Running head: BATS UTILIZE HUMAN REFERENTIAL STIMULI 1 12/20/2014 1 Megachiropteran Bats Utilize Human Referential Stimuli to Locate Hidden Food Nathaniel J. Hall, Monique A. R. Udell, Nicole R. Dorey, and Allyson L. Walsh University of Florida Clive D.L. Wynne University of Florida and Kyung Hee University, South Korea Author Note Nathaniel J. Hall, Monique A.R. Udell, Nicole R. Dorey and Clive D.L. Wynne, Department of Psychology, University of Florida Allyson L. Walsh, Department of Wildlife Ecology and Conservation, University of Florida We are grateful for the help and support provided by The Lubee Bat Conservancy, Gainesville, FL, USA, and thank them for the use of their animals and facilities; specifically we wish to thank Brian Pope and Tasha King who contributed greatly to this project. Correspondence concerning this article should be addressed to Monique A. R. Udell, University of Florida, Department of Psychology P.O. Box 112250 Gainesville, Florida 32611 or fax (352) 392-7985 Email mudell@ufl.edu BATS UTILIZE HUMAN REFERENTIAL STIMULI 2 Abstract 2 Spontaneous point-following behavior has been considered an indicator of advanced 3 social cognition unique to humans. Recently, it has been suggested that a close 4 evolutionary relationship with humans could result in similar social skills in 5 domesticated species. An alternative view is that the mechanism is not genetic 6 domestication alone but instead a combination of phylogenetic and ontogenetic 7 variables. Here we test the necessity of phylogenetic domestication by investigating the 8 point-following behavior of a captive population of non-domesticated megachiropteran 9 bats (Pteropus pumilus, Pteropus rodricensis, Pteropus conspicillatus, Pteropus 10 vampyrus). Three of five subjects were highly successful in following an unfamiliar 11 human’s point to a target location providing the first empirical evidence of cross-species 12 social referencing in bats. The three successful bats were all born in captivity and 13 socialized to humans early in life whereas unsuccessful bats were wild born individuals. 14 This study provides evidence that referential point following is not restricted to 15 domesticated animals and indicates that early experience may be important. 16 Megachiropteran bats may prove to be a useful model for studying social behaviors. 17 18 19 20 21 22 23 24 Keywords: megachiropteran bats, social cognition, human gestures, social referencing BATS UTILIZE HUMAN REFERENTIAL STIMULI 25 3 Megachiropteran Bats Utilize Human Referential Stimuli to Locate Hidden Food 26 Joint attention and point following are considered important markers of socio- 27 cognitive development in human infants and are often used as indicators for theory of 28 mind capabilities in pre-verbal human children (Carpenter, Nagell & Tomasello, 1998; 29 Charman et al., 2000). Recently, comparative research on spontaneous point following 30 behavior in an object-choice task has been used to support the claim that domestic dogs 31 and possibly domesticated species in general, have evolved human-like social skills 32 (Hare & Tomasello, 2005). This domestication hypothesis proposes that selection 33 pressures present during domestication led to the development of heritable human-like 34 social skills that increased genetic fitness (Hare, Brown, Williamson & Tomasello, 35 2002; Miklósi et al., 2003). Thus, dogs, and domesticated species in general, have 36 evolved a specialized capacity to read human social and communicative behavior (Hare 37 & Tomasello, 2005) 38 Support for the domestication hypothesis comes from evidence demonstrating 39 that some domesticated species (goats, Capra hirucs; horses, Equus caballus; cats, Felis 40 catus) show proficiency in following a human point to a target without explicit training 41 (for goats see Kaminski, Riedel, Call & Tomasello, 2005; for horses see Maros, Gácsi & 42 Miklósi, 2008; McKinley & Sambrook, 2000; for cats see Miklósi, Pongrácz, Lakatos, 43 Topál & Csányi, 2005). However, to determine if domestication is necessary for the 44 development of responsiveness to human social and communicative gestures, adequate 45 comparisons to the social behavior of non-domesticated species must also be made. 46 Most of the available literature comparing domesticated and non-domesticated 47 species’ point-following behavior has focused on canids, specifically pet dogs (Canis 48 lupus familiaris) and wolves (C. l. lupus) (Agnetta, Hare & Tomasello, 2000; Gácsi et BATS UTILIZE HUMAN REFERENTIAL STIMULI 4 49 al., 2009; Hare et al. 2002; Miklósi et al., 2003; Udell, Dorey & Wynne, 2008; Virányi 50 et al. 2008). Some researchers have found that wolves do indeed follow human pointing 51 gestures spontaneously (Gácsi et al., 2009; Udell, Dorey & Wynne, 2008), while others 52 report that untrained wolves perform at chance levels (Hare et al., 2002; Virányi et al., 53 2008). In a related line of investigation, Hare et al. (2005) reported that silver foxes 54 artificially bred over many generations for tame behavior (Trut, 1999), were 55 spontaneously more sensitive to human pointing gestures than wild-type foxes. 56 However, the wild type foxes in this study nonetheless followed the experimenter’s 57 point more often than would be expected by chance, suggesting that domestication may 58 have influenced performance but did not predict absolute success or failure on the task. 59 While there is strong evidence for the interplaying roles of phylogeny and ontogeny in 60 the development of dogs’ social skills (Udell & Wynne, 2010), the precise role of 61 domestication is still an ongoing debate within the canid literature. 62 Other non-domesticated species tested for sensitivity to human gestures include 63 dolphins (Tursiops truncates) and fur seals (Arctocephalus pusillus) (Pack & Herman, 64 2004; Scheumann & Call, 2004; Tschudin, Call, Dunbar, Harris & van der Elst, 2001). 65 While some subjects in these studies utilized human points at above chance levels, their 66 status as animals trained for public display or other experiments has sometimes made 67 direct comparisons with untrained domesticated animals difficult (Miklósi & Soproni, 68 2006). More recently, an experiment conducted with a species known to engage in 69 cooperative social behavior in the wild, Jackdaws (Corvus monedula), demonstrated 70 that hand raised individuals would spontaneously follow a point issued by their familiar 71 caretaker (Von Bayern & Emery, 2009). The findings of this study suggest that giving a 72 species that engages in cooperative social behaviors with conspecifics early experiences BATS UTILIZE HUMAN REFERENTIAL STIMULI 5 73 with humans may be important for spontaneous point following behavior. The degree to 74 which point following behaviors may extend to unfamiliar experimenters is unknown. 75 Both Megachiropteran and Microchiropteran bats are considered generally social 76 (for a review see Kerth, 2008). Various bat species have been recorded engaging in 77 conspecific grooming, transfer of information, reciprocal food sharing, and mutual 78 warming (Kerth, 2008; McCracken & Bradbury, 1981; Safi & Kerth, 2007; Wilkinson, 79 1984; Wilkinson, 1986; Wilkinson, 1992). Of the species in this study, Pteropus 80 rodricensis have been recorded aiding conspecifics in parturition (Kunz& Allgaier, 81 1994). In addition, Pteropus vampyrus are considered a gregarious species (Kunz & 82 Jones, 2000). Unfortunately, detailed information on many bat species’ social 83 organization is sparse compared to what is known of other social species (Kerth, 2008). 84 Furthermore, Megachiropterans have vision adequate to participate in traditional object 85 choice tasks (Müller, Goodman & Piechl, 2007). 86 In this study we investigated the performance of five captive bats, belonging to 87 the suborder Megachiroptera, in an object choice task which required them to 88 spontaneously utilize the point of an unfamiliar human to locate a hidden target. 89 90 Methods Subjects/Setting 91 Five megachiroptera from four different species currently residing at a bat 92 conservancy were selected for participation in this study. The bats were kept in outdoor, 93 enclosed, octagonal pens (side length of 4.52 m) with other bats of varying species. 94 While a familiar keeper acted as an assistant, calling or returning the subject back to the 95 starting point between trials, the experimenter who issued the pointing gesture during 96 experimental trials was previously unfamiliar to the bat subjects. BATS UTILIZE HUMAN REFERENTIAL STIMULI 97 6 Three of the five bats were captive-born (Pteropus pumilus, Pteropus 98 rodricensis, Pteropus conspicillatus), while two bats were wild-caught (Pteropus 99 pumilus, Pteropus vampyrus). The captive-born bats had varying histories (mother 100 reared or hand reared on or off the premises of the conservancy) but are distinct from 101 wild-caught bats in that they were born in captivity, provided regular interactions with 102 humans from birth, and reached adulthood in captivity. Wild-caught bats matured to 103 adulthood before being captured and brought to the conservancy, but had spent about 16 104 years in captivity prior to testing. The captive-born bats were also mature adults at 8, 10 105 and 11 years of age and had spent their entire lives in captivity. Thus, the major 106 difference between the groups was that captive-born bats had early experiences with 107 humans handling or entering their enclosure, while wild-caught bats did not. Once at the 108 conservancy, wild-caught and captive-born bats cohabitated and had similar experiences 109 with humans and feeding. 110 Materials 111 The bats were tested individually in vacant triangular outdoor pens (6.2 m x 112 4.52 m x 4.52 cm). Two opaque plastic containers served as the response objects and 113 were fixed 1.6 m apart at an appropriate height for the individual bat to reach (see figure 114 1). Each response object held a 250 ml Mason jar that was out of sight of the bat (see 115 figure 2). Both Mason jars contained 1ml of Kern’s Nectar™ (Kern's Beverages, LLC 116 4002 Westminster Ave. Santa Ana, CA 92703). The lids normally used to seal Mason 117 jars were removed, but the rims used to affix the lid to the jars were retained for the 118 purposes of this study. 119 120 The rims were placed on both Mason jars, however, one jar’s rim was fitted with a fiberglass screen (sham rim) while the other jar’s rim was not manipulated (free rim) BATS UTILIZE HUMAN REFERENTIAL STIMULI 7 121 (figure 2A).The screen on the sham rim permitted air flow from both jars, and hence 122 equalized smell from the jars, while preventing the bat access to the nectar in the sham 123 jar. The other rim provided free access to the nectar. As demonstrated previously, 124 hiding equal amounts of nectar in each box was necessary, because some species of 125 megachiroptera can not only determine the location of food by odor alone in a two 126 choice experiment, but can even discriminate whether the fruit is ripe (Luft, Curio & 127 Tacud, 2003). One bat in our study, Easter, was also initially suspected of using 128 additional scent cues that she may have left on the rim of the free lid when allowed to 129 access the free food during her first set of control trials. This potential confound was 130 resolved by cleaning both jars between each trial, both control and experimental, after 131 which her performance on control trials dropped to chance while her performance on 132 experimental trials remained significantly above chance. 133 Experimental Testing 134 Prior to participation in the study each bat was required to readily approach and 135 take food from the experimenter and experimental apparatus when it was freely 136 available. This was done on the day of testing by the experimenter holding up a 137 container of nectar next to the bat. If the bat approached the container, the container was 138 placed inside one of the experimental boxes. This procedure was repeated until the bat 139 consumed food from the each experimental box twice. The procedure ensured that the 140 bat subjects were food motivated, and were not fearful or distracted within the testing 141 environment. Four bats never approached the experimenter or the experimental 142 apparatus despite being given hours to habituate to the apparatus. These four bats never 143 began testing. BATS UTILIZE HUMAN REFERENTIAL STIMULI 144 8 Each experimental trial began with the assistant bringing the bat to the starting 145 location 1m back from the experimental boxes (figure 2B). Once the bat was in position 146 the experimenter raised his ipsilateral arm to chest height and directed his hand towards 147 the bat. The point started this way to allow for a range of motion of the arm and to gain 148 the bat’s attention. The experimenter called the bat’s name, and then moved his arm 149 laterally in the direction of the correct box. Movement stopped once the arm was 150 pointing to the correct box and the finger was approximately 12 cm from the box (figure 151 2C). The gesture, a dynamic proximal point as defined by Miklósi & Soproni (2006), 152 did not move further until the bat made a choice or the trial timed out after 2 minutes. 153 The point is considered dynamic because the subject is able to see the movement of the 154 point, and the arm remains in the gesturing position throughout the trial. While 155 gesturing, the experimenter looked straight ahead and avoided eye contact. 156 Trials lasted 2 minutes to give the bats sufficient time to move from the start 157 location to the target container by pulling themselves across the enclosure roof (the 158 bats’ locomotor method of choice) in a climbing motion. None of the bats flew in the 159 experimental pen during testing. An observer outside the enclosure recorded the bats’ 160 choice for each trial as it occurred. The observer recorded a choice when the bat 161 physically touched one of the response objects (figure 2D), which were located 1.6 m 162 apart. In every trial where a bat made a choice, the bat subsequently entered the 163 response object to obtain accessible or inaccessible nectar. Thus, the bats’ choice 164 between response objects was unambiguous. If a bat made a correct choice, but had 165 difficulties in consuming the nectar, the experimenter assisted the bat by lifting the jar 166 to increase accessibility. If the bat made an incorrect choice, the assistant called the bat 167 back to the start position without it receiving nectar. If the bat moved towards the BATS UTILIZE HUMAN REFERENTIAL STIMULI 9 168 testing apparatus but had not made a choice within two minutes, the assistant called the 169 bat back to start the next trial and a ‘no choice’ was recorded. No choice responses 170 were analyzed as incorrect responses. Only one ‘no choice’ response occurred during 171 experimental trials (Easter, session 1). 172 After each trial, the experimenter removed both jars from the response objects. 173 If the bat had made a correct choice on the previous trial and subsequently consumed 174 the nectar from the jar in the target container, the experimenter removed both lids from 175 the jars and re-filled the emptied jar with 1ml of nectar. If the bat had made an incorrect 176 choice on the previous trial (and therefore did not consume the nectar), the experimenter 177 removed the jars and the lids, but did not add any more nectar to the jars. Out of the 178 view of the bat, the experimenter simultaneously replaced both lids and placed the lid 179 allowing access to the nectar on the appropriate jar that was then put into the target 180 container. Both jars were then placed in their respective boxes. 181 The target container for each trial was determined pseudorandomly with the 182 stipulation that no side could be used more than twice in a row or for more than 50% of 183 the trials. 184 Each subject received a total of 20 experimental trials, divided into two sessions 185 of ten trials each. If a bat satiated with nectar, experimental trials for that day were 186 suspended. Satiation was determined when the bats failed to take food offered by the 187 assistant (a familiar caretaker). Experimental trials were resumed on another day. Four 188 of the five bats participating in experimental trials completed testing, while one bat, 189 Arthur, refused to participate after the first six trials. 190 Control trials BATS UTILIZE HUMAN REFERENTIAL STIMULI 191 10 Three to seven control trials followed every 10 experimental trials. Control 192 trials were identical to experimental trials, except that no pointing cue was administered. 193 Statistical analysis 194 A one-sample t-test was used to determine if the group performance of the bats 195 was better than would be expected by chance within each session. A paired t-test was 196 used to determine if performance differed between session one and session two. Both 197 tests were two-tailed and had an alpha level of .05. Binomial tests were used to assess 198 the individual performance of each bat for each session, with success criterion set at 199 8/10 trials correct or better within an individual session (p< .05). Analyses were 200 conducted using Excel and SPSS. 201 Results 202 As a group, the bats were successful in following a human point in each testing 203 session (one sample t-test, session 1: t3 = 4.37, p = .022; session 2: t3 = 7.35, p = .005). 204 At the individual level, all three captive-born bats, Grace, Kuri, and Easter, were 205 successful in reliably following a human point to the target in each session (session 1 206 and session 2: binomial test, p < .05, see figure 3), and pooled across both sessions 207 (Grace: 16 out of 20, p ≤ .01; Kuri: 18 out of 20, p < .01; Easter: 16 out of 20 p ≤ .01). 208 Neither wild-caught bat performed significantly above chance on the task. Arthur, one 209 of the wild-caught bats, began testing but only completed six experimental trials before 210 refusing to participate, even after multiple breaks and revisits on other days. Out of 211 those six completed trials, he chose correctly only twice (33% correct). Taba, the other 212 wild-caught bat, readily approached a container in search of the accessible food on 213 every trial, however her performance never reached statistical significance (session 1 214 and session 2: binomial test, p > .05). BATS UTILIZE HUMAN REFERENTIAL STIMULI 11 215 We also analyzed first trial performance for each bat, and compared performance in 216 the first session to that in the second session of testing to assess whether performance 217 was more likely a product of a bat’s capacity before testing or learning within the course 218 of the experiment. There was no difference in performance between the first and second 219 sessions of testing (paired t-test, t3 = 1.00, p = .39). A bat’s first response on the first 220 trial was not a good predictor of subsequent performance. The three successful bats 221 (Easter, Grace and Kuri) made a correct choice on the first trial of testing, so did one of 222 the unsuccessful bats, Taba. Arthur was the only bat to make an incorrect response on 223 the first trial. 224 Control trials were conducted during and after testing in which the bat was 225 allowed to choose a container in the absence of a human point. Bats performed at 226 chance levels on control trials, both on average and at the individual level (mean 227 performance on control trials for all subjects, 49% correct, binomial test, p = .50), 228 indicating that they were not locating the accessible food based on smell or 229 unintentional cueing by the experimenter. As mentioned above, Easter was initially 230 suspected of using smell cues to locate the target container on control trials. This was 231 resolved by cleaning the lids and jars in between trials. Prior to cleaning the lids she 232 scored 8 correct out of 10 trials; after the introduction of the cleaning lid method, 233 performance on control trials dropped to chance, 6 out of 14 (43% correct, binomial 234 test, p = .77). Performance on experimental trials remained above chance despite 235 cleaning between trials (80% correct, binomial test, p ≤ .05). The control trials of the 236 other bats did not indicate the use of any unintentional cues (Taba session 1 and session 237 2: 50% correct on controls; Grace session 1 and session 2: 25% correct; Kuri session 1 238 and session 2: 55% correct; Arthur did not complete control trials). BATS UTILIZE HUMAN REFERENTIAL STIMULI 239 12 Discussion 240 These results demonstrate that several species of experimentally naïve bats are 241 capable of spontaneously utilizing human points to find the location of concealed food 242 in an object-choice task. This suggests that domestication is not necessary for the 243 responsiveness to the human gesture tested. Instead, the proclivity of bats towards 244 conspecific social interactions may be important in understanding their ability to engage 245 in heterospecific interactions such as following human points. Nonetheless, 246 responsiveness to the actions of heterospecifics does not appear to be automatic. 247 Human socialization and regular interaction with humans from an early age emerged as 248 an important predictor of an individual bat’s performance. All three captive-born 249 individuals that participated in the experiment followed the experimenter’s point to the 250 target location successfully. Both wild-born bats were unsuccessful on the task despite 251 having spent 16 years of their adulthood in captivity. 252 Early and intensive exposure to humans has been linked to the success of other non- 253 domesticated species on human-guided tasks (Udell et al., 2008; Von Bayern & Emery, 254 2009). Taken together with this study, this may suggest that there is an important 255 sensitive period for socialization that can help predict success on human guided tasks. 256 However, an additionally important factor in this study was that the successful bat 257 subjects were able to follow the point of an unfamiliar experimenter; demonstrating the 258 ability to generalize their response to referential stimuli provided by humans in general. 259 While in previous reports, chimpanzees have been reported to perform at lower 260 levels than both human children and domestic dogs on human guided tasks (Hare et al., 261 2002; Bräuer, Kaminski, Riedel, Call & Tomasello, 2006), this might have been due, at 262 least in part, to significant differences in the quantity and quality of daily exposure to BATS UTILIZE HUMAN REFERENTIAL STIMULI 13 263 humans. Differences between enculturated and wild born chimpanzees indicate 264 environmental factors contribute to the ontogeny of primate social cognition (Tomasello 265 & Call, 2004). Tomasello and Call hypothesized that growing up in the presence of 266 humans changed the way individuals attended and reacted to human actions and 267 enculturation acted as a catalyst for further development of an already present socio- 268 cognitive capacity (Tomasello & Call, 2004). 269 The individual socialization experiences of the bats in this study differ only at 270 birth until one to two years of age. Captive-born individuals received at least some close 271 human interaction from birth, where wild-caught bats did not. Experiences as adults in 272 captivity were nearly identical for all subjects. The hypothesized role of early 273 socialization in the development of responsiveness to human gestures is consistent with 274 the differences found between our captive-born and wild-caught bats thus far. 275 The success of species generally considered social may indicate an important 276 phylogenetic component to point following behavior in object choice tasks. Species that 277 regularly interact with conspecifics may be more apt at cooperative interactions with 278 heterospecifics. Thus a two-pronged strategy may be needed to further identify species 279 and individuals likely to display sensitivity to human gestures: 1. recognizing species 280 characteristics that suggest a phylogenetic capacity, including evidence of conspecific 281 social interactions of the species, while 2. also taking into account ontogenic factors. 282 With the current sample size, accurate first trial responding was difficult to 283 interpret conclusively, however larger sample sizes might be more conducive to a first 284 trial analysis and may be able to provide information about spontaneous responding in 285 future studies. In addition, further research is needed to determine whether the bat’s 286 performance should be explained as a learned association between human hands and the BATS UTILIZE HUMAN REFERENTIAL STIMULI 14 287 presentation of food, a form of local enhancement, stimulus enhancement, or as 288 requiring an understanding of reference or intentionality. More point types should also 289 be used in future studies to ascertain whether success with proximal pointing would 290 extend to more distal points. With the current data, we propose that the social proclivity 291 of bats and early exposure to humans during ontogeny both likely contributed to the 292 development of increased responsiveness to humans. 293 Bats could serve as an important non-domesticated animal model for investigating 294 the origins of human-like social cognition. More generally, bats are rich but 295 underrepresented subjects in animal cognition research, and ought to be better 296 represented in future studies. 297 298 299 300 301 302 303 304 305 306 307 308 309 310 BATS UTILIZE HUMAN REFERENTIAL STIMULI 311 312 313 314 15 References Agnetta, B., Hare, B., & Tomasello, M. (2000). Cues to food location that dogs (Canis familiaris) of different ages do and do not use. Animal Cognition, 3, 107-112. Bräuer, J., Kaminski, J., Riedel, J., Call, J., & Tomasello, M. (2006). Making 315 inferences about the location of hidden food: social dog, causal ape. Journal of 316 Comparative Psychology, 120, 38-47. 317 Carpenter, M., Nagell, K., & Tomasello, M. (1998). Social cognition, joint attention, 318 and communicative competence from 9 to 15 months of age. Monographs of the 319 Society for Research in Child Development, 6, 1-174. 320 Charman, T., Baron-Cohen, S., Swettenham, J., Baird, G., Cox, A., & Drew, A. 321 (2000). Testing joint attention, imitation, and play as infancy precursors to 322 language and theory of mind. Cognitive Development 15, 481-498. 323 Gácsi, M., Győri, B., Virányi, Z., Kubinyi, E., Range, F., Belényi, B., & Miklósi, Á. 324 (2009). Explaining dog wolf differences in utilizing human pointing gestures: 325 selection for synergistic shifts in the development of some social skills. Plos 326 One, 4, e6584, doi:10.1371/journal.pone.0006584. 327 Hare, B., Plyusnina, I., Ignacio, N., Schepina, O., Stepika, A., Wrangham, R., & Trut, L. 328 (2005). Social cognitive evolution in captive foxes is a correlated by-product 329 of experimental domestication. Current Biology, 15, 226-230. 330 331 332 333 334 Hare, B., Brown, M., Williamson, C., & Tomasello. M. (2002). The domestication of social cognition in dogs. Science, 298, 1634-1636. Hare, B., & Tomasello, M. (2005). Human like social skills in dogs? TRENDS in Cognitive Science, 9, 439-444. Kaminski, J., Riedel, J., Call, J., & Tomasello, M. (2005). Domestic goats, Capra BATS UTILIZE HUMAN REFERENTIAL STIMULI 16 335 hircus, follow gaze direction and use social cues in an object choice task. Animal 336 Behaviour, 69, 11-18. 337 338 Kerth, G. (2008). Causes and consequences of sociality in bats. Bioscience, 58, 737746. 339 Kunz, T.H., & Allgaier, A.L. (1994). Allomaternal care: Helper-assisted birth in the 340 rodrigues fruit bat, Pteropus rodricensis (chiroptera: Pteropodidae). Journal of 341 Zoology, 232, 691-700. 342 Kunz, T.H., & Jones, D.P. (2000). Pteropus vampyrus. Mammalian Species, 642, 1-6. 343 Luft, S., Curio, E., & Tacud, B. (2003). The use of olfaction in the foraging behavior 344 of the golden-mantled flying fox, Pteropus pumilus, and the greater musky fruit 345 bat, Ptenochirus jagori (Megachiroptera: Pteropodidae). Naturwissenschaften, 346 90, 84-87. 347 Maros, K., Gácsi, M., & Miklósi, Á. (2008). Comprehension of human pointing 348 gestures in horses (Equus caballus). Animal Cognition, 11, 457-466. 349 McCracken, G.F., & Bradbury, J.W. (1981). Social organization and kinship in the 350 polygynous bat Pyllostomus hastatus. Behavioral Ecology and Sociobiology, 8, 351 11-34. 352 353 354 McKinley, J., & Sambrook, T. (2000). Use of human-given cues by domestic dogs (Canis familiaris) and horses (Equus caballus). Animal Cognition, 3, 13-22. Miklósi, Á., Kubinyi, E., Topál, J., Gácsi, M., Virányi, Z., & Csányi, V. (2003). A 355 simple reason for a big difference: wolves do not look back at humans, but dogs 356 do. Current Biology,13(9),764-766. 357 Miklósi, Á., Pongrácz, P, Lakatos, G., Topál, J., & Csányi, V. (2005). A comparative BATS UTILIZE HUMAN REFERENTIAL STIMULI 358 study of the use of visual communicative signals in interactions between dogs 359 (Canis familiaris) and humans and cats (Felis catus) and humans. Journal of 360 Comparative Psychology, 119, 179-186. 361 362 363 17 Miklósi, Á., & Soproni, K. (2006). A comparative analysis of animals' understanding of the human pointing gesture. Animal Cognition, 9, 81-93. Pack, A. A., & Herman, L. M. (2004). Bottlenosed dolphins (Tursiops truncatus) 364 comprehend the referent of both static and dynamic human gazing and pointing in 365 an object-choice task. Journal of Comparative Psychology, 118, 160-171. 366 Safi, K., & Kerth, G. (2007). Comparative analyses suggest that information transfer 367 promoted sociality in male bats in the temperate zone. American Naturalist, 170, 368 465-472. 369 370 371 372 373 374 375 Scheumann, M., & Call, J. (2004). The use of experimenter-given cues by south african fur seals (Arctocephalus pusillus). Animal Cognition, 7, 224-230. Tomasello, M., & Call, J. (2004). The role of humans in the cognitive development of apes revisited. Animal Cognition, 7, 213-215. Trut, L. 1999. Early canid domestication: the farm-fox experiment. American Scientist, 87, 160-169. Tschudin, A., Call, J., Dunbar, R. I. M., Harris, G., & van der Elst, C. (2001). 376 Comprehension of signs by dolphins (Tursiops truncatus). Journal of Comparative 377 Psychology, 115, 100-105. 378 Udell, M. A. R., Dorey, N. R., & Wynne, C. D. L. (2008). Wolves outperform dogs in 379 following human social cues. Animal Behaviour, 76, 1767-1773. 380 doi:10.1016/j.anbehav.2008.07.028. BATS UTILIZE HUMAN REFERENTIAL STIMULI 381 Udell, M. A. R., & Wynne, C. D. L. (2010). Ontogeny and phylogeny: both are essential 382 to human-sensitive behavior in the genus Canis. Animal Behaviour, 9, e9-e14. 383 doi:10.1016/j.anbehav.2009.11.033. 384 Virányi, Z, Gácsi, M, Kubinyi, E, Topál, J, Belényi, B., Ujfalussy, D., & Miklósi, Á. 385 (2008). Comprehension of human pointing gestures in young human-reared 386 wolves (Canis lupus) and dogs (Canis familiaris). Animal Cognition, 11, 373- 387 387. 388 18 Von Bayern, A.M.P., & Emery, N.J. (2009). Jackdaws respond to human attentional 389 states and communicative cues in different contexts. Current Biology 19, 602- 390 606. 391 392 393 394 395 396 397 398 399 400 401 402 403 404 Wilkinson, G.S. (1984). Reciprocal food sharing in the vampire bat. Nature, 308, 181184. Wilkinson, G.S. (1986). Social grooming in the common vampire bat desmodusrotundus. Animal Behaviour, 34, 1880-1889. Wilkinson, G.S. (1992). Information transfer at evening bat colonies. Animal Behaviour, 44, 501-518. BATS UTILIZE HUMAN REFERENTIAL STIMULI 19 405 406 Figure 1. Testing Layout. Experimenter 1 (E1) pointed to the target object and 407 continued to point until the bat made a choice. A familiar caretaker, Experimenter 2 408 (E2), served as the call back experimenter. E2 retrieved the bat and placed the bat at the 409 correct starting place before every trial begun. 410 411 412 413 414 415 416 417 418 419 420 421 BATS UTILIZE HUMAN REFERENTIAL STIMULI 20 422 423 424 Figure 2. Experimental Materials and Design. (A) The jar on the left has an open lid 425 making the nectar inside accessible; this jar was placed in the target container. The right 426 jar has a wire mesh lid making the nectar inaccessible, this jar was located in the 427 incorrect testing container. (B) A subject being recalled to the start by the assistant. Bats 428 travelled between the assistant and experimenter by pulling themselves across the 429 roofing. (C) An unfamiliar experimenter is making a dynamic proximal point towards 430 the target container. (D) The subject has made a correct response, entered the target 431 container and is obtaining nectar. BATS UTILIZE HUMAN REFERENTIAL STIMULI 21 432 433 Figure 3. Individual performance in each testing session. The number of successful 434 trials, out of 10 for each session, for each bat that completed testing is displayed. Three 435 captive-born bats (Easter, Grace and Kuri) and one wild-caught bat (Taba) are shown. 436 The second wild-caught bat (Arthur) did not complete testing (completed only 6 trials, 437 two of which were correct choices) and is not shown in this figure. * indicates 438 performance significantly above chance in that session (binomial test, p ≤ .05). The 439 dashed line indicates chance responding on the task.