Advertisement

The high-output singing displays of a lekking bat encode information on body size and individual identity

Original Article

Abstract

A growing body of research suggests that songs are an important part of the courtship behavior of many bat species, however there is little information on the basic characteristics of these vocalizations, or how they may function as a courtship signal. Lekking male lesser short-tailed bats (Mystacina tuberculata) appear to use vocal displays as a primary method for attracting mates, but it is unclear if these vocalizations constitute songs, and what characteristics females may use for mate selection. We recorded 16 lekking males and described the frequency and temporal properties of their vocalizations. We identified four notes (upsweeps, downsweeps, trills, and tones) that comprise courtship vocalizations, and males produced these notes either singly, or combined them linearly to form composite syllables. We classified 51 distinct syllable types (with an average of 29 types per male), with four (trills, upsweep-trills, trill-downsweeps, and upsweep-trill-downsweeps) comprising 69% of all syllables produced. The duration of trill-downsweeps scaled negatively with forearm length (a proxy for body size in bats), and all four main syllable types showed evidence of individuality. Based on the behavioral characteristics and contexts of these courtship vocalizations, we posit that this behavior constitutes singing. Furthermore, M. tuberculata potentially has one of the highest sustained song outputs yet described. Our results suggest the singing displays of M. tuberculata are signals that provide useful, honest cues of male characteristics and identity to females, and are as complex as the songs of many passerines.

Significance statement

Male courtship displays are predicted to honestly advertise aspects of male traits. Like many birds, some bat species produce songs to attract mates, but the characteristics of these songs are not well understood within the context of sexual selection. We demonstrate that the courtship vocalizations of Mystacina tuberculata – a species that likely relies on singing as its primary method of mate attraction – are a complex set of signals. Males have large syllable repertoires, encode four of their most-common syllables with individual signatures, and have one of the highest recorded song outputs for either birds or bats. Moreover, the length of one of the most commonly used syllable types is inversely related to male size, providing females the opportunity to appraise male size by auditory cues alone. Our work is part of a growing body of research demonstrating singing behavior in bats.

Keywords

Allometry Chiroptera Courtship Lek breeding Sexual selection Song 

Introduction

Songs are generally defined as long, complex vocalizations produced by males during the breeding season, and serve the dual purpose of attracting mates and repelling rivals (Catchpole and Slater 2003; Collins 2004). Singing is typically associated with passerines, and birdsong has been studied intensively since the 1950s (Marler 2004). By contrast, singing is relatively rare in mammals (Smotherman et al. 2016) and has thus received comparatively less attention. However, increasing research suggests that singing is an important behavior for some bats.

Due to their nocturnal behavior, bats rely on acoustic signals for both orientation (Fenton 2013) and social purposes (e.g. Wilkinson and Boughman 1998; Pfalzer and Kusch 2003). The importance of echolocation may have predisposed bats for song, as they possess finer control over their vocalizations than other mammals (Smotherman 2007; Pollak and Casseday 2012). To date, only 20 out of over 1300 bat species are known to produce songs or singing-like behaviors, however this number will likely increase with further research (Smotherman et al. 2016). The few studies that have examined bat songs have shown that many species produce songs as complex as those of birds (e.g. Behr and von Helversen 2004; Jahelková et al. 2008; Bohn et al. 2009; Schmidt 2013; Smarsh and Smotherman 2015). For example, the sac-winged bat (Saccopteryx bilineata) is a harem-polygynous species where males produce daytime courtship songs directly at females in their harems, and augment these songs with wing-flapping and olfactory displays (Behr and von Helversen 2004). Song traits, such as the output and frequency of male songs, have been linked with reproductive success (Behr et al. 2006; Voigt et al. 2008), similar to results found songbirds (e.g. Byers 2006; Poesel et al. 2006).

The lesser short-tailed bat (Mystacina tuberculata) is one of only two terrestrial mammals endemic to New Zealand, and one of only two lek-breeding bat species described to date (Bradbury 1977; Toth and Parsons 2013; Toth et al. 2015b). Further, it is one of the 20 bat species known to produce song-like vocalizations. During the breeding season (approximately January to May), some males occupy small tree cavities termed “singing roosts” from which they vocalize for large portions of the night, presumably to attract passing females. These singing roosts are aggregated in the areas surrounding large communal roost trees used by the population as day roosts, likely to increase female encounter rates, and are on average 60 m apart (although they can be as little as 1.5 m; Toth et al. 2015b). Females visit singing males within their roosts for mating purposes, leaving before dawn (Toth et al. 2015b). Both females and singing roosts are defended by resident males from sneaker males (Toth et al. 2018), and males even visit their singing roosts outside of the breeding season (but do not sing; Toth et al. 2015a), as observed in other lekking species (Höglund and Alatalo 1995). While some singing roosts are “solitary” (i.e. occupied by a single male), others are ‘shared’ by two to five males, and are known as “timeshares” (Toth et al. 2015b). Timeshare males occupy their singing roost sequentially (i.e. they do not overlap temporally) each night, and show high fidelity to their roosts on a night-to-night basis. Further, some of these arrangements can last over multiple seasons (Toth et al. 2018).

Although the vocal displays of this species have previously been referred to as “songs” in the literature (Carter and Riskin 2006; Toth et al. 2018), and the courtship structures are referred to as “singing roosts”, it is unknown if male courtship vocalizations meet the general criteria usually required to distinguish calls from songs (e.g. complex vocalizations with specific behavioral and temporal characteristics; Catchpole and Slater 2003). Regardless, M. tuberculata makes a convenient study subject to investigate courtship vocalization behavior in bats, as both solitary and timeshare males show high fidelity to their singing roosts (allowing individuals to be consistently identified and recorded through passive methods; Toth et al. 2015b), and may vocalize for upwards of 8 h per night. This system also allows for useful comparisons to songbird systems. For example, the short inter-roost distances and long-term use of singing roosts may facilitate the sharing of vocal traits between neighboring (and timeshare) males, as seen in many bird species (Marler and Slabbekoorn 2004). Further, as a lek-breeding species, male vocalizations may honestly reflect male traits. Lek theory predicts that female selection of mates is based solely on honest male advertisements, as female choice is not confounded by territory quality, resource ownership, or paternal care (Höglund and Alatalo 1995). This is also true for M. tuberculata, as males do not provide nuptial gifts to visiting females, and females do not use the singing roosts as day roosts (Toth et al. 2015b). Given that females appear to select males (at least initially) using only vocalizations, we may expect sexual selection to have shaped these courtship signals to be individually distinctive and carry information – such as morphology – on the singer.

Previous research on M. tuberculata has shown that body size predicts some courtship behavior and mating success; smaller males have higher lek attendance (i.e. spend more time in their singing roost), higher duty cycles (i.e. spend a greater amount of time vocalizing within a given period), and produce more offspring than larger males (Toth et al. 2018). Furthermore, males of disparate sizes appear to pursue different lekking strategies, as solitary males tend to be smaller than timeshare males (Toth et al. 2018). Given the disparity in size, timesharing may represent a cooperative strategy by larger males to cumulatively increase lek attendance and song output (see Toth et al. 2018). Females in many species have been observed to use male courtship songs as cues for body size (e.g. Ryan 1980; Gray 1997; Galeotti et al. 2005). Frequency characteristics of vocalizations can carry accurate information on the size of the male producing it (e.g. Linhart and Fuchs 2015), as the size of the vocal organs generally scale with body size (known as “size-pitch allometry”; Bradbury and Vehrencamp 1998). However, few studies have demonstrated female preference for males of certain sizes based on courtship vocalization traits in mammals (e.g. Charlton et al. 2007, 2012).

The purpose of this paper is to present an overview of the courtship vocalizations of male M. tuberculata, including an analysis of the frequency and temporal characteristics of the most-common syllable types produced by males, and to determine if male vocalizations meet the criteria of songs. We hypothesize that as the primary mate-attraction method by lekking M. tuberculata, characteristics of male vocalizations and courtship behavior function as an honest indicator of male traits and are individually distinctive. Based on the findings of Toth et al. (2018), we used body size as a predictor and hypothesize that syllable characteristics, specifically duration (as smaller males have been found to have higher duty cycles), display allometry. Further, we hypothesize that neighboring males may share syllable types, as seen in some neighboring songbirds. Specifically, males may share more syllables within timeshare roosts than between them, and neighboring solitary males may share more syllables than those further apart.

Methods

Field methods

We studied Mystacina tuberculata in the Pikiariki Ecological Area of Pureora Forest Park (38°26′S, 175°39′E), central North Island, New Zealand, between November and May 2011–2012 and 2012–2013, and between February and April in 2013–2014. As part of a larger study (see Toth et al. 2015b), we captured 712 individuals in this population (either using mist nets or harp traps) and marked them with passive integrated transponder (PIT) tags injected subcutaneously between the shoulder blades for individual identification. After marking, we weighed individuals with a Pesola scale (0.1 g precision) and measured their forearms with calipers as a proxy for body size. It was not possible to record data blind because our study involved focal animals in the field.

Audio recording

We recorded the courtship vocalizations of 16 males at 12 singing roosts between February 19 and April 5, 2014. All males were recorded between 1900 and 2130 h on nights with relatively clear weather (i.e. no rain or strong winds). We recorded male vocalizations using a D980 bat detector (Pettersson Eletronik AB, Uppsala, Sweden; frequency response 10–200 kHz). Sound was passed through the D980 directly to a Sound Devices 722 digital recorder (Sound Devices, Reedsburg, USA) where it was digitized at sampling rate of 192 kHz with 24-bit precision. We recorded within 10 m of the roost tree, generally just after the male arrived at dusk. However, with timeshare roosts other resident males had to be recorded as they arrived later in the evening. In total, six solitary males and ten timeshare males (from six roosts, including two roosts with recordings of multiple males) were recorded. The identities of recorded males were determined using a Biomark HPR Plus automatic PIT-tag reader (Biomark, Idaho, USA) mounted on the entrances of the singing roosts, and we only recorded the vocalizations of PIT-tagged males. We recorded males for 10 min of what was deemed to be continuous vocalizing (i.e. no significant pauses) to ensure an adequate sample of the male’s repertoire. Additional detailed information on the study site, population size, and singing roost selection can be found in Toth et al. (2015b).

Classifying vocalizations

We used RavenPro 1.4 (Cornell Laboratory of Ornithology, Ithaca, USA) to perform the acoustic analyses of male vocalizations. We generated spectrograms using a 1024-sample discrete Fourier transformation (Hanning window) with 95% overlap, 188 Hz frequency resolution, and 0.135 ms time resolution. Due to the high song output produced by M. tuberculata, classifications were only made for every other minute in each 10-min recording (first minute, third minute, fifth minute, etc.). When classifying, we first used visual discrimination of the frequency-time contour to define discrete syllables. Generally, these were vocalizations separated from subsequent vocalizations by at least 20 ms of silence (gaps shorter than this were considered continuous syllables). The syllables produced by M. tuberculata can be broadly classified into three categories: trills-based syllables, tone-based syllables, and trill-tone combinations. However, considerable structural variation exists within these three categories that is lost with such broad classifications, and so we used a fine-scale classification scheme to further differentiate syllables. We categorized syllables based on the presence and linear combination of four basic notes: tones, trills, upsweeps, and downsweeps (Fig. 1). A trill was defined as a note that rapidly fluctuated in frequency (although not always directly connected; Bohn et al. 2009; Fig. 1a), while tones – although they may gradually increase or decrease in frequency – did not fluctuate (Fig. 1b). Upsweeps and downsweeps were defined as notes that rapidly increased or decreased in frequency, respectively (Fig. 1c, d). Thus, syllables could be composed of a single note or a composite of several notes combined in a linear fashion (e.g. an upsweep immediately followed by a trill was classified as an “upsweep-trill”). While relatively rare, some syllables contained two instances of a note, and this was reflected in their classification. For example, a tone followed by an upsweep, tone, and downsweep would be classified as a tone-upsweep-tone-downsweep.
Fig. 1

Sample spectrograms of the four basic notes that comprise M. tuberculata syllables: a trills, b tones, c upsweeps, and d downsweeps. Syllables may be comprised of several notes in combination, or as solitary notes

Syllables were also scored on their quality: syllables with complete structures (i.e. the entire syllable was clearly visible) were given a score of 1, syllables with incomplete structures (e.g. sometimes the high pitch components were faint on spectrograms) were given a score of 2, and those with poor structures (i.e. only duration could be identified) were given a score of 3. This allowed syllables to be sorted based on the requirements of later analyses. RavenPro generated five frequency measurements from syllables: lowest frequency (LF; the lower bound of the selection box), highest frequency (HF; the higher bound of the selection box), center frequency (CF; the frequency that divided the syllable into two intervals of equal energy), peak frequency (PF; the frequency at which maximum power occurs), and syllable duration (DU).

Data Availability

The datasets analyzed from the current study are available from the corresponding author on reasonable request.

Statistical analyses

All statistical analyses were performed using the program JMP version 13.0 (SAS Institute Inc. 2016) except for Kaiser-Meyer-Olkin scores, which were generated using the package “psych” (Revelle 2018) available for the statistical program R version 3.2.3 (R Development Core Team 2013). Mean ± SD values are reported.

Repertoire size

To determine if number of recordings and the rate of subsampling were sufficient to reveal the population-level syllable repertoire we used the exponential function developed by Wildenthal (1965), modified by Davidson and Wilkinson (2002) (see the Online Resources).

It was not overtly clear if syllables were arranged into longer bouts. As such, we performed a bout analysis to distinguish inter-syllable intervals from vocal-bouts intervals (Sibly et al. 1990). We fit a two-process model on the log-frequency of inter-syllable intervals against interval length, and the resulting bout criterion indicated bouts were denoted by gaps of at least 1.42 s (which comprised 0.4% of all intervals). This gave a total of 101 vocal bouts across all sampled males (7.3 ± 8.7 bouts/male). However, by this criterion, four males had only one continuous bout, while one male had 36 (Table 1). Given this large discrepancy, we continue to focus on syllables as the main unit of analysis.
Table 1

Overview of the vocal output of 16 M. tuberculata males (n = 6 solitary, n = 10 timeshare) in five minutes of recording

Roost type/numbera

Male IDb

Total number of syllables

Number of syllable types

Number of vocal boutsc

Solitary

    

 5

356,606,506

1776

29

1

 6

356,602,237

1820

35

10

 8

356,559,620

2339

29

1

 33

190,967,474

1666

28

9

 52

190,962,945

1827

30

1

 61

356,753,899

1066

32

11

 

Mean ± SD

1749 ± 408.3

30.5 ± 2.6

5.5 ± 5.0

Timeshare

 12

356,559,297

726

25

36

 12

167,879,784

1357

24

12

 12

356,799,541

1464

27

6

 24

190,976,138

2713

33

1

 35

356,750,114

1266

33

1

 36

356,549,813

1754

31

2

 36

356,551,179

1532

24

5

 36

167,788,152

1156

31

4

 48

190,977,468

1579

25

6

 58

356,766,438

1660

21

11

 

Mean ± SD

1520.7 ± 511.9

27.4 ± 4.3

8.4 ± 10.4

 

Overall mean ± SD

1606.3 ± 475.2

28.6 ± 4.0

7.3 ± 8.7

aTimeshare males with the same roost number are roostmates

bMales are denoted by their PIT-tag number

cVocal bouts delineated by gaps of >1.42 s, as indicated by a bout analysis

Syllable sharing

We determined the level of syllable-type sharing between neighboring solitary males, and males within timeshare roosts. The proportion of shared syllables between males was calculated using the formula developed by McGregor and Krebs (1982):
$$ P=\frac{2{N}_S}{\left({R}_1+{R}_2\right)} $$
where Ns denotes the number of shared syllables between the two males, and R1 and R2 are the syllable repertoire sizes of the two males. We calculated shared proportions for each combination of solitary males (resulting in 15 dyads), between timeshare males in the same roost (two roosts with three males each, resulting in six dyads), and between timeshare males in separate roosts (resulting in nine dyads). We also regressed the proportion of shared syllables between solitary males against the distance between their singing roosts to determine if fewer syllables are shared between distant neighbors (e.g. McGregor and Krebs 1982). For details on the spacing between singing roosts, see Toth et al. (2015b).

Frequency characteristics

We examined the basic frequency characteristics of the ten most-common syllables produced by males, as these syllables represent a range of combinations for the four main notes (Fig. 2). This allowed us to 1) examine how the addition or removal of notes affects the frequency parameters of a syllable, and 2) confirm visual classification of syllables, as some appear structurally similar. We extracted PF from all high-quality syllables (i.e. those scoring a 1) of the ten syllables and performed a categorical mixed model with PF as the response variable and syllable type as a predictor, as well as male identity and an interaction between male identity and syllable type as random effects. Further, we performed pairwise comparisons of least squares means generated from the model using Tukey-HSD tests to determine between-syllable frequency differences. We also compared the LF and HF measures of upsweep-trills, upsweep-trill-downsweeps, tone-upsweep-trills, and tone-upsweep-trill-downsweeps, as these represent syllables that are structurally similar, except for the addition of a short tonal note at the beginning of the latter two syllables (see Fig. 2). These analyses consisted of two categorical mixed models with LF or HF as the response variable, syllable type as a predictor, as well as male identity and an interaction between male identity and syllable type as random effects. LF was log-transformed to meet the assumptions of normality. We then used post-hoc pairwise Tukey-HSD tests to compare differences between the syllable types.
Fig. 2

Sample oscillograms (time x voltage) and spectrograms (frequency x time) of M. tuberculata vocalizations, showing the ten most-common syllables (a – j) and an example of 1 s of continuous singing (k). a Trill, b trill-downsweep, c upsweep-trill-downsweep, d upsweep-trill, e tone-trill, f tone-upsweep-trill-downsweep, g trill-tone, h tone-upsweep-trill, i upsweep-tone, j tone. k This section of singing contains upsweep-trills, upsweep-trill-downsweeps, and upsweep-tone-downsweeps. Spectrograms were generated using a 1024-sample discrete Fourier transformation (Hanning window) with 95% overlap, frequency resolution 188 Hz, time resolution 0.135 ms

Allometry

We investigated if male body size influenced syllable characteristics. These analyses were performed on the four most-common syllable types (i.e. those that comprised >10% of all observed syllable utterances), as it is likely that these would be most important for mate attraction. We randomly selected 20 high-quality syllables from each male for each of the four major syllable types and used a principal component analysis (PCA) to reduce the number of frequency variables, as these parameters were largely correlated. Trill-downsweeps and upsweep-trill-downsweeps included syllables from only 15 males each, as two males (one for either syllable) did not produce sufficient high-quality syllables of those types. Kaiser-Meyer-Olkin tests (all indices >0.6) and Bartlett’s test (P < 0.0001) indicated that the data were appropriate for factor analysis. We interpreted the resulting principal components (PCs) on the presence of variables with loadings >0.7. We retained two components subjected to varimax rotation of eigenvalue to ensure at least 80% of the variance was accounted for by the components, and to increase the interpretability of results. We used three linear mixed-effects models to test whether male size influenced inter-male differences in syllable characteristics. The models included syllable PCs and duration as the dependent variables, forearm length, syllable type, and an interaction between forearm and syllable type as fixed variables, with male identity and an interaction between male identity and syllable type as random effects.

Potential for individual identity coding

We used both univariate and multivariate methods to determine if male syllables contain information of individual identity. First, we compared the variability of syllable parameters (i.e. frequency and temporal measures) within individuals and between individuals to determine if syllables have the potential to encode individual identity (e.g. Vignal et al. 2004, 2008). Using the same syllables from the allometry analysis (above), we calculated the within (CVi) and between-individual (CVb) coefficients of variation for the frequency and temporal parameters of the four main syllable types using the formulas:
$$ {CV}_b=\frac{SD}{mean}\times 100\kern2.75em {CV}_i=\frac{SD}{mean}\times 100\times \left(1+\frac{1}{4n}\right) $$
where CVb is calculated with the overall mean and standard deviation (SD) of each parameter, and CVi is calculated with the SD and mean of each parameter per individual, with n as the number of exemplars per individual (see Sokal and Rohlf 1995). We used the ratio of CVb to the mean of CVi (mCVi) to give a measure of individuality, with values >1.0 indicating a signal’s between-individual variability is greater than its within-individual variability, and may therefore be used for individual identification.

We used quadratic discriminant function analyses (DFA) with leave-one-out cross validation to determine if each of the four syllables could be assigned to individuals. To do this, we performed a PCA for the frequency measures of each of the four main syllable types, using the same methods described above. We performed two sets of DFAs: one set in which all males were included, and a second set with smaller, “biologically relevant” groupings of males. Given that it is improbable to hear 16 males concurrently from any location within the aggregation due to the spacing of singing roosts (Toth et al. 2015b), the first DFA likely exaggerates the discrimination problem faced by individuals in the lek (Beecher 1989). Therefore, to determine an appropriate group size for the second set of DFAs, we estimated how many males would be audible from random locations within the lek. We used the “Minimum Bounding Geometry” tool in ArcMap v.10.4.1 (ESRI 2012) to delineate a convex polygon around 59 singing roosts in the “main” aggregation (Toth et al. 2015b), then generated 100 random points within the resulting 65 ha polygon using the “Create Random Points” tool (see Online Resources). While the hearing sensitivity of M. tuberculata is unknown, we created buffer polygons with a radius of 75 m around each point to estimate observer listening areas, as this approximates the absolute maximum distance male songs are audible to human observers (Toth et al. 2015b). We then used the “Spatial Joins” tool to determine how many singing roosts fell within each buffer polygon. This resulted in 47 polygons that intersected ≥1 singing roost, with a median of 3 singing roosts per polygon (average: 2.7 ± 1.5, range: 1–6 roosts/polygon). Thus, for the second set of DFAs we created five random groupings of three males (without replacement) for comparisons. We used syllable PCs and DU as predictor variables in the DFAs, and conducted directed binomial tests to evaluate DFA performance against random classifications (Mundry and Sommer 2007). We used directed tests instead of one-tailed tests to account for the possibility of an unexpected outcome (Rice and Gaines 1994).

Results

Repertoire size

We quantified 25,701 vocalizations uttered by 16 lekking males (1606.4 ± 475.1 vocalizations/male, range: 726–2713), comprised of 51 distinct syllables (28.6 ± 3.9 syllables/male, range: 21–35; Table 1). The four most-common syllable types were trills (6956 syllables, 27.1% of total), trill-downsweeps (3791, 14.8%), upsweep-trill-downsweeps (3517, 13.7%), and upsweep-trills (3460, 13.5%; Fig. 2). These syllables ranged in frequency from approximately 8.6 to 29.6 kHz, and in length from 46.1 to 62.6 ms, on average (Table 2). All other syllables individually comprised <4% of total syllables produced, including the remaining three basic notes (for a list of all syllable types recorded, please see the Online Resource). Overall, syllables ranged in duration from 3 (single upsweeps and downsweeps) to 409 ms (an upsweep-trill-downsweep), averaging 53 ms (± 21 ms). Generally, syllables contained a combination of two or three notes, but some contained four or five (although these were rare; see Online Resources).
Table 2

Parameters measured from the four most-common syllable types produced by M. tuberculata, the between-individual and mean within-individual coefficients of variation (CVb and mCVi), the results of PCAs that combined frequency characteristics, and success rate of two DFAs to assign syllables to the males who produced them using PCA and DU measures

 

Syllable parameters (Mean ± SD)

mCVi

CVb

CVb/mCVib

PC1c

PC2

DFA classification success – all malesd

Mean DFA classification success – three-male groupse

Trills (n = 320)a

      

30%

68%

 LF (kHz)

12.16 ± 4.85

31.71

39.86

1.26

0.32

0.92

  

 HF (kHz)

26.26 ± 7.41

21.34

28.21

1.32

0.92

0.3

  

 CF (kHz)

17.62 ± 5.22

25.13

29.63

1.18

0.74

0.62

  

 PF (kHz)

17.21 ± 5.61

29.04

32.59

1.12

0.64

0.67

  

 % Variance explained

    

47.6%

44.4%

  

 DU (ms)

46.08 ± 20.26

39.78

43.98

1.11

    

Trill-downsweeps (n = 300)

      

33%

73%

 LF (kHz)

10.38 ± 3.95

26.45

38.05

1.44

0.32

0.91

  

 HF (kHz)

29.69 ± 5.67

13.78

19.1

1.39

0.59

0.58

  

 CF (kHz)

18.84 ± 5.17

22.9

27.46

1.2

0.88

0.4

  

 PF (kHz)

18.09 ± 5.92

27.74

32.73

1.18

0.91

0.31

  

 % Variance explained

    

51.2%

35.8%

  

 DU (ms)

62.22 ± 23.22

24.47

37.32

1.53

    

Upsweep-trill-downsweeps (n = 300)

      

28%

66%

 LF (kHz)

8.63 ± 1.74

17.16

20.17

1.18

0.19

0.97

  

 HF (kHz)

31.1 ± 4.92

13.31

15.81

1.19

0.65

0.32

  

 CF (kHz)

19.0 ± 4.45

21.1

23.4

1.11

0.93

0.15

  

 PF (kHz)

17.67 ± 5.57

28.82

31.51

1.09

0.88

0.13

  

 % Variance explained

    

52.9%

27%

  

 DU (ms)

62.66 ± 20.97

21.45

33.46

1.56

    

Upsweep-trills (n = 320)

      

29%

72%

 LF (kHz)

9.54 ± 2.32

19.11

24.32

1.27

0.19

0.97

  

 HF (kHz)

28.88 ± 5.64

15.77

19.53

1.24

0.65

0.32

  

 CF (kHz)

18.4 ± 4.52

21.22

24.57

1.16

0.93

0.15

  

 PF (kHz)

17.06 ± 5.46

29.01

31.99

1.1

0.88

0.13

  

 % Variance explained

    

52.9%

27%

  

 DU (ms)

57.98 ± 23.06

23.97

34.24

1.43

    

aSample sizes refer to total number of syllables used to calculate averages, CVs, PCs, and perform DFAs; 20 high-quality syllables were chosen from either 15 or 16 males (two males were not included, one for trill-downsweeps and one for upsweep-trill-downsweeps, as they did not produce enough high-quality syllables of those types)

bCVb/mean CVi ratios of >1.0 indicate between-individual variation is greater than within-individual variation for a parameter

cLoadings >0.7 are highlighted in bold

dChance of success via random classification is approximately 6%

eChance of success via random classification is 33%

Syllable sharing

There was no difference in the proportion of syllables shared between solitary males (0.74 ± 0.04, range: 0.67–0.83), between individuals in the same timeshare roost (0.76 ± 0.02, range: 0.73–0.79), or between individuals in different timeshare roosts (0.79 ± 0.07, range: 0.69–0.9; one-way ANOVA, F2,29 = 2.28, P = 0.12). Solitary male singing roosts were on average 586 m (± 582.3, range: 99–1444 m) apart, and there was no relationship between syllable sharing and distance between solitary roosts (R2 = 5.35 × 10−5, F1,15 = 0.0007, P = 0.98).

Frequency characteristics

We compared PF of the ten most-common syllable types produced by males. This analysis included trills (n = 1981), trill-downsweeps (n = 1637), upsweep-trill-downsweeps (n = 1285), upsweep-trills (n = 871), tones (n = 454), upsweep-tones (n = 371), tone-trills (n = 368), trill-tones (n = 316), tone-upsweep-trill-downsweeps (n = 214), and tone-upsweep-trills (n = 85). Syllable type was a significant predictor of PF (adj. R2 = 0.43, F9 = 26.57, dfd = 110.7, P < 0.0001), with both male identity (P = 0.017) and the interaction between male identity*syllable type (P < 0.0001) as significant sources of variation. Post-hoc pairwise Tukey-HSD tests indicate that syllables containing tones tended to have lower peak frequencies than those that did not contain them, and syllables that have downsweeps tend to have higher frequencies that those without (Fig. 3). Interestingly, even a simple reversal of notes had substantial effects on frequency, as PF was significantly different between trill-tones and tone-trills (Fig. 3). We performed a further analysis comparing the LF and HF of upsweep-trills, upsweep-trill-downsweeps, tone-upsweep-trills, and tone-upsweep-trill-downsweeps. Syllable type was a significant predictor of HF (adj. R2 = 0.4, F3 = 3.69, dfd = 26.52, P = 0.024) and log-LF (adj. R2 = 0.41, F3 = 16.43, dfd = 26.43, P < 0.0001), and pairwise Tukey-HSD tests indicated significant differences for log-LF between those syllables containing the tonal modifiers and those without (Table 3), but not for HF (all P-values >0.05). Male identity and the interaction between male identity*syllable type were significant sources of variation in both models (all P-values <0.05).
Fig. 3

Box plots of peak frequency across ten of the most-common syllables produced by male M. tuberculata (x-axis labels correspond to the notes that comprise each syllable; T = trill, U = upsweep, D = downsweep, O = tone). Overhead letters correspond to significance groupings provided by pairwise Tukey-HSD comparisons using Holm-Bonferroni correction. Syllables containing downsweeps tend to have higher peak frequencies, while syllables containing tones tend to have lower peak frequencies. Box plots show 25% and 75% quartiles (boxes), medians (lines in the boxes), and outermost values within the range of 1.5 times the respective quartiles (whiskers)

Table 3

Comparisons of LF measures of four structurally similar syllable types: upsweep-trills, upsweep-trill-downsweeps, tone-upsweep-trills, and tone-upsweep-trill-downsweeps. a means and standard deviations of LF for the four syllable types. b Significance levels from pairwise Tukey-HSD tests on least square means

 

Upsweep-trills (n = 871)

Upsweep-trill-downsweeps (n = 1285)

Tone-upsweep-trills (n = 85)

Tone-upsweep-trill-downsweeps (n = 214)

A)

    

 LF Mean ± SD (kHz)

9.55 ± 2.6

8.72 ± 2.09

7.34 ± 1.68

7.21 ± 1.28

B)

    

 Upsweep-trills

    

 Upsweep-trill-downsweeps

*

   

 Tone-upsweep-trills

***

*

  

 Tone-upsweep-trill-downsweeps

***

*

 

*P < 0.05

***P < 0.001

Allometry

The average forearm length of recorded males was 41.9 ± 0.8 mm (range: 40.5–43.7 mm). Syllable repertoire size was not correlated with forearm length (R2 = 0.01, F1,15 = 1.52, P = 0.24). The two PCs generated from the four main syllable types together accounted for 86% of the variance in frequency characteristics, with PC1 (described by HF, CF, and PF) and PC2 (described by LF) explaining 53.5% and 32.8% of the variation, respectively (Table 4). There was no relationship between any predictor and either of the two frequency PCs (Table 5), however duration was significantly related to syllable type, and pairwise Tukey-HSD tests indicated that the duration of trills was significantly lower than the other three syllables (all P values <0.05). Further, while there was no overall effect of forearm length on syllable duration (Table 5), there was a significant negative interaction between the duration of trill-downsweeps and forearm length (β = −0.0044, t = −2.1, P = 0.042). Syllable type was a significant predictor of both PCs and DU, and male identity and the interaction of male identity*syllable type were significant sources of variation in every model (Table 5).
Table 4

Results of PCA combining the frequency characteristics of the four main syllable types produced by M. tuberculata

 

PC1a

PC2

LF

0.22

0.95

HF

0.9

0.084

CF

0.85

0.41

PF

0.75

0.49

% Variance explained

53.5%

32.8%

aLoadings >0.7 are highlighted in bold

Table 5

Mixed effects models estimating predictors of frequency and temporal characteristics, including principal components (used to combine frequency measurements) and duration of the four most common syllables produced by M. tuberculata

 

Fixed effects

Random effects

Response

Predictor

Adj R2

df

dfd

F

P a

Effect

% of Total variation

Wald’s P

PC1

 

0.41

       

Forearm

 

1

14.1

0.008

0.078

Male ID

21.8

0.026

Syllable type

 

3

39.93

17.38

< 0.0001

Male ID

*Syllable type

11.28

0.0006

Forearm

*Syllable type

 

3

40.14

1.5

0.23

   

PC2

 

0.45

       

Forearm

 

1

14.55

0.16

0.69

Male ID

15.41

0.047

Syllable type

 

3

40.33

28.61

< 0.0001

Male ID

*Syllable type

15.38

0.0003

Forearm

*Syllable type

 

3

40.62

0.68

0.57

   

Duration

 

0.49

       

Forearm

 

1

14.35

4.12

0.061

Male ID

25.44

0.024

Syllable type

 

3

40.2

13.5

< 0.0001

Male ID

*Syllable type

13.79

0.0002

Forearm

*Syllable type

 

3

40.39

1.7

0.18

   

aSignificant P values highlighted in bold

Potential for individual identity coding

Male syllables encoded individual information. All measured parameters for the four main syllable types displayed a CVb/mCVi ratio > 1.0, indicating that within-individual variation was lower than between-individual variation (Table 2). The two PCs for each of the four main syllable types together accounted for 92% of the variance in trills, 87% of the variance in trill-downsweeps, 80% of the variance in upsweep-trill-downsweeps, and 84% of the variance in upsweep-trills (Table 4). A DFA including all males could correctly assign syllables to the males that produced them between 28 and 33% of the time (Table 2). When group sizes were kept to biologically relevant sizes of three individuals, DFAs could correctly classify trills on average 68% of the time (range: 55–87%), trill-downsweeps 73% of the time (range: 52–87%), upsweep-trills 66% of the time (range: 52–85%), and upsweep-trill downsweeps 72% of the time (range: 57–88%) (Table 2). All classifications were significantly better than the success rate expected by random chance (directed binomial tests, all Pdir values <0.05).

Discussion

Our results indicate that the courtship vocalizations of male M. tuberculata are a complex set of signals that have the potential to function as an honest indicator of male size and identity. Males possess large syllable repertoires, have syllable traits that display greater variation between individuals than within individuals and can be correctly assigned to individuals greater than expected by chance, and display negative allometry between body size and the length of the second most-common syllable type. Given these features, the courtship vocalizations of M. tuberculata appear to be useful traits for female selection of males.

Do the courtship vocalizations of male M. tuberculata constitute songs, or are they merely song-like calls? There is no clear definition that separates “calls” (generally considered to be shorter, monosyllabic vocalizations) from “songs” (generally longer, complex vocalizations), and the two are often considered to lie on two ends of a spectrum (Catchpole and Slater 2003). The courtship behavior of M. tuberculata shares some qualities with species whose vocalizations would be considered calls, such as anurans, as males produce very short, discrete syllables at high rates for extended periods time (Wells 2010). However, call-like qualities alone do not preclude singing. For example, many songbird species (e.g. wood warblers, parulids) possess only one or a limited number of song types, with complexity introduced through variation in frequency, temporal, and/or structural components (e.g. Morton and Young 1986; Spector 1992; Christie et al. 2004; Mennill and Otter 2007; Rivera-Gutierrez et al. 2010). Stereotyped syllables have also been considered parts of songs in other bat species (e.g. Bohn et al. 2016). Behaviorally, M. tuberculata courtship vocalizations appear to adhere to the general indices used by ornithologists to distinguish songs from calls: they are only produced by males, and only during the breeding season (CAT, pers. obs.). Vocalizations appear to be the primary method of mate attraction for lekking males (although scent-marking with urine has also been described; Toth et al. 2015b), and these males possess large, individually distinctive repertoires, like many songbirds (e.g. Molles and Vehrencamp 1999). As such, we consider the vocalizations of lekking M. tuberculata to be singing, even if they do share call-like qualities.

Mystacina tuberculata syllables can be described with four notes – tones, trills, upsweeps, and downsweeps – that are either produced singly, or combined linearly with other notes to form complex syllables. Individual M. tuberculata in our study possessed an average of 29 syllable types, with a population-level repertoire of 51, which is larger than many songbirds (see Beecher and Brenowitz 2005). There was no evidence for repertoire sharing between either neighboring solitary males, or between timeshare males within the same timeshare roost. This differs from findings in several songbird species which have been shown to learn the songs of territorial neighbors after hatching, or learn the songs of their neighbors following territory settlement in their first year (e.g. Nordby et al. 1999, 2000). How M. tuberculata acquire syllables is beyond the scope of this study, but vocal learning has been shown in the territorial songs of S. bilineata; young pups produce “babbling” bouts that serve as song precursors (Knörnschild et al. 2006) that eventually grow to resemble the songs of their adult tutors (Knörnschild et al. 2010). Mystacina tuberculata pups are reared within maternity roosts, and thus would likely not be exposed to male songs until after they are volant (approximately four weeks of age; Lloyd 2001), at which point they could theoretically learn from any males in the lek. Further, PIT-tagged juvenile males have been recorded outside of singing roosts (CAT, unpublished data). Syllables may also be innate, and not learned at all. Another possibility to explain the lack of syllable sharing is that our method of classification was too fine, and artificially inflated the repertoire sizes of singing males. For example, a trill-tone may be interchangeable with a tone-trill from a bat’s perspective, but were considered as different syllables by our classification criteria. However, we found that tone-trills and trill-tones had significantly different peak frequencies, which suggests some level of differentiation. Furthermore, even small structural changes appeared to impact the frequency components of syllables, as upsweep-trills and upsweep-trill-downsweeps that contained short tonal components had significantly lower frequencies (as measured by their absolute lowest values) than those without. Future work investigating if structural changes to syllables translate to functional differences (e.g. mate attraction vs. territorial defense purposes), perhaps with playback experiments, would be useful.

While it is unclear if all the described syllables have functional significance, four syllables – trills, upsweep-trills, trill-downsweeps, and upsweep-trill-downsweeps – appear to be important, as they were the only syllables that individually comprised >10% of the total syllables produced, and they all displayed the same basic structure between males (see Online Resources). Trills have been shown to be important for mate selection in several species of songbirds, likely because they require high levels of vocal performance to produce (see Podos et al. 2004). Additionally, trills are the most-frequent call type in the courtship songs of S. bilineata (Behr and von Helversen 2004), are common in the songs of Mexican free-tailed bats (Tadarida brasiliensis; Bohn et al. 2009), and have been recorded in the social calls of multiple European species (Pfalzer and Kusch 2003), suggesting a common function across taxa. As each of these common syllables produced by M. tuberculata contain trill notes, they may serve a mate-attraction purpose. Further work comparing syllables produced by males from other populations and subspecies (of which there are three; Carter and Riskin 2006), and whether the same modified trills are as prevalent, would be useful.

Male vocalizations encode identity information. All frequency and temporal characteristics for the four main syllable types showed greater variation between individuals than within individuals, and a DFA including all individuals could correctly assign syllables to the males that produced them approximately 30% of the time, which is more than expected by chance. However, given that the spacing of singing roosts is akin to an “exploded lek” (Höglund and Alatalo 1995; Toth et al. 2015b), sixteen males presents an unrealistic discrimination problem faced by individuals in the lek (e.g. Bee et al. 2001). When the number of males in each DFA was limited to a biologically relevant level of three, average classification success rose to between 66 and 73%. Individual differences in syllable production have been found in the songs of Pipistrellus nathusii (Russ and Racey 2007) and S. bilineata (Davidson and Wilkinson 2002; Eckenweber and Knörnschild 2013). Individual vocal signatures might allow females to track male display performance across the relatively long breeding season before making mate-choice decisions, and/or allow males to identify their neighbors (e.g. Godard 1991). Individual signatures may also facilitate relationships between timeshare males, allowing members of a roost to track who is singing, and for how long. However, it should be noted that the measures of distinctiveness used in this study only incorporated syllables produced by males during a single recording session, which may have undersampled the variation present in syllable production (see Ellis 2008).

Additionally, we found size-duration allometry in one syllable type; larger males had significantly shorter trill-downsweeps (the second most-common syllable). Our results compliment previous research showing that courtship vocalization duty cycles scale negatively with body size in M. tuberculata, and that smaller males have higher paternity success (Toth et al. 2018). Song output is a useful signal for females across various taxa, as singing is an energetically demanding signal that appears to be constrained by energetic costs in many species (see Vehrencamp 2000; Gil and Gahr 2002). In bats, S. bilineata males that possess higher song rates have been found to sire more offspring (Behr et al. 2006), and males with longer calls had more females on their territories (Davidson and Wilkinson 2002, 2004). Recent playback experiments by Smarsh and Smotherman (2017) have shown that territorial male heart-nosed bats (Cardioderma cor) respond more strongly to intruders with longer songs, suggesting that those intruders represent a greater threat. Given that other research has found no relationship between male size and song rate for M. tuberculata (Toth et al. 2018), females may perceive smaller males (with their longer syllables and higher duty cycles) as more attractive. However, whether females can perceive these relatively slight differences between male vocal traits will require further research.

None of the four of the most-common syllable types produced by male M. tuberculata displayed size-pitch allometry. While size-pitch allometry has been demonstrated in non-courtship vocalizations of mammals (Ey et al. 2007; Sell et al. 2010; including bats, Puechmaille et al. 2014), it is thought to be rare both in small animals (Jones and Siemers 2011) and animals with high vocal performance (Patel et al. 2010). Further, fundamental frequency can be a poor predictor of male body mass within species (Charlton and Reby 2016). Mystacina tuberculata appears to adhere to these rules as well.

It is worth noting how high the vocal outputs of M. tuberculata are in comparisons to other species. The Eastern whippoorwill (Caprimulgus vociferus) and the red-eyed vireo (Vireo olivaceus) are two bird species regarded as having some of the highest sustained song outputs (Kroodsma 2005). Across one night of observation (9 h), a single whippoorwill was recorded singing approximately 21,000 songs at a rate of 0.66 songs/s (Kroodsma 2005). Similarly, a single red-eyed vireo has been recorded as producing 22,197 songs within a 14-h period at a rate of 0.44 songs/s (Kroodsma 2005). In contrast, M. tuberculata males produced over 3200 syllables on average over a five-minute period, at a rate of approximately 5 syllables/s. A bout analysis of vocal gap lengths indicated that over a quarter of the males in our study sang almost continuously. Given that the average amount of time spent by males in the singing roosts is approximately 6 h (Toth et al. 2018), many males may be producing over 100,000 syllables nightly. A comparable example from the bat literature would be the songflight (i.e. courtship) calls of the common Pipistrelle bat (P. pipistrellus), which have been recorded at 1.73 calls/s (Jones 1997). Even though songflights can last for hours (Sachteleben and von Helversen 2006), it is unlikely they reach the same outputs as M. tuberculata. Regardless, M. tuberculata likely has one of the highest sustained song outputs yet described.

Based on the evidence presented in this study, the syllables and singing behavior of male M. tuberculata fit the profile of important signals for mate attraction: portions of syllables honestly advertise male characteristics, and syllables contain information on male identity. However, many details of this behavior still need to be examined. Beyond a broad classification of the syllables produced by individual males we have avoided a detailed analysis of male syllable repertoires. Future studies examining if males order their syllables into predictable ‘phrases’ (e.g. Bohn et al. 2009), and if these phrases differ between individuals, would be of interest. The acoustic properties of singing roosts are also currently unknown. Given the importance of singing roosts for displaying males, it may be expected that roosts are selected for their acoustic characteristics. For example, males may use singing roosts as resonating chambers to increase the amplitude of their vocalizations (e.g. Lardner and bin Lakim 2002).

Notes

Acknowledgements

We are grateful to the New Zealand Department of Conservation – particularly T. Thurley and D. Smith – for assistance and support. We thank R. Germain for statistical advice, and three anonymous reviewers for their comments on previous versions of this manuscript.

Funding

Funding was provided by the Australasian Society for the Study of Animal Behaviour, the Australasian Bat Society, Bat Conservation International, and the University of Auckland.

Compliance with Ethical Standards

Conflict of Interest

The authors declare that they have no conflict of interest.

Ethical approval

All applicable international, national, and/or institutional guidelines for the care and use of animals were followed. The capture and marking of individuals was approved by both the University of Auckland Animal Ethics Committee (permit AEC#000920) and the New Zealand Department of Conservation (High Impact Research and Collection Permit WK-32184-RES).

Supplementary material

265_2018_2496_MOESM1_ESM.docx (19.8 mb)
ESM 1 (DOCX 20231 kb)

References

  1. Bee MA, Kozich CE, Blackwell KJ, Gerhardt HC (2001) Individual variation in advertisement calls of territorial male green frogs, Rana clamitans: implications for individual discrimination. Ethology 107:65–84CrossRefGoogle Scholar
  2. Beecher MD (1989) Signalling systems for individual recognition: an information theory approach. Anim Behav 38:248–261CrossRefGoogle Scholar
  3. Beecher MD, Brenowitz EA (2005) Functional aspects of song learning in songbirds. Trends Ecol Evol 20:143–149CrossRefPubMedGoogle Scholar
  4. Behr O, von Helversen O (2004) Bat serenades—complex courtship songs of the sac-winged bat (Saccopteryx bilineata). Behav Ecol Sociobiol 56:106–115CrossRefGoogle Scholar
  5. Behr O, von Helversen O, Heckel G, Nagy M, Voigt CC, Mayer F (2006) Territorial songs indicate male quality in the sac-winged bat Saccopteryx bilineata (Chiroptera, Emballonuridae). Behav Ecol 17:810–817CrossRefGoogle Scholar
  6. Bohn KM, Schmidt-French B, Schwartz C, Smotherman M, Pollak GD (2009) Versatility and stereotypy of free-tailed bat songs. PLoS One 4:e6746CrossRefPubMedPubMedCentralGoogle Scholar
  7. Bohn K, Montiel-Reyes F, Salazar I (2016) The complex songs of two Molossid species. In: Ortega J (ed) Sociality in Bats. Springer, Switzerland, pp 143–160Google Scholar
  8. Bradbury JW (1977) Lek mating behavior in the hammer-headed bat. Z Tierpsychol 45:225–255CrossRefGoogle Scholar
  9. Bradbury JW, Vehrencamp SL (1998) Principles of animal communication. Sinauer Associates Inc., Sunderland, MAGoogle Scholar
  10. Byers BE (2006) Extrapair paternity in chestnut-sided warblers is correlated with consistent vocal performance. Behav Ecol 18:130–136CrossRefGoogle Scholar
  11. Carter GG, Riskin DK (2006) Mystacina tuberculata. Mamm Species 790:1–8CrossRefGoogle Scholar
  12. Catchpole CK, Slater PJB (2003) Bird song: biological themes and variations. Cambridge University Press, Cambridge, UKGoogle Scholar
  13. Charlton BD, Reby D (2016) The evolution of acoustic size exaggeration in terrestrial mammals. Nat Commun 7:12739CrossRefPubMedPubMedCentralGoogle Scholar
  14. Charlton BD, Reby D, McComb K (2007) Female red deer prefer the roars of larger males. Biol Lett 3:382–385CrossRefPubMedPubMedCentralGoogle Scholar
  15. Charlton BD, Ellis WAH, Brumm J, Nilsson K, Fitch WT (2012) Female koalas prefer bellows in which lower formants indicate larger males. Anim Behav 84:1565–1571CrossRefGoogle Scholar
  16. Christie PJ, Mennill DJ, Ratcliffe LM (2004) Pitch shifts and song structure indicate male quality in the dawn chorus of black-capped chickadees. Behav Ecol Sociobiol 55:341–348CrossRefGoogle Scholar
  17. Collins S (2004) Vocal fighting and flirting: the functions of birdsong. In: Marler P, Slabbekoorn H (eds) Nature's music: the science of birdsong. Academic Press, New York, pp 39–79CrossRefGoogle Scholar
  18. Davidson SM, Wilkinson GS (2002) Geographic and individual variation in vocalizations by male Saccopteryx bilineata (Chiroptera: Emballonuridae). J Mammal 83:526–535CrossRefGoogle Scholar
  19. Davidson SM, Wilkinson GS (2004) Function of male song in the greater white-lined bat, Saccopteryx bilineata. Anim Behav 67:883–891CrossRefGoogle Scholar
  20. Eckenweber M, Knörnschild M (2013) Social influences on territorial signaling in male greater sac-winged bats. Behav Ecol Sociobiol 67:639–648CrossRefGoogle Scholar
  21. Ellis JMS (2008) Decay of apparent individual distinctiveness in the begging calls of adult female white-throated magpie-jays. Condor 110:648–657CrossRefGoogle Scholar
  22. ESRI (2012) ArcMap 10. Environmental Systems Research Institute, Redlands, CAGoogle Scholar
  23. Ey E, Pfefferle D, Fischer J (2007) Do age-and sex-related variations reliably reflect body size in non-human primate vocalizations? A review. Primates 48:253–267CrossRefPubMedGoogle Scholar
  24. Fenton MB (2013) Questions, ideas and tools: lessons from bat echolocation. Anim Behav 85:869–879CrossRefGoogle Scholar
  25. Galeotti P, Sacchi R, Rosa DP, Fasola M (2005) Female preference for fast-rate, high-pitched calls in Hermann's tortoises Testudo hermanni. Behav Ecol 16:301–308CrossRefGoogle Scholar
  26. Gil D, Gahr M (2002) The honesty of bird song: multiple constraints for multiple traits. Trends Ecol Evol 17:133–141CrossRefGoogle Scholar
  27. Godard R (1991) Long-term memory of individual neighbours in a migratory songbird. Nature 350:228–229CrossRefGoogle Scholar
  28. Gray DA (1997) Female house crickets, Acheta domesticus, prefer the chirps of large males. Anim Behav 54:1553–1562CrossRefPubMedGoogle Scholar
  29. Höglund J, Alatalo RV (1995) Leks. Princeton University Press, PrincetonCrossRefGoogle Scholar
  30. Jahelková H, Horáček I, Bartonička T (2008) The advertisement song of Pipistrellus nathusii (Chiroptera, Vespertilionidae): a complex message containing acoustic signatures of individuals. Acta Chiropterol 10:103–126CrossRefGoogle Scholar
  31. Jones G (1997) Differences in songflight calls and social calls between two phonic types of the vespertilionid bat Pipistrellus pipistrellus. J Zool 241:315–324CrossRefGoogle Scholar
  32. Jones G, Siemers BM (2011) The communicative potential of bat echolocation pulses. J Comp Physiol 197:447–457CrossRefGoogle Scholar
  33. Knörnschild M, Behr O, von Helversen O (2006) Babbling behavior in the sac-winged bat (Saccopteryx bilineata). Naturwissenschaften 93:451–454CrossRefPubMedGoogle Scholar
  34. Knörnschild M, Nagy M, Metz M, Mayer F, von Helversen O (2010) Complex vocal imitation during ontogeny in a bat. Biol Lett 6:156–159CrossRefPubMedGoogle Scholar
  35. Kroodsma D (2005) The singing life of birds: the art and science of listening to birdsong. Houghton Mifflin Harcourt, New YorkGoogle Scholar
  36. Lardner B, bin Lakim M (2002) Animal communication: Tree-hole frogs exploit resonance effects. Nature 420:475–475CrossRefPubMedGoogle Scholar
  37. Linhart P, Fuchs R (2015) Song pitch indicates body size and correlates with males' response to playback in a songbird. Anim Behav 103:91–98CrossRefGoogle Scholar
  38. Lloyd BD (2001) Advances in New Zealand mammalogy 1990–2000: short-tailed bats. J Roy Soc New Zeal 31:59–81CrossRefGoogle Scholar
  39. Marler P (2004) Science and birdsong: the good old days. In: Marler P, Slabbekoorn H (eds) Nature's Music: the science of birdsong. Academic Press, New York, pp 1–38Google Scholar
  40. Marler P, Slabbekoorn H (2004) Nature's music: the science of birdsong. Elsevier Academic Press, San Diego, CaliforniaGoogle Scholar
  41. McGregor PK, Krebs JR (1982) Mating and song sharing in the great tit. Nature 297:60–61CrossRefGoogle Scholar
  42. Mennill DJ, Otter KA (2007) Status signaling and communication networks in chickadees: complex communication with a simple song. In: Otter KA (ed) Ecology and behavior of chickadees and titmice: an integrated approach. Oxford University Press, Oxford, pp 215–234CrossRefGoogle Scholar
  43. Molles LE, Vehrencamp SL (1999) Repertoire size, repertoire overlap, and singing modes in the banded wren (Thryothorus pleurostictus). Auk 116:677–689CrossRefGoogle Scholar
  44. Morton ES, Young K (1986) A previously undescribed method of song matching in a species with a single song “type”, the Kentucky warbler (Oporornis formosus). Ethology 73:334–342CrossRefGoogle Scholar
  45. Mundry R, Sommer C (2007) Discriminant function analysis with nonindependent data: consequences and an alternative. Anim Behav 74:965–976CrossRefGoogle Scholar
  46. Nordby JC, Campbell SE, Beecher MD (1999) Ecological correlates of song learning in song sparrows. Behav Ecol 10:287–297CrossRefGoogle Scholar
  47. Nordby JC, Campbell SE, Burt JM, Beecher MD (2000) Social influences during song development in the song sparrow: a laboratory experiment simulating field conditions. Anim Behav 59:1187–1197CrossRefPubMedGoogle Scholar
  48. Patel R, Mulder RA, Cardoso GC (2010) What makes vocalisation frequency an unreliable signal of body size in birds? A study on black swans. Ethology 116:554–563CrossRefGoogle Scholar
  49. Pfalzer G, Kusch J (2003) Structure and variability of bat social calls: implications for specificity and individual recognition. J Zool 261:21–33CrossRefGoogle Scholar
  50. Podos J, Huber SK, Taft B (2004) Bird song: the interface of evolution and mechanism. Annu Rev Ecol Evol S 35:55–87CrossRefGoogle Scholar
  51. Poesel A, Kunc HP, Foerster K, Johnsen A, Kempenaers B (2006) Early birds are sexy: male age, dawn song and extrapair paternity in blue tits, Cyanistes (formerly Parus) caeruleus. Anim Behav 72:531–538CrossRefGoogle Scholar
  52. Pollak GD, Casseday J (2012) The neural basis of echolocation in bats. Springer, BerlinGoogle Scholar
  53. Puechmaille SJ, Borissov IM, Zsebok S, Allegrini B, Hizem M, Kuenzel S, Schuchmann M, Teeling EC, Siemers BM (2014) Female mate choice can drive the evolution of high frequency echolocation in bats: a case study with Rhinolophus mehelyi. PLoS One 9:e103452CrossRefPubMedPubMedCentralGoogle Scholar
  54. R Core Team (2013) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna. https://www.R-project.org/
  55. Revelle W (2018) Psych: Procedures for Personality and Psychological Research, Northwestern University, Evanston, Illinois. https://www.CRAN.R-project.org/package=psychVersion=1.8.3
  56. Rice WR, Gaines SD (1994) ‘Heads I win, tails you lose’: testing directional alternative hypotheses in ecological and evolutionary research. TREE 9:236–237Google Scholar
  57. Rivera-Gutierrez HF, Pinxten R, Eens M (2010) Multiple signals for multiple messages: great tit, Parus major, song signals age and survival. Anim Behav 80:451–459CrossRefGoogle Scholar
  58. Russ JM, Racey PA (2007) Species-specificity and individual variation in the song of male Nathusius’ pipistrelles (Pipistrellus nathusii). Behav Ecol Sociobiol 61:669–677CrossRefGoogle Scholar
  59. Ryan MJ (1980) Female mate choice in a neotropical frog. Science 209:523–525CrossRefPubMedGoogle Scholar
  60. Sachteleben J, von Helversen O (2006) Songflight behaviour and mating system of the pipistrelle bat (Pipistrellus pipistrellus) in an urban habitat. Acta Chiropterol 8:391–401CrossRefGoogle Scholar
  61. SAS Institute Inc. (2016) JMP Pro 13. Cary, NC: SAS Institute Inc.Google Scholar
  62. Schmidt S (2013) Beyond echolocation: emotional acoustic communication in bats. In: Altenmuller E, Schmidt S, Zimmermann E (eds) Evolution of emotional communication: from sounds in nonhuman mammals to speech and music in man. Oxford University Press, Oxford, pp 92–104CrossRefGoogle Scholar
  63. Sell A, Bryant GA, Cosmides L, Tooby J, Sznycer D, von Rueden C, Krauss A, Gurven M (2010) Adaptations in humans for assessing physical strength from the voice. Proc R Soc Lond B 277:3509–3518CrossRefGoogle Scholar
  64. Sibly RM, Nott HMR, Fletcher DJ (1990) Splitting behaviour into bouts. Anim Behav 39:63–69CrossRefGoogle Scholar
  65. Smarsh GC, Smotherman M (2015) Singing away from home: Songs are used on foraging territories in the African megadermatid bat, Cardioderma cor. In: Proceeding of Meetings on Acoustics 25:010002. Acoustical Society of America, JacksonvilleGoogle Scholar
  66. Smarsh GC, Smotherman M (2017) Behavioral response to conspecific songs on foraging territories of the heart-nosed bat. Behav Ecol Sociobiol 71:142CrossRefGoogle Scholar
  67. Smotherman MS (2007) Sensory feedback control of mammalian vocalizations. Behav Brain Res 182:315–326CrossRefPubMedPubMedCentralGoogle Scholar
  68. Smotherman M, Knörnschild M, Smarsh G, Bohn K (2016) The origins and diversity of bat songs. J Comp Physiol A 202:535–554CrossRefGoogle Scholar
  69. Sokal RR, Rohlf FJ (1995) Biometry: The principles and practice of statistics in biological research. W.H. Freeman and Co., New YorkGoogle Scholar
  70. Spector DA (1992) Wood-warbler song systems. Curr Ornithol 9:199–238CrossRefGoogle Scholar
  71. Toth CA, Parsons S (2013) Is lek breeding rare in bats? J Zool 291:3–11CrossRefGoogle Scholar
  72. Toth CA, Cummings G, Dennis TE, Parsons S (2015a) Adoption of alternative habitats by a threatened, "obligate" deep-forest bat in a fragmented landscape. J Mammal 96:927–937CrossRefGoogle Scholar
  73. Toth CA, Dennis TE, Pattemore DE, Parsons S (2015b) Females as mobile resources: communal roosts promote the adoption of lek breeding in a temperate bat. Behav Ecol 26:1156–1163CrossRefGoogle Scholar
  74. Toth CA, Santure AW, Holwell GI, Pattemore DE, Parsons S (2018) Courtship behaviour and display-site sharing appears conditional on body size in a lekking bat. Anim Behav 136:13–19CrossRefGoogle Scholar
  75. Vehrencamp SL (2000) Handicap, index, and conventional signal elements of bird song. In: Espmark Y, Amundsen T, Rosenqvist GO (eds) Animal signals. Tapir Academic Press, Trondheim, Norway, pp 277–300Google Scholar
  76. Vignal C, Mathevon N, Mottin S (2004) Audience drives male songbird response to partner's voice. Nature 430:448–451CrossRefPubMedGoogle Scholar
  77. Vignal C, Mathevon N, Mottin S (2008) Mate recognition by female zebra finch: analysis of individuality in male call and first investigations on female decoding process. Behav Process 77:191–198CrossRefGoogle Scholar
  78. Voigt CC, Behr O, Caspers B, von Helversen O, Knörnschild M, Mayer F, Nagy M (2008) Songs, scents, and senses: sexual selection in the greater sac-winged bat, Saccopteryx bilineata. J Mammal 89:1401–1410CrossRefGoogle Scholar
  79. Wells KD (2010) The ecology and behavior of amphibians. University of Chicago Press, ChicagoGoogle Scholar
  80. Wildenthal JL (1965) Structure in primary song of the mockingbird (Mimus polyglottos). Auk 82:161–189CrossRefGoogle Scholar
  81. Wilkinson GS, Boughman JW (1998) Social calls coordinate foraging in greater spear-nosed bats. Anim Behav 55:337–350CrossRefPubMedGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  1. 1.School of Biological SciencesUniversity of AucklandAucklandNew Zealand
  2. 2.Department of Biological SciencesBoise State UniversityBoiseUSA
  3. 3.School of Earth, Environmental and Biological SciencesQueensland University of TechnologyBrisbaneAustralia

Personalised recommendations