roger Mundry helped me with my data analysis. I thank John Ratcliffe, Paul Handford, Rob Dean, Beth and Scott MacDougall-Shackleton for advice, the Rosamund Gifford Zoo in Syracuse, and the entire Bat Lab for support. An Ontario Graduate Scholarship and an NSERC Grant to B. Fenton provided funds.

v

Michelle Nowak has supported me more than anyone while writing the thesis. She helped me in the field, edited my manuscript, gave me advice and ideas, encouraged me tirelessly, and made me laugh everyday. Kristin Nowak, Carter Takacs, and Skylar have been a most welcoming second family while I wrote this thesis. Katy Griffin and Kristal and Luc DuBois kept me sane while I was in London. Finally, I thank my parents. Kin altruism is an amazing force; when I consider how much unconditional support have parents given me over the years, I get all tearyeyed. I truly appreciate their efforts.

vi

Table of Contents

ii. Certificate of Examination iii. Abstract and Keywords iv. Dedication v. Acknowledgements vii. Table of Contents ix. List of Tables x. List of Figures

Chapter 1. Introduction 1.1 Vocal communication in bats: a literature review

1

1.2 Why do they duet? Antiphonal calling behaviour in animals

10

1.3 Antiphonal calling among adult bats: the case of the white-winged vampire

12

Chapter 2. Methods 2.1 Animals

21

2.2 Sound equipment 2.3 Isolation experiments 2.4 Preliminary playback experiments

22

2.5 Antiphonal calling experiment

26

2.6 Individual variation in call structure

28

2.7 Vocal discrimination experiment

31

vii

Chapter 3. Results 3.1 Isolation experiments

36

3.2 Preliminary playback experiments 3.3 Antiphonal calling experiments

37

3.4 Individual variation in call structure

41

3.5 Vocal discrimination experiments

Chapter 4. Discussion 4.1 Temporal pattern of antiphonal calling

46

4.2 Possible functions of antiphonal responses 4.3 Function of double-syllable calls

47

4.4 Evidence for individual recognition

51

4.5 Possible adaptive benefits of long-distance recognition of individuals

53

4.6 Conclusions

57

References

58

Appendix

67

Curriculum Vitae

68

viii

List of Tables Table 1. Variables and transformations used in discriminant function analyses

29

Table 2. Playback sequences included in statistical analyses

32

Table 3. Results of playback experiments

44

Table 4. Hypotheses for call function

48

ix

List of Figures Figure 1. Spectrograms of double-note calls from several different species of bats

8

Figure 2. Screech call (contact call) of adult Phyllostomus hastatus

9

Figure 3. Spectrograms of 5 basic call types of Diaemus youngi

15

Figure 4. Spectrograms of social call types from 6 European bat species

18

Figure 5. Typical playback sequence

33

Figure 6. Frequency distribution of conspecific responses within 5 seconds of a social call

37

Figure 7. Percentage of social calls that occurred as antiphonal responses (within 0.5 seconds of a conspecific call) in 9 bats

40

Figure 8. Mean responses to periods of a habituation-discrimination playback sequence.

43

Figure 9. Difference in signal strength between double-syllable social calls and echolocation pulses in Diaemus youngi

50

Figure 10a. Initiation of reciprocal food sharing in Diaemus youngi

55

Figure 10b. Food sharing at the wound by two females

x

Chapter 1. Introduction

1.1 Vocal communication in bats: a literature review Bats are arguably the most vocal animals. Consider that echolocating bats call almost constantly while flying. They use a tremendous frequency range (1 to >200 kHz) and produce the most intense airborne sounds of any animal (e.g. Holderied et al. 2005, Holderied & von Helversen 2003). In the lab, bats have been model organisms for the study of neural mechanisms of audition (e.g. Kanwal & Rauschecker 2007), and the sensory feedback control of vocalizations (reviewed by Smotherman 2007). Many bat calls are used for communication rather than biosonar. Indeed, when bats are active, their roosts become filled with chittering or squawks. Yet vocal communication in bats has received relatively little attention. A query on Google Scholar, an online academic paper search engine, for the term “vocal communication” returns 5,380 hits. “Vocal communication” and “birds” returns 2,370. Using “primates” instead of “birds” gives us 2,410 hits. Replacing “primates” with “bats”, we get 693. Bats (order Chiroptera: ca. 1,100 spp.) are the second most speciose, and perhaps the most ecologically diverse Mammalian order; yet, “vocal communication” and “baboons” (genus Papio: 5 species) yields an almost equal count of 692 hits. Not surprisingly, most bioacoustics work on bats has focused on biosonar calls (a search for “echolocation” and “bats” yields 5,820 hits). Non-echolocation vocalizations, or “social calls”, have been understudied for several reasons. The social aspects of bat behaviour themselves are largely mysterious, partly because they

1

are elusive and nocturnal. Researchers require special tools such as infrared lights and microphones sensitive to ultrasound to monitor social interactions of bats. Like nonvolant mammals, bats use other modalities such as touch and smell to communicate at close range. Close range vocal signals can be low intensity and difficult to record, even in captivity. For example, a “hum” or purring (see Owings & Morton 1998 for a review of purring in mammals) emitted during mother-pup interactions has been recorded in only one study (Schmidt-French et al. 2006), which required touching the microphone to the bat. From a distance, it can be difficult to discern when sounds are being used for echolocation, communication, or perhaps both. Whether intentional or not, echolocating bats constantly advertise their positions while in flight. Echolocation pulses can transfer information two ways. During eavesdropping, bats respond to conspecific biosonar (e.g. Fenton 1985, Balcombe and Fenton 1988, Wilkinson 1992, Masters et al. 1995, de Fanis & Jones 1996, Kazial & Masters 2004, Ruczynski et al. 2007, Gillam 2007). Bats use echolocation calls as intentional signals when mothers coordinate echolocation pulses to pup calls (Matsumara 1981, Brown et al. 1983). In many instances, distinguishing biosonar from social signals using call design alone can be difficult (Schwartz et al. 2007) or impossible (e.g. Kanwal et al. 1994). Such issues further complicate the study of vocal communication in bats. Comparisons with other taxa- Mere paucity of data does not make a subject worthwhile. Bats also provide much insight for the evolutionary biologist. Comparing vocal signal diversity and design in bats with vocal communication in birds, cetacean, and primates is revealing. Comparative analyses can identify the

2

environmental factors selecting for various aspects of vocal behaviour. Bats and birds share some selective pressures on their vocal communication systems. Flight makes both highly vagile. Many species of bats and birds migrate long distances at night, or travel and forage in groups. Some birds live in caves and have even independently evolved echolocation (e.g. cave swiftlets Collocalia spp., and oilbirds Steatornis caripensis: Griffin 1953). In some bats, males use complex songs to advertise male quality much like songbirds (e.g. sac-winged bats Saccopteryx bilineata: Behr & Helversen 2004, Davidson & Wilkinson 2004, Behr et al. 2006). Some cetaceans and most bats use sound for both communication and echolocation, which might exert opposing selective forces on behaviour or morphology. Are there evolutionary trade-offs between the ability to “see” things and “say” things with sound? Available evidence to date suggests that bats partition these roles without making significant sacrifices in either. For example, hearing sensitivities in bats correspond to the ranges of both echolocation and pup social calls (Bohn et al. 2006), and evidence suggests that separate brainstem areas are designated to controlling echolocation versus communication (Fenzl & Schuller 2007). Some bats are similar to primates in demonstrating some complex social behaviours such as reciprocal food sharing (Wilkinson 1984), adult vocal learning (Esser 1994, Boughman 1998), infant babbling (Knornschild et al. 2006), or the kind of social learning required for cultural transmission (Page & Ryan 2006). According to a social complexity hypothesis for communication, larger and more stable social groups selected for more complex vocal communication, and eventually language in humans (e.g. Dunbar 2003, Pinker 2003, McComb and Semple 2005, Freeberg 2006).

3

Supporting evidence that social complexity drives vocal complexity comes from both comparative data on primates (McComb and Semple 2005) and experimental work on birds (Freeberg 2006). After reviewing the literature on bats, Wilkinson (2003) found a relationship between pup call complexity and colony size among different species, but also tentatively concluded that social complexity in bats seems to have no real correlation to vocal complexity based on the limited available evidence. Is vocal complexity in bats better predicted by a different factor such as ecological niche or phylogeny, or do species that are more social actually have greater vocal complexity, as expected? More comparative data on bats are needed. Structure and function are interconnected themes in biology. Much work describes structural variation of bat social call repertoires (e.g. Nelson 1964, Fenton 1977, Kanwal et al. 1994, Russo & Jones 1999, Christesen & Nelson 2000, Kingston et al. 2000, Leippert et al. 2000, Andrews & Andrews 2003, Ma et al. 2006, Melendez et al. 2006). Some studies report a tremendous diversity of syllable types, syllable combinations, and rules of syntax (Kanwal et al. 1994, Ma et al. 2006). Fewer cases, however, attempt to investigate function. For instance, only one recent vocal repertoire study (Bohn et al., submitted) reports the behavioural context of calls. Function of social calls- Call structure is evolutionarily selected not simply to convey information, but rather to influence others’ behaviour in ways that benefit a signaler’s reproductive success (Dawkins & Krebs 1978, Owings & Morton 1998). Yet adaptive functions have been demonstrated, or speculated upon, in only a few cases, listed below. Male songs in Saccopteryx bilineata (Davidson & Wilkinson 2004, Behr et al. 2006) advertise quality and attract mates. Male songs are also

4

reported in Pipistrellus pipistrellus, P. pygmaeus and P. nathusii (Lundberg & Gerell 1986, Russ and Racey 2007). Isolation calls by dependent pups (e.g. Gould 1973, 1977, Thomson et al. 1985, Gelfand & McCracken 1986, Esser & Schmidt 1989, Scherrer & Wilkinson 1993, Balcombe 1990, Bohn et al. 2007), and directive calls by mothers (e.g. Gould 1977, Esser & Schmidt 1989, Balcombe 1992, Zhang et al. 2005) facilitate reunions. Social calls in Pipistrellus pipistrellus repel competing conspecifics from food patches (Barlow & Jones 1997a). Similarly, aggressive calls of Desmodus rotundus (Sailler & Schmidt 1978), and probably other species (e.g. Myotis lucifugus Barclay et al. 1979), are likely used to convey willingness to fight over resources. Screech calls in Phyllostomus hastatus coordinate group movements and recruit groupmates, probably for monopolizing food resources (Wilkinson & Boughman 1998). There is anecdotal evidence that “distress calls” may function to startle predators (Nagel 2006), or elicit mobbing from reciprocal altruists (Russ et al. 2004). As in most cases, experiments are needed to test assumptions made from observation. Structure and function of bat social calls are not always correlated in an obvious way. For example, in P. pipistrellus, both sexes emit agonistic social calls to repel conspecifics, while males emit “songflight” calls during the mating season. Yet surprisingly, although these calls have different functions, they are almost identical in structure (Barlow & Jones 1997b). One possible explanation is that social call function might depend more on context than on structure (Barlow & Jones 1997b). Perhaps a single call type that conveys identity or dominance rank can be used

5

differently in various social situations. As Rendall and Owren (2002) remark, who is saying it might be more important than what is being said. Contact calls and the “double-note”- Another example of an apparent mismatch between structure and function is the contact call. Three kinds of contact calls have been described: isolation calls, directive calls, and Phyllostomus hastatus screech calls. In most (possibly all) bats, pups emit isolation calls (Gould 1977, Wilkinson 2003, Bohn et al. 2006). In a few species, directive calls are emitted by mothers searching for young (Brown 1976, Gould 1977, Fenton 1985), and anecdotal evidence suggests they may also function as contact calls between adults (Brown 1976, Fenton 1985). Both signals often have a “double-note” structure- usually two overall downward-sweeping tonal syllables with multiple harmonics (Fig. 1, Gould et al. 1973, Gould 1977, Brown 1976, Barclay et al. 1979, Porter 1979, Brown et al. 1983, Fenton 1985, Sterbing 2002, Bohn et al. 2007). Such frequency-modulated, multisyllabic, harmonically-rich calls are a common design for recognition signals (Bradbury & Vehrencamp 1984). On the other hand, screech calls in P. hastatus are the only clearly-demonstrated example of a contact call so far reported among adults; the calls are used for recognition and group cohesion (Boughman 1997, Wilkinson & Boughman 1998, Boughman & Wilkinson 1998). However, these calls are completely atonal (Fig. 2)- a surprising structure for a contact call, and quite the opposite of a double-note call. Screech calls, albeit relatively “simple” in structure, are still complex enough to function as badges of group membership (Boughman & Wilkinson 1998). Their “noisy” design may instead facilitate long-distance travel and localization (Boughman 1997.

6

7

Figure 1. Spectrograms of double-note calls from several different genera of bats. A. Carollia pup; B. Leptonycteris pup; C. Phyllostomus pup; D. Molossus pup; E. Eptesicus mother; F. Myotis mother; G. Macrotus pup; H. Antrozous pup; I. Desmodus pup (adapted from Gould 1973); J. Phyllostomus hastatus pups at 0 days and 10 days (taken with permission from Bohn et al. 2007). X-axis shows time; in A through I, Y-axis shows frequency (for spectrograms) and relative intensity (thin lines).

J

8

Figure 2. Screech call (contact call) of adult Phyllostomus hastatus. Taken with permission from Boughman 1998.

The vocal communication system of group-living P. hastatus (family Phyllostomidae) is better understood than in any other bat species. Isolation calls are individually distinct double-syllable frequency-modulated (FM) signals, and individual pups are discriminated from each other by adults (Bohn et al. 2007). Adults, however, produce group-specific screech calls, so listening bats can discriminate groupmates from non-groupmates, but apparently cannot recognize individuals (Boughman & Wilkinson 1998). In contrast, adults in many other bat species produce multisyllabic FM social signals, rather than atonal screech calls. Here, I report on vocal communication in another group-living phyllostomid where “double-note” FM signals are common in adults.

9

1.2 Why do they duet? Antiphonal calling behaviour in animals Vocal exchanges in animals draw a lot of attention from biologists. They appeal to our ears, perhaps because they sound like conversations (e.g. Snowdon & Cleveland 1984). In most cases however, the functions of coordinated vocal exchanges, such as antiphonal calling and duetting, are elusive. Consequently, some researchers define these terms by their temporal characteristics, rather than by participants or putative functions (Langmore 2002). Under such definitions, antiphonal calling is the interactive exchange of calls in alternating sequences (Soltis et al. 2005), and vocal duetting is either simultaneous or antiphonal calling that occurs repeatedly at consistent intervals with stereotyped structure (Langmore 2002, Hall 2004). Other researchers define “duetting” as vocal exchanges between males and females, whereas “antiphony” occurs between any two individuals (Yoshida & Okanoya 2005). Multiple definitions have led to these terms being used both differently and interchangeably in the literature. Regardless, most studies aim to explain function in their own specific case. Understanding why some species perform duets or antiphonal calling, while others do not, provides insight into comparative social communication and behaviour. In determining the function of vocalizations, context is crucial. Precise vocal exchanges can be competitive, cooperative, or a mix of both. Territorial males engage in many forms of acoustic competition (acoustic dueling in grasshoppers Ligurotettix planum: Greenfield & Minckley 1993; countersinging in birds: e.g. Krebs et al. 1981; chorusing in frogs: e.g. Grafe 2003; roaring in red deer stags Cervus

10

elaphus: Clutton-Brock & Albon 1979). Other kinds of vocal exchanges stem from various levels of cooperation. Avian song duets usually, but not always, occur between mated pairs (Langmore 2002). Some potential functions include mate guarding, joint resource defense, and signaling commitment (reviewed by Hall 2004). Antiphonal exchanges of contact calls occur in a number of social species, such as budgerigars Melopsittacus undulatus (e.g. Brown et al. 1988), Japanese macaques Macaca fuscata (e.g. Sugiura 1993, 1998), cotton-top tamarins Saguinus oedipus (e.g. Weis et al. 2001, Jordan et al. 2004), African elephants Loxodonta africana (Soltis et al. 2005), bottlenose dolphins Tursiops truncatus (Janik & Slater 1998), killer whales Orcinus orca (Miller et al. 2004), and naked mole-rats Heterocephalus glaber (Yosida et al. 2007). Contact calls are often emitted to coordinate pair or group activities or when individuals are visually separated (Cortopassi & Bradbury 2006), and can signal the identity and location of a caller (e.g. Rendall et al. 1996, Wanker et al. 1998, Weis et al. 2001). In crowded situations, contact calls can mediate interactions between two specific individuals. For example, many bat pups that live in colonies produce isolation calls that convey both location and individual identity to mothers (e.g. Gelfand & McCracken 1986, Scherrer & Wilkinson 1993, Bohn et al. 2007). Isolation calls probably originated early in the evolution of bats, and may be common to all bat species. Indeed, bats have evolved an auditory specialization for the relatively low frequencies of isolation calls (Bohn et al. 2006). In a few bat species, adults vocally respond (e.g. Gould 1973, Brown 1976, Fenton 1985). Antiphonal calling has been documented in at least two species, and suggested in others.

11

Matsumara (1981) demonstrated that mother and pup greater horseshoe bats (Rhinolophus ferrumequinum) mutually coordinate their calls in time. Brown et al. (1983) reported “duetting” between mother-pup pairs in lesser fishing bats (Noctilio albiventris). Observations suggest antiphonal calling patterns help mothers identify and localize their pup during reunions at roosts (Gould 1977, Matsumara 1981) and in flight (Brown et al. 1983).

1.3 Antiphonal calling among adult bats: the case of the white-winged vampire The white-winged vampire Diaemus youngi (Chiroptera: Phyllostomidae) is one of three species of Neotropical parasitic vampire bats, which feed exclusively on blood. Very little is known about D. youngi behaviour other than its preference for bird hosts and arboreal feeding (Greenhall & Schutt 1997). The common vampire bat Desmodus rotundus, the sister taxon, is one of the most thoroughly studied bat species, but D. youngi is rare, and their social behaviours are only known from studies on captive bats (Schutt et al. 1999, this study). In 2003, I observed a peculiar calling behaviour between a male and female Diaemus youngi captured in Trinidad. The two bats alternately emitted audible chirps when separated from each other, seeming to precisely coordinate their vocalizations. Similar vocal exchanges were observed in captive bats feeding on chickens (Faure 1995). It appears that D. youngi are unique in the degree to which adult bats seem to duet. I found that white-winged vampire bats share a similar vocal repertoire (Fig. 3) with distantly-related vespertilionid bats (Fig. 4, Pfalzer & Kusch 2003). Signal

12

design roughly corresponds to Morton’s (1977) motivation-structural rules, reported for other mammals and birds. Although D. youngi has a diverse vocal repertoire, only one social call type- a double-syllable (DS) call (Fig. 4b)- is exchanged in a duet-like fashion. These calls consist of two (sometimes one or 3, rarely more) tonal, downward FM syllables with multiple harmonics. As mentioned above, many isolation calls and directive calls in other species have a similar “double-note” structure (Gould 1973, 1977). Following Bohn et al. (submitted), I use the term “double-syllable” rather than “double-note”. Nothing is known about the function of double-syllable calls or antiphonal calling among adult bats. Some researchers (Brown 1976, Fenton 1985, Pfalzer & Kusch 2003) suggest that similar social call structures might be used as contact calls for individual recognition or group cohesion. In the wild-caught white-winged vampires from Trinidad, antiphonal calling between a male and female also suggested the possibility of a mating-related function.

13

14

Figure 3. Spectrograms of 5 basic call types of Diaemus youngi (family Phyllostomidae). FFT displays (amplitude over frequency) are shown for the frequency of most energy. X-axis units are ms; note varying scales. Echolocation pulses (a) are emitted during flight or while scanning surroundings. Intense double-syllable calls (b) are common throughout the night, at least in captive bats. When encountering potential predators or aggressive conspecifics, bats may hiss (c) during an aggressive display or emit a buzz (d) while fleeing. When females are present, males emit complex mono- and multisyllabic calls (e), which may represent a form of song.

15

16

17

Figure 4. Spectrograms of social call types from 6 European bat species (family Vespertilionidae). Taken with permission from Pfalzer & Kusch (2003). Notice the similarities with call types found in D. youngi (Fig. 3): Type A calls and a hiss (Fig. 3c), type B calls and a buzz (Fig. 3d), type C calls and a doublesyllable call (Fig. 3b), and type D calls and complex male songs (Fig. 3e).

18

D. youngi is an intriguing case because data on their social calls can allow us to address several key questions about vocal communication in bats: Are doublesyllable calls used as contact signals by adults (as well as pups)? If so, like pup isolation calls, do they allow individual recognition? Or do adult contact calls only allow identification of groupmates as in P. hastatus? Alternatively, are adult doublesyllable calls used for a different function such as resource defense (e.g. Barlow & Jones 1997a) or mate attraction (e.g. Barlow & Jones 1997b). I investigated the structure and function of antiphonal double-syllable (DS) calls in a captive colony of Diaemus youngi. I considered the predictions of 7 hypotheses regarding their primary function: 1. If white-winged vampire bat antiphonal calling is analogous to avian duetting, then antiphonal calling should occur only between males and females. 2. If DS calls serve a function related to mating, such as attracting members of the opposite sex, then calls should be produced by adults during the mating season. 3. If DS calls are used to advertise male quality in acoustic competition, then this chorusing should occur only between males. 4. If DS calls are contact calls, then they should be easy to localize, produced by isolated individuals and carry information about caller identity. 5. If DS calls are alarm calls, they should be given in the presence of conspecifics, when a predator is detected. 6. If DS calls advertise or defend food, then calling rates should be affected by presence of food.

19

7. Finally, if DS calls are actually signals to heterospecifics such as predators or prey, then calling should not occur among undisturbed bats that are not in the presence of potential prey. I tested these predictions using a series of simple experiments. First, I determined in what contexts isolated bats call. Second, I conducted a playback experiment to measure the effects of call type and social category of caller on response. To determine if calling is exchanged antiphonally, I recorded 4 different individuals simultaneously, using an array of 4 synchronized ultrasonic microphones to characterize temporal emission patterns. I investigated vocal individuality using a statistical analysis of call structure. Finally, I tested for vocal recognition using a habituation-discrimination playback experiment.

20

Chapter 2. Methods

2.1 Animals For all but the final experiment, 16 captive D. youngi served as subjects. Bats were housed within a temperature and humidity-controlled facility in Tijeras, New Mexico (Exhibitor's Permit # 85-C-0021 issued by USDA Animal and Plant Health Inspection Service (APHIS), Animal Care division). Ten of these bats had been wildcaught in Trinidad in 2003, and 6 were born in captivity. Relatedness is unknown. For the vocal discrimination experiment, I used two additional bats. One was born in captivity that year; the other individual was housed at the Burnet Park Zoo in Syracuse, New York. See Appendix. I conducted experiments June – August 2006 and 2007. I marked all individuals with a unique combination of plastic beaded necklaces and wing-punch scars. Walk-in cages housed 4 separate social groups: adult females, old males, young males, and females with pups. Each cage contained one or two wooden roost boxes. Live chickens were herded into cages every night to serve as blood donors (hosts) at a ratio of one chicken per bat, with each chicken bled a maximum of once every 8 days.

2.2. Sound equipment I recorded all sounds with one or more ultrasound condenser microphones (CM 16, Avisoft Bioacoustics, Berlin, Germany) through an Ultrasoundgate 116 (Avisoft) on to a Dell Latitude laptop running the program Recorder USG (Avisoft).

21

Recordings were digitized with 16-bit resolution at a sampling rate of 250 kHz (512 FFT length, 488 kHz resolution). For all playbacks, I used an ultrasonic speaker (Avisoft) connected to an amplifier (Avisoft or custom-made), high-speed DAQcard (6062E, National Instruments, Austin, Texas, USA) and the same laptop running BatSound Pro (Pettersson Electronics).

2.3 Isolation experiments First, I sought to examine the effect of prey and conspecifics calls on vocal behaviour. I isolated an individual bat (n=12) from the rest of the colony under two conditions: with or without a chicken host, alternating the order of presentation. Experiments were located in a different building, and subjects had no visual or auditory contact with colony mates. In a third condition, I spatially isolated an individual bat (n=16) with no chicken present, but allowed the bat to hear the calls of colony mates in an adjacent room. Under all conditions, I used the Avisoft system to remotely monitor the number of DS calls emitted by the isolated bat for two hours. In a second experiment, I sought to examine the effect of DS calls on conspecifics. I remotely observed conspecific response to DS calls from isolated bats. I placed an individual bat (n=6) inside a nylon mesh cage (34 x 34 x 56 cm, Apogee Reptarium), and placed this small cage inside the larger colony cage housing other roostmates (361 x 122 x 226 cm). As a control, I also placed an identical but empty cage inside the colony cage. To record social calls from the isolated bat, a microphone was directed towards the isolated individual and directly away from the

22

colony roost. I used three small wireless LED infrared surveillance cameras to remotely monitor activity.

2.4 Preliminary playback experiments I designed playback experiments to examine three factors that might influence response: call type, caller sex, and caller affiliation (roostmate or not). I also presented playback of social calls from a single male Diaemus youngi recorded in 1993 to investigate response to an unfamiliar conspecific (bat S in Appendix). The chance that subjects are familiar with this bat is extremely slim, because capture dates from different sites in Trinidad differ by 10 years. As a control, I also presented heterospecific echolocation calls from a single big brown bat Eptesicus fuscus, recorded using the same equipment. Playback construction- To record calls from each individual, I isolated a bat in a room (361 x 150 x 302 cm) lined with wedged acoustic foam (noise reduction coefficient @ 4 kHz = 1.08). At the time of recording, a subject was spatially isolated but able to hear and call to other bats in an adjacent room. A mesh cage constrained subjects to within 10-45 cm of the microphone. Due to large differences in amplitude, I could easily discriminate between calls emitted by the subject bat and background calls from other bats. I also used a signal amplitude trigger (set at 12% of measurable intensity) to selectively record calls from the subject. I composed playback stimuli using BatSound Pro. I constructed a playback stimulus for both echolocation pulses and DS calls of every individual. To construct each stimulus, I used the first 18 calls recorded with good signal-to-noise ratio that

23

were between 45% and 99% amplitude. I copied digital call data to a new file and spaced calls with silence. Most calling takes place in spaced apart bouts. Therefore, to provide a more natural temporal pattern, to eliminate temporal variation between stimuli, and to reduce habituation, calls were spaced in a pattern with the following silent intervals (milliseconds): 200, 3000, 3000, 2000, 2000, 1000, 1000, 200, 200, 200, 200, 200, 1000, 1000, 2000, 2000, 3000, 3000, followed by silence until the 30 s mark. This sequence was repeated to produce a 60 s stimulus. I created one 30 s file of silence; which was also repeated to produce a 60 s control stimulus. Playback methodology- All bats received 8 different experimental treatments. On three different days, I presented either two or three playback tests to a subject. Test order was randomized. To avoid habituation, the minimum time between test days for an individual was 4 days. A playback test consisted of a 60 s control period (silent playback) and then 60 s of vocalizations, either echolocation pulses or DS calls. Different echolocation stimuli included calls from: 1. a same-sex roostmate, 2. a same-sex non-roostmate, 3. an opposite-sex non-roostmate, or 4. a heterospecific (E. fuscus). Different DS stimuli included calls from: 5. DS calls from the above same-sex roostmate, 6. the above same-sex non-roostmate, 7. the above opposite-sex non-roostmate, or 8. the one unfamiliar D. youngi male. Pseudoreplication- Because roostmate group sizes were uneven, it was necessary to use one individual’s calls as the same experimental treatment (e.g. different-sex non-roostmate) for two different subjects. Whenever this was the case, I

24

constructed a second unique playback stimulus to represent that individual. I collected playback calls from only one E. fuscus and one unfamiliar D. youngi. Testing procedure- I conducted playback tests between 2030 and 2400 h. I transported subjects individually inside small boxes to a playback room in another building, which was acoustically isolated from all conspecifics. I placed a subject inside a 76 x 76 x 183 cm nylon-mesh testing cage, with one clear acrylic side. The testing cage lay lengthwise, and was surrounded by acoustic foam (described above). I placed subjects into the cage using a small blanket, and testing began once the bat was acclimatized enough to emerge from the blanket. My presence was concealed behind a fabric screen. Each subject received three playback stimuli spaced apart by 9 minutes, so that responses would not linger into the next test. All observable response ceased within 5 minutes. To avoid observer bias during the experiment, I used a procedure that concealed the identity of each playback stimulus during the experiment and response measurement. I gave each playback sequence a number, which corresponded only to its place in the schedule. Thus while recording and quantifying responses I did not know which playback stimulus I was presenting. I recorded ultrasound using the Avisoft system linked to a PC, as described above. I synchronized clocks on both computers at the start of each night. Ultrasound and audio/video recording of responses began simultaneously upon initiation of playback and continued for two min after playback ended. Gains on microphones and the speaker were constant across all trials.

25

Statistics- I used nonparametric statistics in JMP (v.5) to compare the effect of the various stimuli on the vocal response. To control for baseline vocal activity, I calculated the vocal response as: (N during first 3 minutes) – (N during silent control minute x 3) where N = number of social call syllables. Alpha was 0.05.

2.5 Antiphonal calling experiment Experimental setup- In this experiment, my goals were to determine if vocal exchanges are antiphonal, and if so, to determine which- and how many- bats exchange calls. I placed 4 individually-caged bats in the corners of a roughly 4 by 4 m room lined with acoustic foam (described above). I recorded simultaneously from all 4 bats using 4 synchronized microphones (Avisoft), and digitized data with 8-bit resolution. Each microphone was placed at a distance of 5 – 20 cm from a bat, and directed towards that subject and away from all others. Microphones are directional, so calls from a subject showed high signal strength only on its microphone. Furthermore, gains were constantly adjusted to remove any ambiguity of signal origin. I conducted three recording sessions with all-male roostmates, one with allfemale roostmates, one with 1 female and 3 males, and two with 1 male and 3 females. I used 14 bats, omitting two that were caring for pups. I stopped each session when bats seemed habituated to the set-up, as indicated by a sudden decrease in calling activity. Sessions lasted 50-71 minutes. Temporal measurements- I measured onset of call, call duration, and frequency at maximum amplitude using automated procedures in Avisoft

26

SASLabPro. I used frequency at maximum amplitude measurements to discriminate social calls from echolocation pulses. I tested and validated automated measurement parameters of onset times using 268 hand-measured calls. I calculated latency from onset times of social calls. To eliminate redundant calls picked up by more than one microphone, I double-checked call spectrograms where latency was less than 10 ms. This latency period is the maximum amount of time in which sound could arrive at a neighbouring microphone. I also set amplitude detection thresholds to levels that filtered out non-subject calls. Statistics- I considered calls antiphonal if they occurred less than 0.5 seconds after a conspecific call. For each bat in a session, I calculated the number of calls expected to be antiphonal based on random chance alone (E) using the following equation:

E=

0.49N F NC T

where N F is the number of bat calls from the subject, NC is the number of

!

conspecific calls, and T is the total time recorded. For each conspecific call NC , there !

! is an antiphonal response “window” of 0.49 s. The number of subject bat calls ! expected to occur in these windows by chance alone is thus N F *(0.49/T) times the

number of windows, or NC . This approach is mathematically equivalent to another ! analysis of antiphonal calling (Soltis et al. 2005). I used chi-square tests to compare ! actual and expected values for all bats in the antiphonal calling experiments. Alpha

was 0.05. Playback recordings- I applied the above calculations to the data from the previous playback experiment to determine if vocal responses were antiphonal more 27

than expected by chance. I measured latency periods in BatSound Pro, pooling vocal data from all 6 subjects and 18 trials with vocal responses. Bats together versus isolated- If antiphonal calling is related to physical isolation, then the same 4 bats should not call when reunited. I tested this prediction by putting all bats together in a single cage after each session and recording any vocalizations for 10-30 minutes.

2.6 Individual variation in call structure Automated measurements of call structure- I recorded DS calls using methods described above. From each of the 17 bats, I selected all calls, which had the best signal-to-noise ratio, no clipping of signal, unambiguous caller identity, and two syllables (27-64 per individual, 816 total). I marked the beginning and end of each syllable in the program BatSound Pro. Next, a custom-designed Matlab program (Skowronski 2007) used automated frame-based measurements to estimate 58 temporal and spectral parameters for each marked syllable. Variable reduction- The number of variables used in a discriminant function analysis (DFA) must be fewer than the number of levels (calls) in the smallest group (an individual), in this case 27. Furthermore, I found some automated measurements to be unreliable (see below). I therefore reduced the number of variables per call from 116 to 20. First, I removed variables that measured maximum and minimum frequency values, because these measurements might contain inaccurate outliers if my beginning or end marks were off by even one frame. Second, when any two variables were highly correlated (correlation coefficient >0.70), I removed one.

28

Third, I removed variables whose frequency distribution could not be transformed to an approximately normal distribution due to the presence of many outliers. Fourth, I removed variables that measured harmonics above the 4th harmonic, because not all recordings included these data. I assessed normality using Shapiro-Wilk’s tests and from observing frequency histograms. For each variable, I used transformations which best fit data to normal distributions, based on Shapiro-Wilk’s W (Table 1).

Table 1. Variables and transformations used in discriminant function analyses. Units: ms = milliseconds, kHz = kilohertz Variable 1st syllable: Duration (ms) Frequency of the 10th percentile of the fundamental (kHz) Frequency of the 50th percentile of the fundamental (kHz) Frequency of the 90th percentile of the fundamental (kHz) !(kHz) Frequency of most energy (FME) of the fundamental ! st FME of the 1 harmonic (kHz) FME of the 3rd harmonic (kHz) ! Time of FME of the fundamental relative to start (ms) ! Slope of the 50th percentile of the fundamental (kHz/ms) 2nd syllable: Duration (ms) Frequency of the 10th percentile of the fundamental ! (kHz) Frequency of the 50th percentile of the fundamental (kHz) Frequency of the 90th percentile of the fundamental (kHz) FME of the fundamental (kHz) ! FME of the 1st harmonic (kHz) ! FME of the 2nd harmonic (kHz) rd ! FME of the 3 harmonic (kHz) ! Time of FME of the fundamental relative to start (ms) ! Slope of the 50th percentile of the fundamental (kHz/ms) ! Interval: ms between end of 1st syllable and start of 2nd

Transformation

x + 0.5 2

x None x + 0.5 x + 0.5 None None x None

None log(x + 1) x2 x2 x2 x2 x2 None log(x + 1) None log(x + 15)

!

Statistics- I initially used a DFA in SPSS (v.11) to calculate correct ! classification rates for individual identity of 17 bats. I entered 20 variables together 29

(Table 1). I used a random 75% of calls to derive discriminant functions; the remaining 25% were used as test calls. I then used permuted DFA (pDFA, see Mundry & Sommer 2007) in R (v.2.6) to test the null hypothesis that there is no difference in DFA performance when classifying individual identity in the original dataset versus datasets in which individual identity is randomly assigned to calls. An R script (written by Roger Mundry) compared the number of correctly classified calls between the original dataset and the results of 1000 DFAs where subject (the test factor) was randomized among calls recorded in the same year (the control factor). To control for the effect of sex, I performed a pDFA on male and female calls separately. I omitted one bat, born that year, from the analysis because it was recorded in only one year and sex was unknown. Each permutation in a pDFA requires two randomizations (Mundry & Sommer 2007). First, the pDFA randomly selects training calls from each individual bat to derive discriminant functions. The pDFA chooses the number of training calls such that the sample of calls for each combination of individual identity and year of recording are balanced. The remaining (testing) calls are then used for validating the discriminant functions (i.e. cross-classification). Because numbers of calls per bat are not equal, I repeated DFAs with 100 different random selections to determine the average number of correct test assignments for the original dataset. In the second randomization, the pDFA randomly assigns new “individual identity” values to calls to make 1000 randomized data sets, and performs a DFA for each of these to determine a distribution of correct test call assignments. The average number of

30

correct test assignments for the original data set is then compared to this distribution to calculate a one-tailed p-value. Alpha was 0.05.

2.7 Vocal discrimination experiment General experimental design- I used a habituation-discrimination approach because this test is particularly powerful for showing individual recognition (e.g. Rendall et al. 1996, Boughman and Wilkinson 1998, Weiss et al. 2001) and does not require training. In this design, a subject is repeatedly presented with a stimulus (A); after habituation takes place (lowered or no response), a new stimulus (B) is presented. A rebound in response at this point is evidence of discrimination between A and B. Following Rendall et al. (1996), I represented each individual with playback of multiple call exemplars. This design tests for discrimination of individuals, not simply stimuli. Playback sequence construction- I constructed 24 playback stimuli from recordings made using methods described above (2.3). Responses to 8 playback sequences were discarded since I aborted those trials (see below); I only analyzed responses to 16 playback sequences (Table 2). Although the same bats were used several times as habituation stimuli or test stimuli due to call recording availability, every sequence used a unique set of call exemplars. I only used calls with high (>90%) signal-to-noise ratio with signal amplitudes between 45% and 99%. Mean amplitude of playback calls from different individuals did not differ significantly. Median latency between calls of a lone bat is 998 ms (n=500), so I spaced playback calls between randomly generated latencies of 900 to 1100 ms.

31

Table 2. Playback sequences included in statistical analyses. Letters indicate different bats. * males, ? unknown sex

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Source of habituation calls P∗ A I∗ H∗ H∗ G∗ F E E B B B B A C A

Source of test calls F∗ D E S∗ B H∗ C N G∗ S∗ R? L K R? R? L

Length of habituation (seconds) 30 30 90 60 30 90 30 60 30 90 30 60 30 30 60 60

Each playback sequence (Fig. 5) included a 5 s silent period, followed by a habituation period (either 30, 60, or 90 s, Table 2), a test period (10 s), and a rehabituation period (10 s). The habituation and re-habituation periods consisted of different call exemplars from bat A, while the test period presented different call exemplars from bat B. To calibrate intensity of playback to natural levels, I adjusted speaker gain such that calls monitored simultaneously from the speaker, and a live bat facing the speaker, registered with equal amplitude to an equidistant microphone.

32

Figure 5. Typical playback sequence. Y-axis shows intensity; X-axis shows time. The test period is highlighted with a dashed-line box.

Playback methodology- I conducted playback tests between 2030 and 2400 h in complete darkness. I transported subjects inside boxes to an acoustically isolated testing room. Temperature and humidity were maintained at levels similar to housing area. Observations suggest that bats respond to conspecifics most often when isolated after vocal or physical contact is interrupted. Therefore, I initially allowed the subject bat and the bat corresponding to the habituation stimulus to vocally interact alone in complete darkness. I placed the subject bat inside a 13.9 x 7.6 x 7.6 cm testing cage, in which one side was clear acrylic; I placed the non-subject bat in a different cage nearby. After 2-5 min, I removed the non-subject bat. After another acclimation period of 0.5- 2 min, I broadcast a playback sequence. I simultaneously began recording response using a Sony Nightshot DV Camcorder (30 frames/s), equipped with an infrared light, a Sennheiser K6 microphone, and linked directly to a 33

Macintosh G4 PowerBook running the application iMovie HD. I also simultaneously recorded ultrasound using the Avisoft system (described above) linked to a PC laptop. I synchronized clocks on both computers and the DV camcorder at the start of every night. The speaker was located 79 cm from the subject. The ultrasound microphone faced towards the subject and directly away from the speaker. The gain on both microphones and the speaker was constant across all trials. I monitored and recorded response and started playback from an adjacent room. I defined a response as any of the following: 1) vocal response- production of social call, 2) orientation- turning of head and body towards speaker, 3) startle response- change from eyes closed and ears still, to eyes open and ears moving. If bats did not respond to any habituation period or the test period, I played a “post-test stimulus” (following Weiss et al. 2001). The post-test stimulus was a distress call, an aggressive hissing call, a rapid bout of calling, and a series of complex vocalizations, all from different bats. If a subject failed to respond to this series of sounds, then I aborted that trial and discarded those data. I also aborted trials where subjects did not habituate during the habituation period (i.e. showed no visible decrease in response) or when the experiment was interrupted, for instance by an outside noise. Out of 135 trials attempted, I aborted 59. I used data from 8 of 16 individuals in the final analyses. Response measurement- I measured two responses independently: calling rate and physical response duration. To measure calling, I used the program BatSound Pro to count the number of social call syllables produced by the subject during each 10 s period of the playback sequence starting at the onset of the playback. Subject calls

34

could be unambiguously discriminated from playback calls. To measure physical response duration, I analysed frame-by-frame digital video in the program iMovie HD. For each 10 s period, I marked the starting frame of the response until the response or the period ended, yielding a measurement from 0-10 s. I measured all responses blind to identity of the trial. Statistical analysis- I used nonparametric statistics to compare the mean response during the 10 s test period to the average of the mean responses during 10 s periods immediately before and after the test period, i.e. (mean response 10 s before + mean response 10 s after)/ 2). Averaging the mean responses during these two periods is a conservative approach, because some of the response to the 10 s test period may linger into the following 10 s period. Hence, the difference in response between the test and two habituation periods will be less significant than if only the last habituation trial and test trial are compared. To avoid pseudoreplication, I averaged the responses across all trials for each subject (with same sequence), then averaged the responses across all subjects within each sequence. This approach loses power but allows for the use of conservative and straightforward nonparametric tests. I conducted analyses using JMP (v. 5.1) and SPSS 11.

35

Chapter 3. Results

3.1 Isolation experiments Diaemus youngi emitted DS calls when spatially isolated and able to hear conspecifics. When completely isolated with a chicken, no bats called (n=12). When completely isolated with no chicken present, one bat out of 12 emitted a total of 9 calls. When bats were spatially isolated and allowed to hear conspecifics, all 16 of 16 called, a total of 507 calls. Bats are attracted to DS calls from isolated conspecifics. In the second experiment, 5 out of 6 bats began to emit DS calls after lights were turned off. In all 5 of these cases, I recorded rapid vocal exchanges. In all 6 cases, roostmates were attracted to the cage with a bat inside, but ignored the empty cage.

3.2 Preliminary playback experiments Bats’ responses were low overall and decreased rapidly over time, suggesting that they became habituated to the playback tests. Although all subjects moved towards the speaker during at least one of the tests, only 6 of the 16 bats responding with social calls to any playback test. Thus, only 18 of the 128 tests elicited a vocal response. Regardless, several results are noteworthy. Response to white-winged vampire bat social calls was significantly greater than to white-winged vampire bat echolocation pulses (Wilcoxon Z= 2.89, p=0.0038, n=96). Furthermore, responses were not limited to D. youngi callers of any one category; subjects replied to the opposite sex, same-sex, roostmates, non-roostmates and the unfamiliar individual. I

36

thus found no effect of caller sex (Wilcoxon Z= 0.372, p=0.7, n=96) or roosting affiliation (Wilcoxon Z= 0.176, p=0.86, n=96) on vocal response to playback. Of the 6 bats that replied to playback, three responded to the unfamiliar conspecific. No bats called during the control periods. There was almost no response to heterospecific calls; one bat responded to the first echolocation pulse of an E. fuscus, but ignored the rest of the playback.

3.3 Antiphonal calling experiments Calling was not random in time (Fig. 6). The most common latency period was ca. 300 – 350 milliseconds. Ten out of 14 bats called during these experiments, but I excluded two females from analyses due to small sample size (n=2 calls each). For all but one of the remaining bats, calls occurred less than 0.5 seconds after a conspecific call significantly more than expected by chance (Fig. 7).

Figure 6. Frequency distribution of conspecific responses within 5 seconds of a social call. Black bars represent calls considered antiphonal (latency < 0.5 seconds). 37

Antiphonal calling bouts occurred in every session and decreased in frequency over time. Most antiphonal calling (67%) occurred as an initial DS call followed by a single response. The longest antiphonal bout included 5 alternating DS calls among three bats. If all calling consisted of calls with a single antiphonal response, then 50% of calls are expected to be antiphonal responses and the other 50% of calls would be initiating calls. In contrast, approximately 18% of calls (n=1858) were antiphonal responses. Thus many calls received no response. During the playback experiments, 38.5% of responses (n= 703) were antiphonal, i.e. within 0.5 seconds of a playback call. Furthermore, bats generally called when isolated, but not when together. DS call bouts by isolated bats began immediately after the lights were turned off, but I recorded no DS calls from the same bats placed together.

38

39

Figure 7. Percentage of social calls that occurred as antiphonal responses (within 0.5 seconds of a conspecific call) in 8 bats. White bars show percentage expected by chance alone. Black bars show actual percentage observed. Sample sizes (n) are total numbers of calls emitted. Chi-square values (based on raw data) for each bat: D = 568.4, G = 8.3, H = 44.0, I = 1.6, J = 116.7, M = 90.3, O = 636.3, P = 422.1. Df= 1. * values significantly greater than chance (p<0.005).

40

3.4 Individual variation in call structure I found individual differences in structure of DS calls. Correct classification rates for the preliminary DFA in SPSS for test calls were 78.3% for females, 73.8% for males, and 67.2% overall. Number of test calls correctly assigned to individual by pDFA with the original data set was significantly greater than with randomized data sets (females: p=0.005, n=417; males: p=0.001, n=399). For the single bat excluded from the pDFA, 100% of test calls were correctly classified by the preliminary DFA.

3.5 Vocal discrimination experiments Vocal response- Mean calling rate of subjects during the 10 s test period was significantly greater than the average of the mean calling rates during the habituation period (10 s before) and re-habituation period (10 s after- Wilcoxon Sign Rank test: W= 30, p= 0.005, n= 16, Fig. 8a). Mean calling rate during the test period did not differ significantly from the first 10s of habituation (Wilcoxon Sign Rank test: W= 4.5, p= 0.772, n=16, Fig. 8a). Physical response- Mean duration of physical response during the 10 s test period was significantly greater than the average of the mean durations during the habituation period and re-habituation period (Wilcoxon Sign Rank test= 49, p= 0.003, n=16, Fig. 8b). Mean duration during the test period was also greater than the first 10s of habituation (Wilcoxon Sign Rank test= 41, p= 0.007, n= 16, Fig. 8b).

41

42

Figure 8. Mean vocal (a) and physical (b) responses to periods of a habituation-discrimination playback sequence. Error bars represent standard error of the mean. Response during the “middle” period was calculated as mean response per 10 s period between the first and last habituation periods. Last bar shows the average of the responses during the periods before and after the test period. See text for test statistics. a) * p=0.005; b) diamonds: p=0.007; triangles: p=0.003.

43

Results for analysed trials are summarized in Table 3. All other trials were aborted; see Methods for criteria. Successful trials demonstrate that bats could vocally discriminate conspecifics regardless of sex or roosting affiliation (cage).

Table 3. Results of playback experiments. Letters represent different individual bats. * indicates that the sex of bat changes between habituation and test stimuli sources. Ψ A rebound during test response means that the response during the last 10s habituation period was zero and the test period response was greater than zero. Subject

Playback

Habituation source to test source K

F

D

A to R G to H H to B* E to G* B to R I to E* A to L B to L E to N F to C* H to S E to G* A to R B to K O to D* H to B* I to E* H to S B to L P to F I to E*

Affiliation with subject D= different cage S= same cage U=unfamiliar D to S D to D D to D D to D D to S D to D D to S D to S D to D D to D D to U S to D S to D S to D D to S D to S D to S D to U S to D D to D D to S

No. of trials 3 3 1 1 1 3 2 1 2 1 1 2 3 3 1 2 1 1 1 2 1

44

Evidence for discrimination (Rebound during test period) Ψ vocal physical response response X X

X

X X X X X

X

Responses only during habituation

X X X

X

X X

X

X X X

X X

X X X

J

H

P

O I

B to K F to C* A to R E to G* B to S* G to H B to R G to H I to E* H to B* E to G* P to F F to C* I to E* E to G* A to R O to D* H to S A to R C to R G to H E to G* H to B* E to G* B to R E to G* G to H H to B*

S to D D to S S to D S to D D to U D to D S to D S to S S to D S to D D to S D to D D to D S to D D to S D to D D to D S to U D to D D to D D to D D to D D to D D to D D to D D to S S to S S to D

2 1 4 1 1 1 1 2 2 1 1 1 1 3 1 2 1 1 1 1 1 1 1 1 1 1 2 1

45

X

X X

X X X X

X X X X X X

X X X X

X

X X X

Chapter 4. Discussion

4.1 Temporal pattern of antiphonal calling Double-syllable (DS) calls sometimes receive an antiphonal response with a typical latency of about 1/3rd of a second. At other times, there is no response from conspecifics. This basic call/answer pattern is found in primate antiphonal contact calls, albeit with longer mean latency periods (e.g. about 3.7 s in cottontop tamarins Saguinus oedipus: Miller et al. 2005). In the antiphonal calling experiments, most responses ceased after a series of initial bouts lasting roughly 20 minutes, but a few individuals continued to call repetitively with no response from conspecifics.

4.2 Possible functions of antiphonal responses Vocalizations carry costs, such as expending energy and advertising location to predators. Therefore, in order to be selected for, they must provide significant direct or indirect benefits to callers. Responding antiphonally to contact calls might be advantageous to listeners for several reasons. Kin or reciprocal altruists might be seeking to: minimize the amount of time the caller must wait for a response (Hall 2004), signal reception or identity (Sugiura 1993), and/or encode information in the latency period. Another possibility is that listeners use antiphonal responses to increase or update available information. In theory, latencies between behaviours will normally be shorter in paired or grouped animals than in lone animals (Dostalkova & Spinka 2007). Accordingly, individual calling rates should be higher during antiphonal

46

calling than for the same calls performed by lone individuals. Therefore, listeners seeking more signals can rapidly answer calls to increase calling rate, thereby increasing available cues regarding location and identity. For example, when searching for pups, mother bats can increase isolation call rate by responding rapidly (Gould 1977, Matsumara 1981), which would make recognition and localization faster. My results support this explanation for adult white-winged vampires as well, because individual calling rates are lower for bats calling alone. However, this cohesion scenario does not explain why adult bats emit DS calls in the first place, or why adults need to find each other.

4.3 Function of double-syllable calls Several results suggest that DS calls are long-distance contact calls (Table 4). DS calls are more intense and lower frequency than echolocation calls, so they travel farther distances (Fig. 9). DS calling was not correlated with disturbance of any kind or with the presence or absence of hosts. Likewise, DS calling was not elicited by disturbance. Rather, calling stopped when a person appeared. DS calling occurs year-round (personal observations, Daniel Abram, pers. comm.). Calling is not constrained by sex or age class; I recorded antiphonal responses from bats as young as 4 months. Four observations suggest that antiphonal calling behaviour between unrelated adults in D. youngi is similar to- and perhaps evolved from- the antiphonal contact calling that sometimes occurs between bat mother and pups. First, a doublesyllable structure is common in bat pup isolation calls (Gould 1973, 1977). Second,

47

calls are most reliably recorded from bats that can hear conspecifics, but are isolated. Third, conspecifics are attracted to the calls. And lastly, calls are individually distinct.

Table 4. Hypotheses for call function. The only hypothesis fully supported by data is maintaining contact (*). Category: primary function

Antiphonal calling? Yes

Responses from all sexes, ages? Maybe

Call when isolated? Maybe

Mating: attract mates, advertise quality to females

Yes

Maybe

Maybe

Contact calling: localize individuals*

Yes

Yes

Yes

Counter-singing: advertise male quality to other males

Alarm calling: warn conspecifics/kin Pair duetting: various functions

Yes Yes

Maybe

Food calling: advertise food Aggression: defend a resource

Yes Yes

Yes

Signals to chickens: unknown function

48

49

Figure 9. Oscillogram (intensity over time) and spectrogram (frequency over time) showing large difference in signal strength between double-syllable social calls and echolocation pulses of a Diaemus youngi. Calls were recorded at the same distance with an Avisoft CM16 ultrasound microphone (frequency response ± 3 dB at 10 to 100 kHz).

50

On the other hand, DS calls are not only emitted during isolation. Captive white-winged vampires also emit DS calls while emerging from roost boxes, while feeding on chickens, and in some social interactions. For example, I observed one bat calling while atop another individual, so DS calls may also be related to dominance behaviours. These observations support the notion that adult contact calls are used in a variety of contexts.

4.4 Evidence for individual recognition Evidence from call structure- Using a permuted DFA, I have shown that individual identity can be assigned with greater than chance probability using call structure alone. This result alone does not establish vocal recognition, but playback results provide convincing evidence of individual discrimination. Why response to test stimuli was greater than habituation stimuli- In a typical habituation-discrimination test, the initial response to the habituation stimulus should be greater than, or equal to, the test stimulus (e.g. Rendall et al. 1996). However, for the data I present, mean response was greater during the test period than during the initial habituation period (Fig. 8). There are two explanations for this result. First, the initial habituation period was preceded by some prior habituation, when subjects were exposed to the same live bat from which they would later hear playback calls (see Methods). Therefore, response to the “first” habituation period is not actually the response to the first stimulus, i.e. calls from the live bat. The “first” response corresponds instead to the first playback calls. Hence, from the subject’s point of view, these “first” playback calls may have represented a continuation of

51

calling. Indeed in 4 trials, subjects seemed to be already habituated prior to playback presentation; they only responded during the test period, when the caller’s identity changed. A second reason that the test period elicited more responses is that trials with strong responses during the habituation period were more likely to be aborted. When bats responded strongly during the habituation period, they also often failed to habituate before the test period began, and I consequently aborted the trial (see Methods). Evidence from habituation-discrimination experiments- Playback results demonstrate that white-winged vampire bats can discriminate between calls from different individuals. Because relatedness is unknown, it is theoretically possible that every observed case of discrimination was actually a case of kin discrimination, and not individual recognition. However, my data are unlikely to result from kin discrimination alone for two reasons. First, several subjects could discriminate between the same pair of bats. For example, 5 subjects discriminated playback of bat R from bat A (Table 2), and all 5 responded more to playback of bat R. If bats can only discriminate kin based on call, then all 5 of these subjects must be related to one bat, but not the other. It is far more likely that subjects discriminated calls from bats based on identity. In particular, Bat R may have elicited a strong response, because it was born that year. Second, bat S was almost certainly unrelated and unfamiliar to all subjects (see Methods), yet 4 bats could discriminate its calls from others (Table 2). Diaemus youngi are “whispering bats” (family Phyllostomidae) with echolocation pulses of very low intensity (Fig. 9, Griffin 1958). They may thus

52

require low frequency, high intensity social calls to signal effectively at longer distance. In other families of echolocating bats, biosonar pulses are very intense and may also be used for long-distance vocal discrimination by conspecifics. For example, Kazial & Masters (2003) demonstrated that female big brown bats (Eptesicus fuscus) recognize sex based on echolocation calls alone. Many studies have searched for structural variation in bat calls related to various attributes such as age or identity (e.g. Masters et al. 1995, Siemers & Kerth 2006), but more playback studies are needed to determine if bats actually perceive or respond to these differences.

4.5 Possible adaptive benefits of long-distance recognition of individuals The Old World primates (Catarrhini) are known for their complex societies in which individual relationships are highly influenced by relatedness, dominance rank, and previous interactions. In these taxa, social relationships are a major component of fitness (reviewed by Silk 2007). The importance of individual identity is obvious and has selected for a keen ability to discriminate individuals based on acoustic cues, even in vocalizations such as alarm calls (Sproul et al. 2006). Contact calls appear to function primarily as a mechanism to announce identity and location whenever it provides a benefit, such as staying with a group or protecting kin (e.g. Rendall et al. 2000). Like many social primates, common vampire bats (Desmodus rotundus) live in societies where long-term individual social relationships are important components of reproductive success. D. rotundus are long-lived (15 years in the wild: Tschapka

53

& Wilkinson 1999), and their social structure conforms to a fission-fusion model of sociality (Wilkinson 2003). In hollow trees, groups of 8-12 (up to 20 in caves) adult females and their pups also typically include a dominant male and subordinate males (Wilkinson 1985, 1988). Although intra-group relatedness is low (Wilkinson 1985b), individual affiliations are important and maintained partly through allogrooming (Wilkinson 1986). Indeed, such relationships are a prerequisite for the evolutionary stability of the reciprocal altruism that takes place (regurgitated food sharing between kin and unrelated adults: Wilkinson 1984, DeNault & McFarlane 1995). Desmodus rotundus is considered the closest living relative to Diaemus youngi based on molecular and morphological data (Greenhall & Schutt 1996), and several observations suggest individual identity plays an important social role in Diaemus youngi as well. Captive D. youngi perform reciprocal food sharing (Fig. 10) and allogrooming. I have also observed allonursing and cooperative care of young. Certain pairs of individuals were consistently aggressive or friendly to one another. Although roosting location varied, an individuals’ sleeping position in relation to other bats was fairly stable.

54

a

b

Figure 10. a) Initiation of reciprocal food sharing in Diaemus youngi. A female (left) is trying to elicit regurgitation of blood from a male (right). b) Food sharing at the wound by two females. Both photos depict bats used in this study and were taken by the author.

Interestingly, D. rotundus do not seem to use similar contact calls and are not nearly as vocal (Schmidt 1972, Sailler & Schmidt 1978; G. Carter, unpublished observations). This distinction may reflect important differences in social behaviour between the two species. Specifically, I predict that D. youngi social interactions probably differ from D. rotundus outside the roost. In individualized societies, simply advertising one’s identity may be a very effective way to influence conspecifics in ways that benefit the signaler (Rendall & Owren 2001). Antiphonal calling and long distance individual recognition thus have obvious implications for possibilities of social interactions outside the roost. In Phyllostomus hastatus for instance, female roosting groups use group-specific contact calls, termed screech calls, to facilitate cohesion and recruit groupmates to foraging

55

sites (Boughman & Wilkinson 1998, Wilkinson & Boughman 1998). In this case, individual bats may benefit if cooperative foraging allows for the monopolization of food resources and the exclusion of individuals from foreign groups (Wilkinson & Boughman 1998). A similar situation may occur in Diaemus youngi. White-winged vampires likely use antiphonal contact calls to mediate more complex social interactions in the wild. D. youngi might travel in groups or pairs, and/or recruit certain individuals to host or roost locations. Wild-caught D. youngi are often found in nets as pairs (Hawsraj Abraham, Trinidad vampire control, pers. comm.). Group defense of feeding areas has also been suggested for Desmodus rotundus (Wilkinson 1987). Captive D. youngi feed simultaneously from the same wound. Based on these observations, possible adaptive benefits of pairing or grouping while stalking hosts might include sharing resources with kin, cooperatively defending feeding areas, and reciprocal sharing of wounds. However, all hypotheses amount to mere speculation at this point, since there are no reports on social behaviour of wild D. youngi. Further study is needed to determine how white-winged vampires use these contact calls in the wild.

56

4.6 Conclusions 1. Double-syllable calls, which mediate contact between bat pups and mothers in other bat species, are used as contact calls between adults in Diaemus youngi. 2. D. youngi often “answer” conspecific calls with a latency of about 330 ms. 3. Antiphonal calling allow D. youngi to maintain contact with specific individuals at long-distance. 4. D. youngi can discriminate individuals based on double-syllable calls alone. 5. Individual social relationships are important for both D. youngi and Desmodus rotundus within the roost. Outside the roost, I predict D. youngi performs some social behaviours requiring individual recognition, which D. rotundus do not.

57

References Andrews, M. M. & Andrews, P. T. 2003. Ultrasound social calls made by greater horseshoe bats (Rhinolophus ferrumequinum) in a nursery roost. Acta Chiropterologica, 5, 221-234. Balcombe, J. P. 1990. Vocal recognition of pups by mother Mexican free-tailed bats, Tadarida brasiliensis mexicana. Animal Behaviour, 39, 960-966. Balcombe, J. P. & McCracken, G. F. 1992. Vocal recognition in Mexican free-tailed bats: do pups recognize mothers? Animal Behaviour, 43, 79-88. Barclay, R. M. R., Fenton, B. & Thomas, D. W. 1979. Social behavior of Myotis lucifugus. II Vocal Communication. Behavioral Ecology and Sociobiology, 6, 137-146. Barlow, K. E. & Jones, G. 1997a. Function of pipistrelle social calls: field data and a playback experiment. Animal Behaviour, 39, 960-966. Barlow, K. E. & Jones, G. 1997b. Differences in songflight calls and social calls between two phonic types of vespertilionid bat Pipistrellus pipistrellus. Journal of Zoology (London), 241, 315-324. Behr, O. & Helversen, O. v. 2004. Bat serenades - complex courtship songs of the sac-winged bat (Saccopteryx bilineata). Behavioral Ecology and Sociobiology, 56, 106-115. Behr, O., Helversen O. v., Heckel G., Nagy M., Voigt, C. C. & Mayer, F. 2006. Territorial songs indicate male quality in the sac-winged bat Saccopteryx bilineata (Chiroptera, Emballonuridae). Behavioral Ecology, 17, 810-817. Bohn, K. M., Boughman, J. W., Wilkinson, G. S. & Moss, C. F. 2004. Auditory sensitivity and frequency selectivity in greater spear-nosed bats suggest specializations for acoustic communication. Journal of Comparative Physiolology A, 190, 185-192. Bohn, K. M., Moss, C. F. & Wilkinson, G. S. 2006. Correlated evolution between hearing sensitivity and social calls in bats. Biology Letters, 2, 561-564. Bohn, K. M., Wilkinson, G. S. & Moss, C. F. 2007. Discrimination of infant isolation calls by female greater spear-nosed bats, Phyllostomus hastatus. Animal Behaviour, 73, 423-432. Bohn, K. M., Schmidt-French, B., Ma, S. T. & Pollack, G. D. Submitted. Syllable acoustics, temporal patterns, and call composition vary with behavioural context in Mexican free-tailed bats.

58

Boughman, J. W. 1997. Greater spear-nosed bats give group-distinctive calls. Behavioral Ecology and Sociobiology, 40, 61-70. Boughman, J. W. 1998. Vocal learning by greater spear-nosed bats. Proceedings of the Royal Society Biological Sciences Series B, 265, 227-233. Boughman, J. W. & Wilkinson, G. S. 1998. Greater spear-nosed bats discriminate group mates by vocalizations. Animal Behaviour, 55, 1717-1732. Bradbury, J. W. & Vehrencamp, S. L. 1998. Principles of Animal Communication. Sunderland, MA: Sinauer Associates, Inc. Brown, P. 1976. Vocal communication in the pallid bat, Antrozous pallidus. Zeitschrift fur Tierpsychologie, 41, 34-54. Brown, P. E., Brown, T. W. & Grinnell, A. D. 1983. Echolocation, development, and vocal communication in the lesser bulldog bat, Noctilio albiventris. Behavioral Ecology and Sociobiology, 13, 287-298.. Brown, S. D., Dooling, R. J. & O'Grady, K. 1988. Perceptual organization of acoustic stimuli by budgerigars (Melopsittacus undulatus): III. Contact Calls. Journal of Comparative Psychology, 102, 236-247. Christesen, L.-S. & Nelson, J. 2000. Vocal communication in the Grey-headed Flying-fox Pteropus poliocephalus (Chiroptera: Pteropodidae). Australian Zoologist, 31, 447-457. Clutton-Brock, T. H. & Albon, S. D. 1979. The roaring of red deer and the evolution of honest advertising. Behaviour, 69, 145–170. Cortopassi, K. A. & Bradbury, J. W. 2006. Contact call diversity in wild orangefronted parakeet pairs, Aratinga canicularis. Animal Behaviour, 71, 11411154. Davidson, S. M. & Wilkinson, G. S. 2004. Function of male song in the greater white-lined bat, Saccopteryx bilineata. Animal Behaviour, 67, 883-891. Dawkins, R. & Krebs, J. R. 1978. Animal signals: information or manipulation? In: Behavioural Ecology: An Evolutionary Approach (Ed. by Krebs, J. R. & Davies, N. B.), pp. 282-309. Sunderland, MA: Sinauer Associates, Inc. de Fanis, E. & Jones, G. 1995. Post-natal growth, mother-young interactions and development of vocalizations in the vespertilionid bat Plecotus auritus. Journal of Zoology (London), 235, 85-97.

59

DeNault, L. A. & McFarlane D. A. 1995. Reciprocal altruism between male vampire bats Desmodus rotundus. Animal Behaviour 49, 855-856. Dostalkova, I & Spinka, M. 2007 (published online). Synchronization of behaviour in pairs: the role of communication and consequences in timing. Animal Behaviour, in press. Dunbar, R. I. M. 2003. The social brain: mind, language, and society in evolutionary perspective. Annual Review of Anthropology, 32, 163-181. Esser, K. H. 1994. Audio-vocal learning in a non-human mammal: the lesser spearnosed bat Phyllostomus discolor. Neuroreport, 5, 1718-1720. Esser, K. H. & Schmidt, U. 1989. Mother-infant communication in the lesser spearnosed bat Phyllostomus discolor (Chiroptera, Phyllostomidae): evidence for acoustic learning. Ethology, 82, 156-168. Faure, P. A., Brockmann, K., Schutt, W. A. J. & Muradali, F. 1995. Abstracts: Behavior and vocalizations of the white-winged vampire bat during blood feeding. Bat Research News, 36. Fenton, M. B. 1977. Variation in the social calls of little brown bats (Myotis lucifugus). Canadian Journal of Zoology, 55, 1151-1157. Fenton, M. B. 1985. Communication in the Chiroptera. Bloomington: Indiana University Press. Fenzl, T. & Schuller, G. 2007. Dissimilarities in the vocal control over communication and echolocation calls in bats. Behavioural Brain Research, 182, 173-179. Freeberg, T. M. 2006. Social complexity can drive vocal complexity: group size influences vocal information in Carolina chickadees. Psychological Science, 17, 557-561. Gelfand, D. L. & McCracken, G. F. 1986. Individual variation in the isolation calls of Mexican free-tailed bat pups (Tadarida brasiliensis mexicana). Animal Behaviour, 34, 1078-1086. Gillam, E. 2007. Eavesdropping by bats on the feeding buzzes of conspecifics. Canadian Journal of Zoology, 85, 795-801. Gould, E. 1977. Echolocation and communication. In: Biology of Bats of the New World Family Phyllostomatidae (Ed. by Baker, R. J., Jones, J. K. J. & Carter, D. C.), pp. 247-279. Lubbock, TX: Texas Tech University Press.

60

Gould, E., Woolf, N. K. & Turner, C. 1973. Double note communication calls in bats: occurrence in 3 families. Journal of Mammalogy, 54, 998-1001. Grafe, T. U. 2003. Acoustic interactions among chorusing male anurans. In: Animal Communication Networks (Ed. By McGregor, P. K.). Cambridge, UK: Cambride University Press. Greenhall, A. M. & Schutt, W. A. 1996. Diaemus youngi. Mammalian Species, 533, 1-7. Greenfield, M. D. & Minckley, R. L. 1993. Acoustic dueling in tarbush grasshoppers: settlement of territorial contests via alternation of reliable signals. Ethology, 95, 309-326. Griffin, D. R. 1953. Acoustic Orientation in the Oil Bird, Steatornis. Proceedings of the National Academy of Science USA, 39, 884-893. Griffin, D.R. 1958. Listening in the Dark. New Haven, CA: Yale University Press. Hall, M. L. 2004. A review of hypotheses for the functions of avian duetting. Behavioral Ecology and Sociobiology, 55, 415-430. Holderied, M. W., Korine, C., Fenton, M. B., Parsons, P. E., Robson, S. & Jones, G. 2005. Echolocation call intensity in the aerial hawking bat Eptesicus bottae (Vespertilionidae) studied using stereo videogrammetry. Journal of Experimental Biology, 208, 1321-1327. Holderied, M. W. & von Helversen, O. 2003. Echolocation range and wingbeat period match in aerial-hawking bats. Proceedings of the Royal Society of London B Biological Sciences, 270, 2293-2299. Janik, V. M. & Slater, P. J. B. 1998. Context-specific use suggests that bottlenose dolphin signature whistles are cohesion calls. Animal Behaviour, 56, 829-838. Jordan, K., Weiss, D., Hauser, M. & McMurray, B. 2004. Antiphonal responses to loud contact calls produced by Saguinus Oedipus. International Journal of Primatology, 25, 465-475. Kanwal, J. S., Matsumara, S., Ohlemiller, K. K. & Suga, N. 1994. Analysis of acoustic elements and syntax in communication sounds emitted by moustached bats. Journal of the Acoustical Society of America, 96, 12291254. Kanwal, J. S. & Rauschecker, J. P. 2007. Auditory cortex of bats and primates: managing species-specific calls for social communication. Frontiers in Bioscience, 12, 4621-4640.

61

Kazial, K. A. & Masters, W. M. 2004. Female big brown bats, Eptesicus fuscus, recognize sex from a caller's echolocation signals. Animal Behaviour, 67, 855863. Kingston, T., Jones, G., Akbar, Z. & Kunz, T. H. 2000. Social calls in clear-winged woolly bats Kerivoula pellucida from Malaysia. Bioacoustics, 11, 1-16. Knornschild, M., Behr, O. & von Helversen, O. 2006. Babbling behavior in the sacwinged bat (Saccopteryx bilineata). Naturwissenschaften, 93, 451-454. Krebs, J. R., Ashcroft, R. & van Orsdol, K. 1981. Song matching in the great tit (Parus major L.). Animal Behaviour, 29, 918-923. Langmore, N. E. 2002. Vocal duetting: definitions, discoveries and directions. Trends in Ecology and Evolution, 17, 451-452. Leippert, D., Goymann, W., Hofer, H., Marimuthu, G. & Balasingh, J. 2000. Roostmate communication in adult Indian false vampire bats (Megaderma lyra): an indication of individuality in temporal and spectral pattern. Animal Cognition, 3, 99-106. Lundberg, K. & Gerell, R. 1986. Territorial advertisement and mate attraction in the bat Pipistrellus pipistrellus. Ethology, 71, 115-124. Ma, J., Kobayasi, K., Zhang, S. & Metzner, W. 2006. Vocal communication in adult greater horseshoe bats, Rhinolophus ferrumequinum. Journal of Comparative Physiology A, 192, 535-550. Masters, W. M., Raver, K. A. S. & Kazial, K. A. 1995. Sonar signals of big brown bats, Eptesicus fuscus, contain information about individual identity, age and family affiliation. Animal Behaviour, 50, 1243-1260. Matsumara, S. 1981. Mother-infant communication in a horseshoe bat (Rhinolophus ferrumequinum nippon): vocal communication in three-week-old infants. Journal of Mammalogy, 62, 20-28. McComb, K. & Semple, S. 2005. Coevolution of vocal communication and sociality in primates. Biology Letters, 1, 381-385. Melendez, K. V., Jones, D. L. & Feng, A. S. 2006. Classification of communication signals of the little brown bat. Journal of the Acoustical Society of America, 120, 1095-1102.

62

Miller, P. J. O., Shapiro, A. D., Tyack, P. L. & Solow, A. R. 2004. Call-type matching in vocal exchanges of free-ranging resident killer whales, Orcinus orca. Animal Behaviour, 67, 1099-1107. Miller, C. T., Iguina C. G. & Hauser M. D. 2005. Processing vocal signals during antiphonal calling in tamarins. Animal Behaviour, 69, 1387-1398. Morton, E. S. 1977. On the occurrence and significance of motivation-structural rules in some bird and mammal sounds. American Naturalist, 111, 855-869. Mundry, R. & Sommer C. 2007 (published online). Discriminant function analysis with nonindependent data: consequences and an alternative. Animal Behaviour, in press. Nagel, J. J. 2006. Variation in Distress Calls of New World Bats. M.Sc. Thesis. University of Western Ontario. London, ON, Canada. Nelson, J. 1964. Vocal communication in Australian flying foxes (Pteropodidae: Megachiroptera). Zeitschrift fuer Tierpsychologie, 21, 857-870. Owings, D. H. & Morton, E. S. 1998. Animal Vocal Communication: A New Approach. Cambridge, UK: Cambridge University Press. Page, R. A. & Ryan, M. J. 2006. Social transmission of novel foraging behavior in bats: frog calls and their referents. Current Biology, 16, 1201-1205. Pinker, S. 2003. Language as an adaptation to the cognitive niche. In: Language evolution (Ed. by Christiansen M. H. & Kirby S.), pp. 16–37. Oxford, UK: Oxford University Press. Pfalzer, G. & Kusch, J. 2003. Structure and variability of bat social calls: implications for specificity and individual recognition. Journal of Zoology (London), 261, 21-33. Porter, F. L. 1979. Social behavior in the leaf-nosed bat, Carollia perspicillata II. Social communication. Zeitschrift fur Tierpsychologie, 50, 1-8. Rendall, D., Rodman, P. S. & Emond R. E. 1996. Vocal recognition of individuals and kin in free-ranging rhesus monkeys. Animal Behaviour, 51, 1007-1015. Rendall, D., Cheney, D. L. & Seyfarth, R. M. 2000. Proximate factors mediating “contact” calls in adult female baboons (Papio cynocephalus ursinus) and their infants. Journal of Comparative Psychology, 114, 36-46.

63

Rendall, D. & Owren, M. J. 2002. Animal vocal communication: say what?. In: The Cognitive Animal (Ed. by Bekoff, M., Allen, C. & Burghardt, G. M.), pp. 307314. Cambridge, MA: The MIT Press. Ruczynski, I., Kalko, E. K. V. & Siemers, B. 2007. The sensory basis of roost finding in a forest bat, Nyctalus noctula. Journal of Experimental Biology, 210, 36073615. Russ, J. M., Jones, G., Mackie, I. J. & Racey, P. A. 2004. Interspecific responses to distress calls in bats (Chiroptera: Vespertilionidae): a function for convergence in call design? Animal Behaviour, 67, 1005-1014. Russ, J. M. & Racey, P. A. 2007. Species-specificity and individual variation in the song of male Nathusius’ pipistrelles (Pipistrellus nathusii). Behavioral Ecology and Sociobiology, 61, 669-677. Russo, D. & Jones, G. 1999. The social calls of Kuhl's pipistrelles Pipistrellus kuhlii (Kuhl, 1819): Structure and variation (Chiroptera: Vespertilionidae). Journal of Zoology (London), 249, 476-481. Sailler, H. & Schmidt, U. 1978. Social calls of the common vampire bat Desmodus rotundus during aggressive behavior at the feeding place in the laboratory. Zeitschrift fuer Saeugetierkunde, 43, 249-261. Scherrer, J. A. & Wilkinson, G. S. 1993. Evening bat isolation calls provide evidence for heritable signatures. Animal Behaviour, 46, 847-860. Schmidt, U. 1972. Social calls of juvenile vampire bats Desmodus rotundus and their mothers. Bonner Zoologische Beitraege, 23, 310-316. Schmidt-French, B., Gillam, E. & Fenton, M. B. 2006. Vocalizations emitted during mother-young interactions by captive eastern red bats Lasiurus borealis (Chiroptera: Vespertilionidae). Acta Chiropterologica, 8, 477-484. Schutt, W. A., Muradali, F., Mondol, N., Joseph, K. & Brockmann, K. 1999. Behavior and maintenance of captive white-winged vampire bats, Diaemus youngi. Journal of Mammalogy, 80, 71-81. Schwartz, C., Tressler, J., Keller, H., Vanzant, M., Ezell, S. & Smotherman, M. 2007. The tiny difference between foraging and communication buzzes uttered by the Mexican free-tailed bat, Tadarida brasiliensis. Journal of Comparative Physiology A, 193, 853-863. Silk, J. B. 2007. Social components of fitness in primate groups. Science, 317, 13471351.

64

Siemers, B. M. & Kerth, G. 2006. Do echolocation calls of wild colony-living Bechstein’s bats (Myotis bechsteinii) provide individual-specific signatures? Behavioral Ecology and Sociobiology, 59, 453-454. Skowronski, M. 2007. Acoustic analysis of bat calls: methods for read_marks Matlab program. Available online at: http://publish.uwo.ca/%7Emskowro2/ Snowdon, C. T. & Cleveland, J. 1984. “Conversations” among pygmy marmosets. American Journal of Primatology, 7, 15-20. Soltis, J., Leong, K. & Savage, A. 2005. African elephant vocal communication I: antiphonal calling behaviour among affiliated females. Animal Behaviour, 70, 579-587. Smotherman, M. S. 2007. Sensory feedback control of mammalian vocalizations. Behavioural Brain Research, 182, 315-326. Sproul, C., Palleroni, A. & Hauser, M. 2006. Cottontop tamarin, Saguinus oedipus, alarm calls contain sufficient information for recognition of individual identity. Animal Behaviour, 72, 1379-1385. Sterbing, S. J. 2002. Postnatal development of vocalizations and hearing in the phyllostomid bat, Carollia perspicillata. Journal of Mammalogy, 83, 516-525. Sugiura, H. 1993. Temporal and acoustic correlates in vocal exchanges of coo calls in Japanese macaques. Behaviour, 124, 207-225. Sugiura, H. 1998. Matching of acoustic features during the vocal exchange of coo calls by Japanese macaques. Animal Behaviour, 55, 673-687. Thomson, C. E., Fenton, M. B. & Barclay, R. M. R. 1985. The role of infant isolation calls in mother-infant reunions in the little brown bat, Myotis lucifugus (Chiroptera: Vespertilionidae). Canadian Journal of Zoology, 63, 1982-1988. Tschapka, M. & Wilkinson, G. S. 1999. Free-ranging vampire bats (Desmodus rotundus, Phyllostomidae) survive 15 years in the wild. Zeitschrift fuer Saeugetierkunde, 64, 239-240. Wanker, R., Apcin, J., Jennerjahn, B. & Waibel, B. 1998. Discrimination of different social companions in spectacled parrotlets (Forpus conspicillatus): evidence for individual vocal recognition. Behavioral Ecology and Sociobiology, 43, 197-202. Weiss, D. J., Garibaldi, B. T. & Hauser M. D. 2001. The production and perception of long calls by cotton-top tamarins (Saguinus oedipus): acoustic analyses and playback experiments. Journal of Comparative Psychology, 115, 258-271.

65

Wilkinson, G. S. 1984. Reciprocal food sharing in the vampire bat. Nature, 308, 181184. Wilkinson, G. S. 1985a. The social organization of the common vampire bat Desmodus rotundus I. pattern and cause of association. Behavioral Ecology and Sociobiology, 17, 111-122. Wilkinson, G. S. 1985b. The social organization of the common vampire bat Desmodus rotundus II. mating system genetic structure and relatedness. Behavioral Ecology and Sociobiology, 17, 123-134. Wilkinson, G. S. 1986. Social grooming in the common vampire bat, Desmodus rotundus. Animal Behaviour, 34, 1880-1889. Wilkinson, G. S. 1987. Altruism and co-operation in bats. In: Recent Advances in the Study of Bats (Ed. By Fenton, M. B., Racey, P., & Rayner, J. M. V.), pp 299323. Cambridge, UK: Cambridge University Press. Wilkinson, G. S. 1988. Social organization and behavior. In: Natural History of Vampire Bats (Ed. by Greenhall, A. M. & Schmidt, U.), pp. 85-97. Boca Raton, FA: CRC Press. Wilkinson, G. S. 1992. Information transfer at evening bat colonies. Animal Behaviour, 44, 501-518. Wilkinson, G. S. 2003. Social and vocal complexity in bats. In: Animal Social Complexity (Ed. by de Waal, F. B. M. & Tyack, P. L.), pp. 322-341. Cambridge, MA: Harvard University Press. Wilkinson, G. S. & Boughman, J. W. 1998. Social calls coordinate foraging in greater spear-nosed bats. Animal Behaviour, 55, 337-350. Yoshida, S. & Okanoya, K. 2005. Evolution of turn-taking: a bio-cognitive perspective. Cognitive Studies, 12, 153-165. Yosida, S., Kobayasi, K. I., Ikebuchi, M., Ozaki, R. & Okanoya, K. 2007. Antiphonal vocalization of a subterranean rodent, the naked mole-rat (Heterocephalus glaber). Ethology, 113, 703-710. Zhang, L., Jones, G., Parsons, S., Liang, B. & Zhang, S. 2005. Development of vocalizations in the flat-headed bats, Tylonycteris pachypus and T. robustula (Chiroptera: Vespertilionidae). Acta Chiropterologica, 7, 91-99.

66

Appendix 17 White-winged vampire bats (Diaemus youngi) used in this study. NM= housed at New Mexico Bat Research Institute, Tijeras, New Mexico, USA. NY= housed at Rosamund-Gifford Park Zoo, Syracuse, New York, USA. Letter

Name

Cage

Colony

Age/Sex

A

Amber

1

NM

Adult/Female

B

MaryBee

1

NM

Adult/Female

C

Cici

1

NM

Adult/Female

D

Daniela

1

NM

Adult/Female

E

Emily

1

NM

Adult/Female

F

Farouk

1

NM

Adult/Male

G

GaryMcCracken

2

NM

Adult/Male

H

Hermanson

2

NM

Adult/Male

I

Isaac

2

NM

Adult/Male

J

JerryWilkinson

2

NM

Adult/Male

K

Kristin

3

NM

Adult/Female

L

Laurie

3

NM

Adult/Female

M

MelvilleMerlin

4

NM

Adult/Male

captive born

N

Nutella

4

NM

Adult/Female

captive born

O

Oatmeal

4

NM

Adult/Male

captive born

P

Punk

4

NM

Adult/Male

captive born

R

-

3

NM

Subadult/?

captive born to K or L

S

Syracuse

5

NY

Adult/Male

unfamiliar to others

67

Notes wild-caught

Curriculum Vitae Gerald Gunnawa Carter born: April 28, 1983 Education B. Sc. Cornell University, 2001-2005 Honours Thesis: Noninvasive DNA-based identification of avian hosts of vampire bats M.Sc. University of Western Ontario, Jan 2006- Dec 2007 (expected) Masters Thesis: Vocal communication in white-winged vampire bats Peer-reviewed publications Carter, G., Coen, C., Stenzler, L. & Lovette, I. 2006. Avian host DNA isolated from the feces of white-winged vampire bats (Diaemus youngi). Acta Chiropterologica, 8, 255-259. Carter, G. & Riskin, D. 2006. Mystacina tuberculata. Mammalian Species, 790, 1-8. Riskin, D., Parsons, S., Schutt, W., Carter, G. & Hermanson, J. 2006. Terrestrial locomotion of the New Zealand short-tailed Bat (Mystacina tuberculata) and the Common Vampire Bat (Desmodus rotundus). Journal of Experimental Biology, 209, 1725-1736. Non-refereed publications Carter, G. 2005. Bat diversity and abundance in Cusuco National Park core zone, Honduras. Internal report for Operation Wallacea, UK. Carter, G. 2004. A field key to the bats of Trinidad. Distributed to to the Wildlife Section of the Forestry Division, Ministry of Agriculture, Trinidad. Conference presentations Carter, G., Skowronski, M., Faure, P. & Fenton, B. 2007. Vocal communication in white-winged vampires. 14th International Bat Research Conference, Merida, Yucatan, Mexico. Carter, G., Skowronski, M., Fenton, B. & Faure, P. 2007. Do adult vampires

68

duet at night? Antiphonal calling among adult white-winged vampire bats (Diaemus youngi). 87th American Society of Mammalogists Meeting, Albuquerque, NM. Carter, G. 2006. Antiphonal calling behaviour in white-winged vampires (Diaemus youngi). 36th North American Symposium on Bat Research, Wilmington, NC. Carter, G., Lovette, I., Hermanson, J. 2004. Noninvasive identification of the avian host species of white-winged vampire bats (Diaemus youngi) from fecal samples. 34th North American Symposium on Bat Research, Salt Lake City, Utah. (Winner of 2004 Bat Research News Award) Awards and scholarships 2007 Graduate Student Teaching Award, University of Western Ontario 2007 Ontario Graduate Scholarship (only 3% awarded to non-Canadian citizens, $10,000) 2005 Magna cum laude, Cornell University, "with Honors and Distinction in Research" 2005 Paul Schreurs Memorial Award, Cornell University (for undergraduate research, $500) 2004 Bat Research News Award (for "outstanding oral paper", $500) 2001-2005 Robert and Helen Appel Presidential Research Scholarship ($10,000) Relevant teaching experience 2007 TA for Introductory Biology tutorials, University of Western Ontario (UWO) 2006 TA for Animal Behaviour (Bio 336), UWO* ( *2007 Graduate Student Teaching Award) 2006 TA for Introductory Biology (Bio 22/23), laboratory, UWO 2004 TA for Field Biology (NTRES 212), Cornell University Relevant technical experience 2006 Operation Wallacea Bat Specialist- hired to conduct first survey of bat fauna of Cusuco National Park core zone, Honduras 2003-2004 Field assistant to Dan Riskin- mistnetting, filming, design of experimental components, data collection, and calibration of equipment 2002-2005 Research assistant to Dr. I. J. Lovette, Evolutionary Biology molecular lab, Cornell Laboratory of Ornithology.

69