Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

Echolocation is an astonishing behaviour mainly associated with bats, although neither restricted to nor characteristic of them. Echolocation (biosonar) is an active mode of orientation involving animals using echoes of sounds they produce to collect information about their surroundings (Galambos and Griffin 1942; Griffin and Galambos 1941). The data for the echolocator is the difference between what it said and what it hears. For echolocation to work, the animal must register the outgoing signal in its brain for comparison with retuning echo(es), and so ideally outgoing signals should not mask returning echoes. Echolocation has been documented in some species of birds (Apodiformes, Caprimulgiformes) and mammals (Afrotheria [Afrosoricida], Laurasiatheria [Lipotyphla, Chiroptera, Cetartiodactyla]) (Thomas et al. 2002). Not all bats echolocate; specifically most of those species belonging to the family Pteropodidae and those few pteropodids (Rousettus spp.) that do echolocate use tongue clicks as signals. All other bats echolocate and do so using signals produced in the larynx.

Current research on echolocating bats has revealed the complexity and sophistication of the behaviour. First, echolocating bats adjust their calls according to the situation, whether tongue-clicking pteropodids (Rousettus aegyptiacus; Yovel et al. 2010) or laryngeal echolocators (e.g. Jakobsen and Surlykke 2010; Surlykke et al. 2009a, b; Ratcliffe et al. 2011). The echolocation behaviour of tongue-clicking pteropodids is not “unsophisticated” as these bats adjust their flight paths and patterns of call production to maximize information return (e.g. Holland et al. 2004; Yovel et al. 2010). Laryngeally echolocating bats adjust their outgoing signals according to clutter (e.g. Brinklov et al. 2010a, b; Hiryu et al. 2010; Stamper et al. 2009), the presence of conspecifics (Chiu et al. 2008, 2009, 2010) and changes that include the use of harmonics (Hiryu et al. 2010; Stamper et al. 2009), as well as adjustments to the sonar beam (Surlykke et al. 2009a, b). Some also adjust call intensities according to the situation which affects both range and strength of echoes (Brinklov et al. 2010a, b). Still others adjust the frequencies that dominate call components apparently affecting range of operation and conspicuousness to hearing-based defences of prey (Goerlitz et al. 2010). The purpose of this chapter is to consider possible origins of echolocation in bats and the role that echolocation may have played in the diversification of bats.

2 Echolocation Signals

A sampling of echolocation calls and relevant features (Figs. 3.1 and 3.2) sets the stage for the information that follows. The general topic of echolocation call design and diversification is well covered in several reviews (e.g. Jones and Holderied 2007; Maltby et al. 2009; Schnitzler and Kalko 2001).

Fig. 3.1
figure 00031figure 00031

Sampling of echolocation signals, including spectrograms (Part 1) and power spectra (Part 2). Spectrograms include echolocation clicks of Rousettus aegyptiacus (a); two search phase calls of Otonycteris hemprichii (b), Lasiurus cinereus (c) and Rhinopoma hardwickei (d); and one each of Pteronotus parnellii (e) and Rhinolophus capensis (f). Times between pairs of calls have been reduced to show pairs of calls. H harmonic, FM frequency modulated, CF constant frequency, qCF quasi-constant frequency. Part 2 shows power spectra, fast Fourier transforms (FFTs) of echolocation calls including Otonycteris hemprichii (a), Lasiurus cinereus (b), Rhinopoma hardwickei (c), Pteronotus parnellii (d) and Rhinolophus capensis (e). For O. hemprichii, L. cinereus and R. hardwickei, FFTs are of the first of the two calls shown in Fig. 3.1, Part 1. FM frequency modulated, qCF quasi-constant frequency, CF constant frequency, H harmonic

Full size image
Fig. 3.2
figure 00032figure 00032

A comparison of low duty cycle (Lasiurus cinereus, a) and high duty cycle (Pteronotus parnellii, b) call sequences. Note different patterns of calls, short calls separated by long periods of silence (low duty cycle) versus longer calls separated by short periods of silence (high duty cycle)

Full size image

2.1 Time and Frequency

Durations and inter-call intervals are measured in ms, the frequencies of echolocation calls (and their components) in kHz. Although most bats use echolocation signals with most energy >20 kHz (by definition, ultrasonic = beyond the range of human hearing), their echolocation signals range in frequency from about 8 to over 200 kHz. Echolocation is not synonymous with ultrasonic. Using humans as reference points for frequency (ultrasonic) or signal strength (dB) is ill-advised. In a group of ten people, it is unlikely that any two will have exactly the same hearing sensitivity, whether response to higher frequencies or threshold.

2.2 Intensity

The strength (intensity) of echolocation calls is usually measured in (dB) sound pressure level (SPL) at a specific distance from the source (e.g. 130 dB SPL at 10 cm; Holderied et al. 2005; Surlykke and Kalko 2008). Intensity (dB) is measured on a log scale where, by definition, 0 dB is the threshold of human hearing at 1 kHz. Griffin (1958) recognized that some bats produced much quieter echolocation calls than others. He distinguished between whispering bats (e.g. phyllostomids, megadermatids, nycterids) and bats that produced low-intensity echolocation calls (~60 dB SPL at 10 cm) and those using higher intensity signals (>110 dB SPL at 10 cm). But we now know that there is a continuum between “whispering” and other bats and that all echolocating species probably use intensities much higher than early estimates suggested. The situation has been best described in phyllostomids.

The echolocation clicks of R. aegyptiacus range in intensity from 85 to 96 dB SPL at 1 m (Holland et al. 2004). Mora and Macias (2007) demonstrated that the phyllostomid Phyllonycteris poeyi produces more intense echolocation calls than expected, but they did not provide SPL data. Working with arrays of microphones, Surlykke and Kalko (2008) and Brinklov et al. (2009, 2010) showed that, measured at 2 m distant, other phyllostomids produced calls of 99 dB SPL RMS at 10 cm (Carollia perspicillata), 105 dB SPL RMS at 10 cm (Macrophyllum macrophyllum) or 110 dB RMS at 10 cm (Artibeus jamaicensis) (Surlykke and Kalko 2008; Brinklov et al. 2009, 2010a, b).

2.3 Signals and Frequency

Bandwidth refers to the range of frequencies in an echolocation signal. Some narrowband signals are CF (bandwidth ~ 0 kHz; Fig. 3.1e, f), but others are broader in bandwidth (quasi-constant frequency—qCF; Fig. 3.1c, d). The echolocation signals of R. aegyptiacus (Fig. 3.1a, and presumably other echolocating Rousettus spp.) are broadband tongue clicks (Holland et al. 2004). The echolocation signals of all other bats are produced in the larynx and are tonal (Fig. 3.1b–f) showing structured changes in frequency over time. Tonal FM signals can be broadband or narrowband, and some calls include both narrowband and broadband components.

2.4 Harmonics

Harmonics (overtones; Figs. 3.1b, d and 3.2a, c–e) are obvious and a consistent feature of the echolocation calls of some, but not all bats (Fenton et al. 2011). Harmonics effectively broaden the bandwidth of echolocation signals, potentially providing more details to the echolocator (Simmons and Stein 1980). Bats such as Eptesicus fuscus use more harmonics when operating in clutter (Stamper et al. 2009). Detecting harmonics in the echolocation calls of free-flying bats in the field often depends upon on the position of the bat relative to the position of the microphone, so only recordings using arrays of microphones where the bat was in (close) proximity may dependably be used in studies of harmonics (Fenton et al. 2011).

2.5 Clutter

“Clutter” could simply be defined as echoes from anything other than the target of interest, but as Denny (2006) pointed out, the situation can be complicated. The definition of clutter will depend upon several features of the bats involved, such as their size, flight speed and inter-call intervals. For a bat to deal with clutter, it must detect clutter and negotiate a course through it. This means that both echolocation (design of calls and behaviour) and manoeuvrability in flight (wing design) will influence how a bat adjusts to clutter (Aldridge and Rautenbach 1987; Norberg and Rayner 1987). Some bats adjust their echolocation calls (harmonics, Stamper et al. 2009; intensity, Brinklov et al. 2010a, b) when operating in cluttered areas such as thick vegetation. Tactile sensitivity of flight membranes seems to augment flight performance under similar circumstances (Chadha et al. 2011; Miller 2005).

2.6 Range

A combination of spherical spreading loss and frequency-dependent atmospheric attenuation (Lawrence and Simmons 1982a) dictates that echolocation in air is a relatively short-range operation. Kick (1982) demonstrated that using echolocation, E. fuscus first detected a 19 mm diameter target at 5 m and a 4.8 mm target at 3 m. In the field, Holderied et al. (2005) used videogrammetry to document that Eptesicus bottae flew 3–8 m s−1 and first detected large objects at ~40 m. They concluded that the inter-call interval (Fig. 3.3) corresponding to lowest peak in mean wingbeat period provided a useful indicator of maximum range of detection of E. bottae-sized targets at about 20 m. Surlykke and Kalko (2008) used an array of microphones and demonstrated that the effective range of detection for insect-sized prey was 20–30 m in many species and more than 60 m in Cormura brevirostris. The long detection range in C. brevirostris coincides with calls consistently produced in triplets that show consistent changes in frequency as well as prominent harmonics.

Fig. 3.3
figure 00033figure 00033

A comparison of the faces of bats with special reference to ears, tragi, noseleafs and related structures. Included are Balantiopteryx io (a), Nyctalus noctula (b), Eptesicus fuscus (c), Otonycteris hemprichii (d), Murina leucogaster (e), Pteronotus personatus (f), Pteronotus parnellii (g), Rhinopoma hardwickei (h), Asellia tridens (i), Rhinolophus clivosus (j), Megaderma lyra (k), Nycteris grandis (l), Phyllostomus discolor (m), Anoura caudifer (n), Desmodus rotundus (o), Chrotopterus auritus (p) and Macrotus waterhousii

Full size image

Flight speed and effective detection range together have important implications for bats hunting for flying prey. Higher flight speeds mean shorter time from detection to contact and challenge bats accordingly. Several species of bats alternate frequencies between adjacent calls perhaps to thwart insect defences (Goerlitz et al. 2010) or as a means of enhancing the range at which they detect prey (Holderied and von Helversen 2003). Mora et al. (2004) reported neurophysiological specializations in the inferior colliculus of Molossus molossus that enhance the effect of alternating frequencies between calls. Ratcliffe et al. (2011) reported that Saccopteryx bilineata alternate calls while foraging but not when they are in the vicinity of their roost. They also proposed that S. bilineata use call alternation to detect prey at short range in cluttered situations (Ratcliffe et al. 2011).

2.7 Self-Deafening (Forward Masking)

Jen and Suga (1977) reported that in Myotis lucifugus contraction of the stapedius muscle in the middle ear reduced hearing sensitivity as an echolocation call was produced. Contractions of the muscle significantly reduce the movement of the ossicles, thereby reducing sound transmission to the cochlea (self-dampening), yet allow the outgoing pulse to be registered in the brain. Alternately, separating pulse and echo in time minimizes the chances of loud outgoing pulses masking faint returning echoes, but it also means that most echolocating bats are not expected to be able to effectively broadcast and receive at the same time. Veselka et al. (2010a, b) showed that contact between stylohyal and tympanic bones was a characteristic of laryngeally echolocating bats that may further facilitate registering registration of the outgoing pulse for future comparison with returning echoes.

Bats in the families Rhinolophidae and Hipposideridae and the mormoopid Pteronotus parnellii use Doppler-shifted echolocation calls so as to separate pulse and returning echoes in frequency. These bats can broadcast and receive at the same time.

2.8 Duty Cycle

Bats that separate pulse and echo in time produce short echolocation calls separated by long periods of silence (Fig. 3.3a), producing a signal about 5 % of the time they are echolocating—a low duty cycle. Species that separate pulse and echo in frequency produce long calls separated by short periods of silence (Fig. 3.3b). These bats broadcast at a high duty cycle, producing a signal >40 % of the time they are echolocating.

Each echolocation call of a high duty cycle bat is dominated by a single frequency (CF), but most calls begin and end with an FM sweep. Echolocation calls of low duty cycle bats are more variable, being FM signals with a range of bandwidths. Even low duty cycle echolocators producing long, narrowband signals (qCF) echolocate at low duty cycle. The advantage of narrowband signals is that for every ten times reduction in bandwidth, signal strength increases by 10 dB with no additional cost of sound production (Oppenheim et al. 1999).

3 Signal Production

With the exception of Rousettus species that use tongue clicks, echolocating bats produce their signals in the larynx. Earlier work provided details of pulse production and anatomy of the larynx and associated structures (e.g. Fattu and Suthers 1981; Griffiths 1978). Tracheal chambers and nasal cavities may further modify signals after they leave the larynx (Suthers et al. 1989). In Rhinolophus hildebrandtii, emitted echolocation calls have most energy in the second harmonic, and three tracheal cavities are involved in suppressing the fundamental frequency of sounds leaving the larynx (Suthers et al. 1989).

Most bats emit their echolocation calls through their open mouths, but some (rhinolophids, hipposiderids, nycterids, megadermatids and phyllostomids) emit them through their nostrils (Pedersen 1995, 1996, 1998). The difference in call emission influences cranial anatomy, but the role that other cranial features, such as concavities in the basisphenoid and/or basioccipital bones, play in echolocation remains unclear (DeBaeremaeker and Fenton 2003).

Noseleafs and related structures (Fig. 3.4) can further influence the signal leaving the bat. Hartley and Suthers (1987) demonstrated that the dorsal lancet of the noseleaf of C. perspicillata directed echolocation signals (short, broadband FM pulses) vertically. Microcomputed tomography reveals that lancet furrows in the noseleaf complex of Rhinolophus affect the FM sweep that terminates each echolocation call (Zhuang and Mueller 2006). Vanderelst et al. (2010) used a comparison of Micronycteris microtis and Phyllostomus discolor to demonstrate that in phyllostomid bats noseleafs can focus energy and help the bats distinguish between echoes from the periphery from those of objects in the area of focus.

Fig. 3.4
figure 00034figure 00034

A comparison of a feeding buzz (Molossus ater, a) and a social buzz (Tadarida teniotis, b)

Full size image

Flying bats tend to produce echolocation calls during the expiration phase of the wingbeat cycle (Suthers et al. 1972), so that they do not incur additional costs associated with sound production during echolocation (Speakman and Racey 1991, but see Parsons et al. 2010). Contractions of abdominal wall muscles further assist call production in flying P. parnellii (Lancaster et al. 1995; Lancaster and Speakman 2001). Of particular note is the striking difference between high duty cycle P. parnellii and low duty cycle Eptesicus, Pipistrellus and Myotis (Lancaster and Speakman 2001) in terms of how the diaphragm contributes to vocalization.

4 Signal Reception

The pinnae of bats (Fig. 3.4) generally increase the sensitivity of hearing, notably so in high duty cycle species where the pinnae are mechanically tuned to the dominant frequencies of their echolocation calls (Obrist et al. 1993). The degree of specialization of pinnae varies across taxa, and in some species, the ears are better tuned to detecting the sounds of prey than to echolocation signals (Obrist et al. 1993). Gleaning bats, such as Megaderma lyra, detect very low-intensity rustling sounds (Hubner and Weigrebe 2003). Sensitivity arising from tuning of the pinnae is further enhanced by neurological tuning in the inferior colliculus (e.g. Mora et al. 2004).

Some parts of the ear, such as the thickened lower edge of the pinnae in Nyctalus spp. combined with the tragus (Fig. 3.4b), reduce extra side lobes in the echoes reaching the bat (Mueller et al. 2006). Early evidence suggested that the tragus affected vertical localization of targets by E. fuscus (Lawrence and Simmons 1982a, b). More recent work with microcomputed tomography indicates that the tragus in E. fuscus reduces extra side lobes from echoes (Mueller 2004). The variety of tragi and ear structures in bats promises new insights from this emerging area of research.

5 Uses for Echolocation

Some species of Rousettus may use echolocation only to gain access to dark roosts and little more (Holland et al. 2004); however, many other species of bats use echolocation to detect, assess and track prey, a process that is particularly clear during feeding buzzes produced during attacks on prey (Fig. 3.5a) (e.g. Kalko and Schnitzler 1989; Schnitzler and Kalko 2001). Echolocation calls of foraging insectivorous bats have received considerable attention, providing a rich array of examples of specific behaviour and call design (e.g. Chiu et al. 2010; Goerlitz et al. 2010; Stamper et al. 2009). The role of echolocation in foraging of fruit eaters and blood feeders remains unclear; however, some New World flowers use nectar guides that reflect ultrasonic signals in order to attract flower-visiting bats and thereby increase rates of pollination (Von Helversen and von Helverson 1999, 2003; Yovel et al. 2008).

Fig. 3.5
figure 00035

The “old” (a) and “new” (b) phylogenies of bats

Full size image

Möhres (1966) noted that the signals one bat uses in echolocation may also affect the behaviour of other bats, and Barclay (1982) used playback experiments to further explore this aspect of echolocation. More recently, more evidence of the role of echolocation signals in communication has expanded our knowledge (e.g. Bayefsky-Anand et al. 2008; Gillam et al. 2007; Schuchmann and Siemers 2010; Swartz et al. 2007; Ulanovsky et al. 2004; Voight-Heucke et al. 2010). Kingston and Rossiter (2004) proposed that enhanced communication may underlie the taxonomic diversification of some high duty cycle echolocating bats. When several bats fly in the same airspace, they may shift the dominant frequencies in their echolocation calls which suggests jamming avoidance and/or enhanced communication (Ulanovsky et al. 2004; Voight-Heucke et al. 2010). In many situations, foraging bats produce social buzzes (Fig. 3.5b) apparently directed at nearby conspecifics. In some situations E. fuscus stop producing echolocation calls apparently to avoid jamming (Chiu et al. 2008).

Information leakage and short operational range are two important drawbacks to using echolocation as a pursuit strategy. Bat-detecting ears have attracted considerable attention (e.g. Pye 1968; Roeder 1967), and their widespread occurrence among insects (Ratcliffe 2009) testifies to the importance of bats as predators. The success of hearing-based defences in many insects reflects the problem of information leakage. Bats use various strategies to circumvent insect defences; behaviours ranging from an allotonic approach (broadcast outside the frequencies to which insect ears are most sensitive; Fenton and Fullard 1979), to stealth, are producing calls of lower intensity (Goerlitz et al. 2010). Bats such as Lasiurus borealis use echolocation calls that are detected by moths with ears, many of which successfully evade the bats (Acharya and Fenton 1992, 1999); however, L. borealis still preys heavily on these moths (Clare et al. 2009).

Laryngeally echolocating bats use morphological specializations of their pinnae that include tuning to a particular frequency or better sound collection (Obrist et al. 1993) combined with neurological adaptations to increase their hearing sensitivity and, thus, their effective target detection range (Neuweiler 1989, 1990). The echolocation-based hunting strategies of many bats are most impressive (Neuweiler 1989), ranging from those detecting, pursuing and attacking flying insects (e.g. Goerlitz et al. 2010; Lazure and Fenton 2011) to those taking prey from the water’s surface (e.g. Hartley et al. 1989; Kalko and Schnitzler 1989; Schnitzler et al. 1994). Other bats may or may not use echolocation when gleaning prey from surfaces (Faure and Barclary 1994; Hubner and Weigrebe 2003; Siemers and Schnitzler 2004), while others may use vision (Bell 1985). The tactile sensitivity of small sensory hairs located on the bats’ wings has also been shown to augment flight performance during pursuit or gleaning activities (Chadha et al. 2011; Miller 2005).

6 Echolocation and the Phylogeny of Bats

Teeling (2009) and Jones and Teeling (2006) have provided excellent overviews of the three hypotheses concerning the origin of echolocation in bats. Simmons and Geisler (1998) proposed that flight was ancestral, providing the ancestors of bats with mobility, and that echolocation evolved later. Simmons et al. (2008) proposed that the Eocene Onychonycteris finneyi could fly but not echolocate, also supporting the flight-first hypothesis (Simmons 2008). Fenton et al. (1995) proposed that echolocation was ancestral in bats and was the key innovation that provided the ancestors of bats access to flying nocturnal insects as food. Speakman and Racey (1991) proposed that flight and echolocation evolved simultaneously in the ancestors of bats allowing them to produce intense signals on the downstroke, covering the cost of signal production with the costs of flight. It appears that this coincidence may not apply to all bats, for example, Mystacina tuberculata (Parsons et al. 2010) and S. bilineata (Ratcliffe et al. 2011).

Morphology, particularly of the shoulder girdle, supports the view that Onychonycteris and other Eocene bats could fly (Simmons and Geisler 1998; Simmons et al. 2008). In bats, connections between stylohyal and tympanic bones are unambiguous morphological indicators of laryngeal echolocation (Veselka et al. 2010a). Simmons and Geisler (1998) used features of the proximal end of the stylohyal bone and the relative size of the cochlea to argue that other Eocene bats (e.g. Icaronycteris, Hassianycteris, Palaeochiropteryx) could echolocate. The two specimens of O. finneyi are not well enough preserved to assess possible contact between tympanic and stylohyal bones, leaving open the question of whether or not the bat had the capacity for laryngeal echolocation (Simmons et al. 2010; Veselka et al. 2010b).

If laryngeal echolocation is ancestral, then it evolved once in the ancestors of Chiroptera but was lost in the lineage leading to Pteropodidae. Within the pteropodids, echolocation by tongue clicking appears to have evolved secondarily once. If laryngeal echolocation was not ancestral, then it could have evolved twice, once in each suborder of Chiroptera (Yinpterochiroptera and Yangochiroptera, Fig. 3.6). As noted by Teeling (2009), the emergence of the new phylogeny and classification of bats has implications for the evolution of echolocation. In the “old” classification, separation of Pteropodidae as Megachiroptera from all other bats presented a different situation than the “new” classification in which both suborders include laryngeal echolocators. High duty cycle echolocation has evolved independently in both Yinpterochiroptera and Yangochiroptera, along with other features such as low-intensity echolocation calls and the use of prey-generated sounds for finding food (Fenton et al. in press). As of 2012 it appears that only the discovery of additional fossil material would resolve the question of the origin of echolocation among Chiroptera. Data on the incidence of genes associated with hearing and vocalizing (Prestin and FoxP2—Li et al. 2007, 2008, respectively) does not resolve the situation or provide unambiguous support for either the traditional (Megachiroptera, Microchiroptera) or “new” (Yinpterochiroptera, Yangochiroptera) phylogeny.

Fig. 3.6
figure 00036

J. D. Smith’s proposed protobat (Hill and Smith 1984)

Full size image

Most data about the phylogeny of bats suggest that Chiroptera is monophyletic (Eick et al. 2005; Simmons 1994; Simmons et al. 1991; Simmons and Geisler 1998), but this view is not unanimous (Pettigrew 1991; Dell et al. 2010; Kruger et al. 2010).

7 Origin of Bats

Although Jepsen (1970) argued that bats evolved in caves, I prefer the suggestion that they evolved in forests with open understory. The protobat has been depicted as being arboreal (Hill and Smith 1984) and presumed to have glided in pursuit of flying insects (Fig. 3.7). Norberg (1985) proposed a model that demonstrated the feasibility of such a transition from gliding to powered flight. Coordinated elongation of arm, hand and finger bones has been shown to be under genetic control during development (Sears et al. 2006). A related scenario applies to the webbing (patagium) of bat wings but not their hind feet (Weatherbee et al. 2006).

Fenton et al. (1995) argued that to function in detecting and tracking flying insects, echolocation signals had to be more intense to increase the effective range of the call. They contended that the use of high-intensity signals in echolocation constituted a “breakthrough”, providing access to a previously inaccessible food source (flying nocturnal insects). Maltby et al. (2009) proposed that the ancestral echolocation call in laryngeally echolocating bats was short, narrowband and multiharmonic. In the initial phase, I think that the protobat would have glided in pursuit of its prey, being a more fundamental step in detecting and tracking targets than the more complicated scenario of when both predator and prey are flying.

This hypothesis makes the prediction that fossil protobats will have laryngeal cartilages robust enough to anchor the cricothyroid muscles that adjust tension on the vocal cords. These cartilages will match increased subglottal pressures associated with echolocation. Precise control of the vocal cords must have been mirrored by simultaneous development of cricothyroid muscles. The appearance of the stylohyal-tympanic connection may have appeared after development of echolocation, followed by diversification and elaboration of facial and ear structures associated with echolocation. It may be that only exceptionally well-preserved fossils will provide the detail necessary to address and perhaps resolve this prediction.

8 Conclusions

The diversity of bats and echolocation behaviour provides as many opportunities for further work as it did 50 years ago; new discoveries open many doors than definitive answers to old questions close. The present array of tools for examining the details of what bats do and how they do it enhances our capacity for better understanding the evolution and diversification of bats. Additional fossil finds may well oblige us to adjust our theories about the origin of bats and echolocation.