Playback response data for "Common vampire bat contact calls attract past food-sharing partners"
Subjects and social data
Subjects were 31 common vampire bats (Desmodus rotundus), from five matrilines and three zoo populations, housed and cared for by the Organization of Bat Conservation at the Cranbrook Institute of Science (Bloomfield Hills, Michigan) under permits from the USDA (34-C-0117) and the US Fish and Wildlife Service (MB003342-0). Bats were housed at 25-28 degrees Celsius with >33% humidity on a 12 h light/dim light cycle in a 3 x 1.5 x 2 m cage with a soft paper bedding floor that allowed flight and free association among cagemates. Bats were fed on physically or chemically defibrinated bovine blood that was available ad libitum and changed twice daily. All procedures with animals were approved by the University of Maryland Institutional Animal Care and Use Committee (Protocol R-10-63).
Measures of food sharing history were derived from a food-sharing network involving 355 trial-donations and more than 1250 regurgitation observations made from 2010—2014. To quantify food sharing history we used a food sharing index from A to B, which we calculated as the natural log of (X+1) where X is the food sharing rate from A to B (s of mouth-licking per 1 h of experimental trial). To calculate relatedness, we genotyped 19 microsatellite markers (Carter and Wilkinson 2015) and used the Wang estimator (Wang 2002) in the R package related (Pew et al. 2015). We defined kinship by substituting zeros for all negative relatedness values and replacing relatedness values of known maternal kin with the appropriate pedigree kinship values (e.g. mother = 0.5, grandmother = 0.25). Findings remained the same whether we used kinship or relatedness.
Isolated vampire bats produce multi-harmonic contact calls (fundamental typically sweeping down from 31—20 kHz over 6—24 ms) that are variable between and within individual bats (Carter et al. 2012). We recorded calls by placing an adult individual in a small soft mesh butterfly cage at a distance of 10–30 cm of a CM16 ultrasound condenser microphone (frequency range 10–200 kHz, Avisoft Bioacoustics, Berlin, Germany) inside an acoustic recording booth (plastic bin lined with acoustic foam) for 1–24 h. Due to these recording conditions (e.g. ultrasonic microphone close to bat’s mouth, low gain, semi-anechoic chamber), the noise level was low (<0.5% amplitude) and we selected the first 17-30 contact calls that were between 10-99% amplitude (100% amplitude signals are clipped). Sounds were digitized with 16-bit resolution at a sampling rate of 250 kHz through an Avisoft Ultrasoundgate 116 to a laptop running the program Avisoft Recorder.
To create playbacks, we used Batsound Pro (Pettersson Elektronik AB, Sweden) to construct 19 playback sequences of contact calls spaced at random inter-call intervals varying from 1-3 s, and then adjusted intervals so that 30 calls would fit in a 60 s sequence. To remove sounds below 1 kHz, we used a high-pass Butterworth filter (filter order=2). Each 60 s playback sequence was presented and repeated four times in a trial through one of two Avisoft UltraSoundGate Player BL Pro units (henceforth “speakers”), which each contain an integrated digital-to-analogue converter, amplifier, and ultrasonic speaker (see www.ultrasoundgate.com for frequency response curves).
To equalize the intensity of the two speaker signals, we used the same Avisoft recording system to simultaneously monitor the amplitudes of sounds from the two speakers playing the same contact call repeatedly and placed equidistant to the microphone. We then adjusted the speaker gains until the signal amplitudes matched.
Trials took place in darkness and were filmed with an infrared spotlight (IRlamp6, Wildlife Engineering, USA) and a Sony Nightshot DCR-SR85 camera. In each trial, a subject bat was placed in a 76 x 76 cm plastic mesh experimental ‘maze’ with five perpendicular arms. The bat was placed in the arm farthest from the camera and the arms to the left and right each led towards a speaker simultaneously playing either calls from a control bat or a test bat. Other arms allowed the bats to move away from either speaker, by walking forwards or climbing upwards, which bats typically do as an escape response (Figure 1). These options allowed us to remove cases where bats were not motivated to move toward either speaker. Playback trials were of one of two types: past donor trials and kinship trials.
In past donor trials, a subject with a history of food sharing (N=12 females and 3 males) was presented with a choice between approaching contact calls from a frequent food donor (test bat) and an infrequent or non-donor (control bat). Test and control bat differed in their rates of food sharing with the subject, but were matched by age and kinship (Table 1).
In kinship trials, a subject (N=12 females, 19 males) was presented with a choice between contact calls from a close relative (test bat) and distant or non-relative (control bat). Test and control bats differed in their pairwise kinship with subject, but were closely matched wherever possible by age and past food sharing experience (Table 1).
Each trial lasted 10 min. If the bat did not make a choice within the first trial, the trial was discarded and a new trial began. Test stimuli were presented on the left or right randomly. The maze was rotated 90° clockwise between each trial. We presented the two trial types on a random schedule to remove order effects.