88.8 The Chewing Channel

[Bartie Spalitto]

Batsare highly gregarious animals, displaying a large spectrum of social systems with different organizational structures. One important factor shaping sociality is group stability. To maintain group cohesion and stability, bats often rely on vocal communication. The Honduran white bat, Ectophylla alba, exhibits an unusual social structure compared to other tent-roosting species. This small white-furred bat lives in perennial stable mixed-sex groups. Tent construction requires several individuals and, as the only tent roosting species so far, involves both sexes. The bats social system and ecology render this species an interesting candidate to study social behaviour and vocal communication. In our study, we investigated the social behaviour and vocalizations of E. alba in the tent by observing two stable groups, including pups, in the wild. We documented 16 different behaviours, among others play and fur chewing, a behaviour presumably used for scent-marking. Moreover, we found 10 distinct social call types in addition to echolocation calls, and for seven call types we were able to identify the corresponding broad behavioural context. Most of the social call types were affiliative, including two types of contact calls, maternal directive calls, pup isolation calls and a call type related to the fur-chewing behaviour. In sum, this study entails an ethogram and describes the social vocalizations of a tent-roosting phyllostomid bat, providing the basis for further in-depth studies about the sociality and vocal communication in E. alba.

Copyright: 2021 Fernandez et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: This research was funded by the German Academic Exchange Service (CS, SS). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

A: Picture of group 2 roosting in the tent during the day. Clearly visible is the yellow coloration of the ears of the adult individuals, whereas the pups ears are still almost white. B: Temporal colour marking of the fur to distinguish individual bats from group 1.

Since E. alba roosts cryptically during the day [7], we assumed that most social interactions would take place at dawn and dusk when bats are returning to or leaving the roost, or during the night when bats return to their tent [12]. To test this, we collected behavioural data over a total of 22 hours by monitoring group 1 for 3-hour periods spread equally over six days and nights (ranging from 9:47 to 8:03 of the next day). Each 3-hour period was recorded only once, each on a different date. The 22 hours were spread over several dates to avoid a sampling bias (e.g. randomly sampling a day/night which is deviating from average). Based on this 22-hour monitoring, we calculated time budgets of the different behaviours that we observed and determined when bats were most active. Consequently, we restricted our behavioural observations and sound recordings to recordings from dusk until dawn using ad libitum sampling [19]. During 25 nights, we recorded the bats behaviours in group 1 to establish an ethogram. Group 2 was observed occasionally and complemented our behavioural observations of group 1.

Each social behaviour type was considered either a state or an event [19]. States were defined as behaviours with a minimum duration of ten seconds, including behaviours during which the same motor actions were performed repeatedly (e.g., wing fluttering). Events were instantaneous and singular (e.g. wing stretching) and occurred during a state. To calculate time budgets for the different behavioural states, we used the data from our 22-hour monitoring. For each of the adult individuals in group 1 (n = 3) the 22 hours of observation time was split into two biological meaningful periods: the night-period, which included the time from leaving the tent at dusk to returning at dawn, and the day-period, which included the time from returning to the tent at dawn to leaving at dusk. Subsequently, we calculated the duration of each state (in seconds) and converted the durations into percentages to be able to compare these between individuals and day- and night-periods. The pup was not yet weaned and not foraging on his own; therefore, we decided not to split the 22-hour observation period for the pup.

Adult bats exhibited two main activity peaks, one before sunset and one around sunrise. This coincides with the time at which adult bats leave and return from foraging at night. The activity peaks were characterized by increased self-grooming, wing stretching and frequent position changing in the roost. The analysis of the 22-hour observing period (S1 Dataset) revealed that during the day-period, the adult bats spent the majority of time resting (male 1: 91.9%, male 2: 96.7%, female: 77.4% of the time). A short amount of the time they spent auto-grooming (male 1: 8.1%, male 2: 1.5%, female: 1.5% of the time) and, in the case of the female, nursing (21.2%). Male 2 was attacked by a mosquito during the 22-hour observation period and spend considerable time with distress behaviour (i.e. vigorously shaking the body and lashing out with half-opened wings, 1.74% of the time). The female nursed the pup three times before leaving for foraging at sunset and twice in the early morning after returning to the tent at sunrise. The longest nursing duration was observed at 06:00 when the female nursed her pup for one hour and 21 minutes.

After manoeuvring behind the back of a roosting group member, males were observed licking and gently biting the fur between the shoulder blades for a prolonged time (up to 13 minutes, Fig 2). While chewing the fur, males were sometimes simultaneously trembling their folded wings. After chewing, the bitten individual showed a visible patch of wet fur from the saliva biting individual. In most cases, males were chewing fur on the back of a female, but it was also observed that males chewed on the back of each other. In one occasion, a male that returned to the roost performed this behaviour on the pup who was roosting alone in the tent (for about 6 minutes, see S1 Video). The individual being bitten remained mostly still, sometimes started self-grooming, wing stretching and changing the position, with the fur chewing individual firmly clinging on. Eventually, the individual being bitten (if it was not the pup) also engaged in biting/licking another group member.

While being alone in the tent, the pup started to investigate a torn, dangling piece of the roost leaf (see S2 video, Fig 2). First, the pup started sniffing the leaf piece, and soon after used both thumbs/claws and wrists to grasp the leaf piece. Once grasped, the pup started chewing on the piece. The pup chewed on the leaf piece for a few seconds, stopped and started scanning. This behavioural sequence was repeated several times. Sometimes, the pup also started cleaning, wing stretching or moving around after chewing on the leaf piece. At one time, the pup was observed to inspect the modified midrib of the tent (i.e. the part of the leaf which is modified during tent construction to collapse the leaf next to the cut to achieve the typical shape of E. alba tents, see [8]). Afterwards, the pup turned back to the leaf piece and started chewing again, while simultaneously using the wrists and claws grasping and holding on to it. Chewing could get quite vigorous, and eventually, the pup started to bend the leaf piece to some extent.

E. alba produced ten distinct social call types in addition to echolocation calls, and for most call types the broad behavioural context in which they were uttered could be defined. The social call types SC9 and SC10 were not included in the statistical analysis because they were only recorded once with the microphone of the camcorder (S1 Fig in S1 File). Our visual classification of the remaining eight social call types was confirmed by the classification success of the cross-validated DFA (88.8% of all call types were classified correctly; Table 2, Fig 3B, further information about statistical fit: S1 Table in S1 File). The acoustic parameter that contributed most to the distinction of social call types was peak frequency in the centre of the call; followed by peak frequency at the start and the end of the call, whereas duration only played a minor role (S2 Table in S1 File). A conspicuous feature of the social vocalizations in E. alba is the suppression of the fundamental frequency and the lower harmonics in some of the call types (SC1, SC2; SC5-SC7, Table 3, Fig 3A).

A: Spectrograms depicting eight social call types (SC1-SC8) and two echolocation calls (EC) of E. alba for comparative reasons. The spectrograms correspond to the natural appearance of those call types; i.e., suppression of the lower harmonics in types SC2, SC5, SC6, SC7. Social call types SC9 and SC10 are only depicted in the supporting information (S1 Fig in S1 File) as they are not included in acoustic and statistical analyses. Information about acoustic parameter measurements is given in Table 3. The context in which the social call types were emitted is described in Table 4. Spectrograms were created using a 1024-point FFT and a Hamming window with 87.5% overlap. B: The spacing of eight social call types of E. alba in a two-dimensional signal space defined by the two most important discriminant functions. Each social call type is represented by a distinct symbol, small black symbols represent centroids (i.e., the geometric centre of all calls per type in the DFA signal space). Note that EC are not included in the DFA.

Furthermore, for seven out of ten social call types the broad behavioural context in which they were uttered was elucidated (Table 4). Three social call types were uttered in an affiliative context, namely SC2, SC3, SC5 (Fig 3). Two social call types were uttered in the context of mother-pup interactions, namely SC4 and SC6 (Fig 3). In agonistic context, two social call types were uttered (see supplements), namely SC9 and SC10. Echolocation calls were uttered during flight and during alert behaviour in the roost (i.e., scanning, see Table 1). Most social call types were uttered singly and monosyllabic; SC7 once consisted of multiple syllables (Table 4).

3a8082e126