Background

During hunting, insectivorous bats mainly emit either multiple frequency modulating (FM) or constant frequency-frequency modulating (CF-FM) echolocation pulses and analyze the returning echoes to extract information about the target features (e.g., size, shape, texture, velocity, range etc.) (Griffin 1958). Based on the frequency content of these echolocation pulses, these bats have been conveniently called FM or CF-FM bats (Griffin 1958; Neuweiler 2003; Ulanovsky and Moss 2008). The echolocation pulses of these bats are typically multiple harmonics with the first one being the fundamental frequency. Also, the echolocation pulse energy is not evenly distributed among different harmonics. For example, the CF-FM bat emits a long CF component followed by a short FM component, and the pulse energy is predominantly distributed in the second harmonic (H 2) (i.e., predominant harmonic) (Griffin 1958; Jones and Teeling 2006; Neuweiler 2003; Suga 1990; Vanderelst et al. 2013). It has been proposed that the CF component is used for echo detection and evaluation of Doppler-shifted echo frequency while the FM component is effective for echo ranging and fine target feature analysis (Fitzpatrick et al. 1991; Jakobsen et al. 2013; Kober and Schnitzler 1990; O’Neill and Suga 1982; Schnitzler and Denzinger 2011; Simmons and Vernon 1971).

In addition to using the echolocation pulses for hunting, the bats also use these pulses as social signals for communication, individual recognition, threatening, and guidance (Möhres 1967; Knörnschild et al. 2012; Kazial et al. 2008; Yovel et al. 2009; Voigt-Heucke et al. 2010). Other studies also showed that the echolocation pulses could be used for identification of species (Russo et al. 2007; Schuchmann and Siemers 2010), age (Jones and Ransome 1993; Masters et al. 1995), and group affiliation (Jameson and Hare 2009; Masters et al. 1995). Presumably, these social interactions are based on the fact that there are variations on pulse parameters among individual bats and between female and male bats (sexual dimorphism).

Sexual dimorphism in the echolocation pulses has been reported for both FM and CF-FM bats (Feng et al. 2003; Grilliot et al. 2009; Jones et al. 1992; Jones et al. 1993; Kazial and Masters 2004; Knörnschild et al. 2012; Russo et al. 2001; Siemers et al. 2005; Suga et al. 1987). Interestingly, most of these studies have been devoted to examining sexual dimorphism of echolocation pulses in the CF-FM bat (Feng et al. 2003; Jones et al. 1992; Knörnschild et al. 2012; Russo et al. 2001; Siemers et al. 2005). This is perhaps due to the fact that this bat performs unique Doppler-shifted compensation behavior during which the bat appropriately adjusts the frequency of the emitted pulse to compensate for the Doppler-shifted frequency of the returning echo. Furthermore, most of these studies mainly examined the sexual dimorphism of the frequency and duration of the H 2 of the echolocation pulses (conveniently referred as CF hereafter) while neglected other pulse parameters. However, recent studies have shown that the FM component of the CF-FM pulse not only plays an essential role in shaping the response properties of central auditory neurons (the inferior colliculus) (Fu et al. 2010) but also facilitates the selectivity of these neurons in duration (Luo et al. 2008; Macías et al. 2013), amplitude (Fu et al. 2011), and the recovery cycle (Tang et al. 2011). These studies suggest that the CF-FM bat may utilize the information provided by the variation in the CF and FM components of the echolocation pulses throughout the entire course of hunting for successful prey capture. These studies also implicate that bats may utilize the variation in the FM components and other pulse parameters if any for social interactions.

In this study, we studied the sexual dimorphism of the echolocation pulses by analyzing multiple pulse parameters (e.g., duration, frequency, bandwidth of the FM component, and repetition rate of echolocation pulses) of the predominant H 2 which has been shown to be essential for successful echolocation in behavioral and neurophysiological studies (Neuweiler et al. 1987; O’Neill and Suga 1982; Schnitzler and Henson 1980; Schnitzler and Denzinger 2011; Suga et al. 1987; Suga 1990). We report here that the male bat has higher CF, wider FM bandwidth, and higher pulse repetition rate (PRR) while the female bat has longer duration of both CF and FM components (conveniently referred as CF duration and FM duration hereafter).

Methods

Animals and morphological studies

In this study, we used 34 adult Hipposideros pratti (H. Pratti) (13 males and 21 females, 34.5–48.1 g body weight (b. w.)) caught from a cave near Xianning city of Hubei providence of People’s Republic of China in May of 2013 and 2014. They were caught from the same colony by means of mist net with the permission of Central China Normal University. These bats were housed in groups (three to five bats each group) in a cage within a temperature- and humidity-controlled animal room and were given larvae of Tenebrio molitor plus vitamin-mineral-enriched water ad libitum. The bats were trained to eat larvae and drink water for about 1 week. The age category of the bat was determined on the basis of whether phalangeal epiphyses were fused with the diaphysis, and its sex was assessed by inspecting the genitalia (Racey 1988). Each bat was weighed to the nearest 1 g with a Pesola balance, and the length of its right forearms was measured to the nearest 0.1 mm with dial calipers. This study was conducted with the approval of the Institutional Animal Care and Use Committee of Central China Normal University, Wuhan, Hubei, People’s Republic of China.

Recordings of echolocation pulses

The recording of the echolocation pulses was made a few days after each bat was acclimated to the laboratory condition. The recording of the echolocation pulses was made at about 7 p.m. in May of 2013 and 2014. During recording, each bat was hanging on the ceiling of the experimental anechoic room and the recording microphone was placed 1 m below. The echolocation pulses were recorded only from each sedentary bat in real time with a handheld ultrasound detector (Petterson D1000X; Pettersson Elektronik AB, Uppsala, Sweden) before being fed into a notebook computer. Each bat was recorded 1–3 times and each time lasted 3–5 s.

Pulse analysis

The echolocation pulses were analyzed using the software BatSound pro 3.31b (Pettersson Elektronik AB, Uppsala, Sweden), with a fast Fourier transformation (FFT) size of 1024 points and a Hanning window using a cursor and visual determination on a screen. We only analyzed high-quality pulses with good signal-to-noise ratio. To characterize each echolocation pulse, we analyzed the CF, FM bandwidth, and durations of both CF and FM components of the H 2 as well as the PRR. The time parameters were taken from amplitude waveforms, and frequency parameters were taken from spectrograms. In the present study, when we calculated the CF difference between the recorded pulses from male and female bats, we firstly calculated the CF average of all pulses recorded from a same male or female bat as a male or female individual, i.e., as a sampling unit. Then, we put each CF average from each male or female bat together and calculated the total mean CF value from all male and female individuals.

Statistical analysis

All data obtained were processed and plotted using Sigmaplot 2000 (Systat Software, San Jose, CA, USA). They were then quantitatively examined and statistically compared using SPSS 13.0 (Student’s t test at p < 0.05) (SPSS, Chicago, IL, USA).

Results

The forearm and body weight of Hipposideros pratti

The measurement of the forearm and body weight of these 34 bats is shown in Table 1. It is clear that there are differences between male and female bats and the former have longer average forearm length but slightly lighter body weight than the latter. However, these differences are not statistically significant (p > 0.05).

Table 1 Measurements of the forearm and body weight of 13 male and 21 female Hipposideros pratti
Full size table

Echolocation pulses of Hipposideros pratti

In this study, 1574 individual echolocation pulses were recorded from these 34 bats ranging from 24 to 76 pulses (average 46 pulses) from each bat. Fig. 1 shows the sonogram and power spectrum of a representative echolocation pulse. The echolocation pulse consists of 4 harmonics with the first harmonic being the fundamental frequency. Each of these harmonics starts with a long CF component followed by a short FM component. The analysis of the power spectrum of the pulse clearly shows that the second harmonic (H 2) is the predominant one that contains the highest energy. This is evident by the fact that the H 2 has the darkest band in Fig. 1a and the highest peak in Fig. 1b. The measurements of CF, FM bandwidth, CF duration, FM duration, and PRR of the H 2 of echolocation pulses from all these 34 bats are shown in Table 2. Although there are differences among these multiple pulse parameters, the H 2 of this bat has an average of about 6.3-ms CF component at 60 kHz followed by 2.5-ms FM component with 10-kHz bandwidth. Each hanging bat emitted its echolocation pulses at a rate about 14 pulses per second.

Fig. 1
figure 1

The sonogram (a) and power spectrum (b) of a representative echolocation pulse emitted by a hanging Hipposideros pratti. The CF (kHz), FM bandwidth (kHz), CF duration (ms), and FM duration (ms) of this echolocation pulse were 60.0, 12.8, 7.5, and 2.7. Clearly, the second harmonic is the predominant one that contains the highest energy

Full size image
Table 2 Parameters of the predominant second harmonic (H 2) of the echolocation pulses emitted by Hipposideros pratti
Full size table

Sexual dimorphism of echolocation pulse parameters in Hipposideros pratti

We examined the sexual dimorphism of echolocation pulses in this bat species by separately plotting the distribution of average value of these multiple parameters of echolocation pulses obtained from each male and female bat. As shown in Fig. 2, the predominant CF of the male bat was between 58.8 and 60.5 kHz with most above 59.5 kHz (61.5 %, Fig. 2(A-1)) while that of the female bat was between 58.0 and 59.9 kHz with most below 59.5 kHz (57.1 %, Fig. 2(A-2)). This sexual dimorphic difference in CF (0.63 kHz) was statistically significant (59.91 ± 0.43 vs. 59.28 ± 0.56 kHz, n = 13 for male, n = 21 for female, T = 3.509, p < 0.01, t test, Fig. 2(A-3)). The FM bandwidth of the male bat ranged from 9.1 to 12.7 kHz with most above 9.5 kHz (69.28 %, Fig. 2(B-1)) while that of the female bat ranged from 8.1 to 10.9 kHz with most below 9.5 kHz (61.9 %, Fig. 2(B-2)). This sexual dimorphic difference in FM bandwidth was also statistically significant (10.08 ± 1.01 vs. 9.38 ± 0.81 kHz, n = 13 for male, n = 21 for female, T = 2.239, p < 0.05, t test, Fig. 2(B-3)). The PRR of the male bat ranged between 4.0 and 31.0 Hz with most (61.5 %) higher than 15.0 Hz (Fig. 2(C-1)). Different from this observation, the PRR of the female bat ranged between 2.7 and 27.0 Hz and most (85.7 %) of them were below 15 Hz (Fig. 2(C-2)). A statistical analysis revealed a significant sexual dimorphic difference in the average PPR of echolocation pulse of this bat species (16.3 ± 7.8 vs. 10.9 ± 6.2 Hz, n = 13 for male, n = 21 for female, T = 2.288, p < 0.05, t test, Fig. 2(C-3)).

Fig. 2
figure 2

Sexual dimorphism of the CF, FM bandwidth, and PRR of echolocation pulses emitted by hanging Hipposideros pratti. A-1, B-1, C-1 vs. A-2, B-2, C-2: the distribution of CF, FM bandwidth, and PRR of echolocation pulses recorded from male and female bats. A-3, B-3, C-3: statistical comparison of these three pulse parameters recorded from male and female bats. The number within each parenthesis indicates the sample size. Student’s t test: *p < 0.05; **p < 0.01

Full size image

The distribution of the averaged CF and FM duration of the predominant H 2 component of the echolocation pulses of each male and female H. pratti is shown in Fig. 3. The CF duration of the predominant H 2 of the male bat was between 4.2 and 8.8 ms with more than half (69.2 %) shorter than 6 ms (Fig. 3(A-1)). In contrast, the CF duration of the predominant H 2 of the female bat was between 5.4 and 10.9 ms with most (81.0 %) longer than 6.0 ms (Fig. 3(A-2)). This sexual dimorphic difference in the CF duration of the predominant H 2 was statistically significant (5.7 ± 1.1 vs. 7.3 ± 1.6 ms, n = 13 for male, n = 21 for female, T = −3.066, p < 0.05, t test, Fig. 3(A-3)). The FM duration of the predominant H 2 of the male bat was between 1.3 and 2.9 ms with 61.5 % shorter 2.5 ms while that of the female bat ranged from 2.0 to 3.2 ms with most (71.4 %) longer than 2.5 ms (Fig. 3(B-1) vs. (B-2)). The sexual dimorphic difference in this measurement was also statistically significant (2.4 ± 0.3 vs. 2.7 ± 0.4 ms, n = 13 for male, n = 21 for female, T = −3.058, p < 0.05, t test, Fig. 3(B-3)).

Fig. 3
figure 3

Sexual dimorphism of the CF and FM duration of echolocation pulses emitted by hanging Hipposideros pratti. A-1, B-1, C-1 vs. A-2, B-2, C-2: the distribution of CF and FM duration of echolocation pulses recorded from male and female bats. A-3, B-3, C-3: statistical comparison of these two temporal parameters of echolocation pulses recorded from male and female bats (see Fig. 2 for legends)

Full size image

Discussion

The forearm length, body weight, and echolocation pulses of Hipposideros pratti

The forearm length, body weight, and the predominant CF of H. pratti measured in the present study (Fig. 1, Tables 1 and 2, bats caught in Hubei province) were similar to the measurements of the same bat species reported in a previous study (bats caught in Fujian, Guangxi, and Henan provinces, Zhang et al. 2009) but different from that reported in another study (bats caught in Guizhou province, Chen et al. 2002). The difference in these measurements of the same bat species may be due to geographic difference. Regardless of these different reports, our present study did not find any sexual dimorphism in the forearm length and body weight of H. pratti (Table 1).

Sexual dimorphism of multiple parameters of the predominant H 2

In this study, we observed the sexual dimorphism of multiple parameters in the predominant H 2 of the echolocation pulses of H. pratti (Figs. 2 and 3, Table 2). On average, the male bat has higher predominant CF, wider FM bandwidth, and higher PRR while the female bat has longer duration of the CF and FM components. This observed sexual dimorphism in the multiple pulse parameters in this bat species is not likely related to the difference in the forearm length and body weight since these two latter physical parameters were not significantly different between the male and female bats (Table 1).

Previous studies on sexual dimorphism of echolocation pulses of the CF-FM bats were mainly focused on the predominant CF (Jones et al. 1992, 1993; Knörnschild et al. 2012; Russo et al. 2001; Feng et al. 2003; Siemers et al. 2005; Suga et al. 1987). For comparison, data on the sexual dimorphism of the predominant CF in different CF-FM bats examined in these previous and the present studies are assembled in Table 3. It is clear that significant sexual dimorphism in the predominant CF is observed in all but two bat species (e.g., Rhinolophinae euryale and Rhinolophinae mehelyi, Russo et al. 2001). It is also clear that the predominant CF of the same bat species (Rhinolophinae euryale) varies with geographic location (Russo et al. 2001; Siemers et al. 2005). This finding is similar to a recent report (Jiang et al. 2010).

Table 3 Sexual dimorphism in the CF in echolocation pulses of CF-FM bats
Full size table

As shown in Table 3, sexual dimorphism in the predominant CF varies with bat species. There is no consistent pattern in the sexual dimorphism in the predominant CF difference in these bat species. For example, in Rhinolophidae, the CF was higher in the female than in the male bat but the opposite was observed in other studies. This is also true in the predominant CF in all Hipposiderinae bat species studied. On the one hand, these different observations on the sexual dimorphism of the predominant CF might be due to species-specific difference, different methods of analysis of the echolocation pulses, or due to the different ways that the echolocation pulses were recorded (Luo et al. 2007). On the other hand, it has been suggested that these different observations might be due to geographic location, different dimensions of vocal cord, and sex hormone level (Neuweiler et al. 1987; Russo et al. 2001). While the largest sexual dimorphic CF difference is 10.49 kHz in Hipposideros caffer (Jones et al. 1993), the smallest sexual dimorphic CF difference is only 0.8 kHz in Rhinolophinae euryale (Siemers et al. 2005). The CF difference of 0.63 kHz between male and female bats (H. pratti) in the present study is comparable and significant (Fig. 2(A-1) vs. (A-2)). Because the bats were from a same colony of the same cave, there were the possible relativity among the individuals and the possible similarity among the recorded pulses.

In addition to examining sexual dimorphism in the predominant CF, we also observed that the FM bandwidth was significantly wider and the PRR was significantly higher in male than in female H. pratti (Fig. 2). On the contrary, the duration of both CF and FM was significantly longer for female than for male bat (Fig. 3). Whether these multi-parametric dimorphic differences are due to the length and width of vocal cord between the male and female bats remains to be studied. If they were different significantly, combined with the fact that the prey detection may be a primary selective force influencing the echolocation features of bats, it might help us to realize our data. In the case of female bats, they have lower CF than male bats due to their laryngeal structure; to increase pulse intensity, they may shorten the bandwidth (Lazure and Fenton 2011); and to get adequate information, they may increase the duration, simultaneously leading to the PRR reduction considering that low-duty-cycle bats separate pulse and echo in time (Lazure and Fenton 2011; Ho et al. 2013). And the echolocation features of male bats are opposite to the female bats, which may also be beneficial for them to make best use of the predation space and distinct each other and resist jamming.

Possible biological significance of this study

Although there have been studies on sexual dimorphism of bat communication sounds (Kanwal et al. 1994; Liu et al. 2013; Ma et al. 2006), many studies have been mainly focused on the sexual dimorphism of orientation pulses in the CF-FM bat by analyzing the predominant CF (Jones et al. 1992, 1993; Knörnschild et al. 2012; Russo et al. 2001; Feng et al. 2003; Siemers et al. 2005; Suga et al. 1987). In the present study, we have clearly shown that sexual dimorphism of echolocation pulses in the CF-FM bat can also be found in other multiple pulse parameters. For this reason, the present study not only complements but also advances all previous studies with more detailed analysis of sexual dimorphic differences in the echolocation pulses of the CF-FM bat.

While the pulse energy is unevenly distributed on the multiple harmonics of the echolocation pulses of the CF-FM bat, the predominant H 2 has been shown to be essential for successful echolocation in behavioral and neurophysiological studies (Neuweiler et al. 1987; O’Neill and Suga 1982; Schnitzler and Henson 1980; Schnitzler and Denzinger 2011; Suga et al. 1987; Suga 1990). Because of the time constraint, as the first step, we studied the sexual dimorphism in the multiple pulse parameters of the predominant H 2 of echolocation pulses emitted from sedentary bats mimicking the hanging bats in the roosting cave. An analysis of these multiple parameters of echolocation pulses emitted by flying bat during hunting or foraging should be conducted to further study the sexual dimorphism in echolocation pulse parameters of this bat species. Conceivably, sexual dimorphism in the predominant H 2 of echolocation pulse parameters may also be observed under field study.

During hunting, the bat systematically varies the echolocation pulses in multiple parametric domains to obtain as much information as possible from the returning echoes for successful prey capture and avoidance of obstacles (Griffin 1958; Jen and Kamada 1982; Schnitzler and Henson 1980). It has been shown that the dynamic variation in multiple pulse parameters throughout the prey approach sequence improves the response sensitivity of bat auditory neurons in multiple parametric domains and likely increases the success of hunting (Jen et al. 2001; Wu and Jen 2006, 2008). As mentioned in the “Background” section, many studies have shown that the bats use echolocation pulses for prey capture as well as for social signals for communication, individual recognition, threatening, and guidance (Jameson and Hare 2009; Jones and Ransome 1993; Kazial et al. 2008; Knörnschild et al. 2012; Masters et al. 1995; Möhres 1967; Russo et al. 2007; Schuchmann and Siemers 2010; Voigt-Heucke et al. 2010; Yovel et al. 2009). If our observed sexual dimorphism of echolocation pulses also occurs in nature, then the bats may potentially use this sexual dimorphism cues in pulse parameters for social communication and species and sex identification.

Conclusions

The present studies examined the sexual dimorphism in echolocation pulse parameters of the CF-FM bat, H. pratti. The data showed that on average the male bat has higher predominant CF, wider FM bandwidth, and higher pulse repetition rate while the female bat has longer duration of the CF and FM components. These observations suggest that bats may potentially use this sexual dimorphism in echolocation pulse parameters for social communication and species and sex identification.