The communicative potential of bat echolocation pulses
Ecological constraints often shape the echolocation pulses emitted by bat species. Consequently some (but not all) bats emit species-specific echolocation pulses. Because echolocation pulses are often intense and emitted at high rates, they are potential targets for eavesdropping by other bats. Echolocation pulses can also vary within species according to sex, body size, age, social group and geographic location. Whether these features can be recognised by other bats can only be determined reliably by playback experiments, which have shown that echolocation pulses do provide sufficient information for the identification of sex and individual in one species. Playbacks also show that bats can locate conspecifics and heterospecifics at foraging and roost sites by eavesdropping on echolocation pulses. Guilds of echolocating bat species often partition their use of pulse frequencies. Ecology, allometric scaling and phylogeny play roles here, but are not sufficient to explain this partitioning. Evidence is accumulating to support the hypothesis that frequency partitioning evolved to facilitate intraspecific communication. Acoustic character displacement occurs in at least one instance. Future research can relate genetic population structure to regional variation in echolocation pulse features and elucidate those acoustic features that most contribute to discrimination of individuals.
KeywordsSignalling Biosonar Auditory Vocalisations Eavesdropping
The key functions of echolocation pulses are to generate echoes that permit a bat to negotiate its three-dimensional environment and, for many bats, to find food (Neuweiler 1989, 1990; Schnitzler et al. 2003). Bat echolocation pulses often are very loud (>120 dB SPL 10 cm in front of the bat’s head for bats in open space; Holderied and von Helversen 2003; Surlykke and Kalko 2008) and are emitted at repetition rates of about 2–20 pulses/s when the bat is in search or commuting flight. Thus, they provide a rich potential source of information to other bats that are within hearing distance. Due to the high frequencies typically used, echolocation pulses will attenuate quickly as they travel through air. Nevertheless, the range over which a bat can be heard by other bats is much larger than the range over which it can receive supra-threshold echoes of its own pulses. For a bat with an assumed detection threshold at 20 dB SPL, a 12 kHz pulse will be audible over 281 m, a 20 kHz pulse over 128 m, a 50 kHz pulse over 35 m and a 100 kHz pulse over 16 m (calculated for 20°C, 80% humidity, 102 Pa air pressure and open airspace; for details of our calculations see, e.g. Hoffmann et al. 2007). For the same frequencies and conservative echo detection threshold (20 dB SPL), the bat would receive an audible echo from a large insect (target strengths 20 dB) over only 14, 11, 6 and 3 m, respectively, and from a flying predator or conspecific (target strengths 0 dB) over only 32, 22, 9 and 5 m, respectively (for details of echo level calculations, see e.g. Safi and Siemers 2010; compare also Holderied and von Helversen 2003, who used a lower echo detection threshold of 0 dB). Echolocation pulses obviously have to be loud in order to generate echoes over a distance large enough to permit efficient spatial orientation and foraging. The pulses thus are frequently audible to other bats, and other animals, including prey and predators, i.e. echolocation inevitably has a ‘communicative’ function (Fenton 1985). This is the flip side of active sensing.
Echolocation pulses evolved for orientation, and in many cases for the detection, localization and classification of prey. Their features are shaped according to task-specific acoustic challenges, and also by phylogenetic constraints (Jones and Teeling 2006). Bats have evolved a wide range of social calls specifically for communication. For several species, the repertoires of both echolocation pulses and social calls are now well described (e.g. Rhinolophus ferrumequinum (Jones and Rayner 1989; Ma et al. 2006); Pteronotus parnellii (Kanwal et al. 1994; Macías et al. 2006); Tadarida brasiliensis (Gillam and McCracken 2007; Bohn et al. 2008) and Megaderma lyra (Schmidt et al. 2000; Bastian and Schmidt 2008; Janssen and Schmidt 2009)). Hence bats have distinctive repertoires of echolocation pulses and social calls, and although the latter are the primary signals used in communication, echolocation pulses nevertheless provide public information (Danchin et al. 2004) that can potentially be exploited by other bats.
Hearing in a given bat species is especially sensitive to the frequencies used in echolocation (Bohn et al. 2006). Hence receivers may be especially sensitive to frequencies emitted by conspecifics, and the potential for the broadcast of inadvertent social information (Danchin et al. 2004) is large. Bats that emit long constant frequency (CF) components in pulses have auditory foveae, meaning that their hearing is tuned sharply to the frequencies they use in echolocation (Schuller and Pollack 1979). Some CF bats emit ‘personalised frequencies’ that are matched to the frequencies of their acoustic foveae, and this may reduce the problems of confusing echoes from their own pulses with those of conspecifics (‘jamming’) (Suga et al. 1987). Bats may reduce the impact of jamming by changing spectral and/or temporal characteristics of pulses (e.g. Ulanovsky et al. 2004). There is even recent behavioural evidence suggesting that bats do not only use information about other bats contained in their pulses, but may also use pulses and potentially even echoes of other bats’ pulses for spatial orientation (Chiu et al. 2008). The evidence comes from bats that were flown in dyads in a flight room. Repeatedly, one bat ceased calling for several hundreds of milliseconds, especially when the animals flew in parallel. Likely, the bats use silence as a strategy to avoid interference from sonar vocalizations of the neighbour, while listening to conspecific-generated acoustic pulses to guide orientation or even by ceasing calling and exploiting information from conspecific echolocation (Chiu et al. 2008).
In this paper, we review the available evidence for a role of echolocation pulses in communication. As it is not yet clear in many cases whether bats intend to convey information to other bats in their echolocation pulses or rather do so inadvertently, and whether there is any selection pressure acting on the information content of echolocation signals to potential receivers other than the calling bat, we use a broad definition of communication. It encompasses both the intentional transmission of information from a sender to a receiver (communication sensu strictu; compare Bradbury and Vehrencamp 1998) and the use of echolocation pulses as cues by unintended receivers; i.e. by eavesdroppers. We first compile the published evidence for signatures in echolocation pulses with respect to species, sex, age, body size and condition, geography, individual identity, group or colony membership and current behaviour of the calling bat. In doing so, we distinguish between pulse signatures that were revealed by an array of digital signal analysis techniques and evidence from behavioural tests showing that bats can actually extract the respective information from the pulses. We then discuss the roles of echolocation in providing ‘private bandwidths’ that facilitate intraspecific communication in bat communities. We argue that interspecific differences in echolocation pulse frequency among bats occupying adjacent frequency bands evolve for effective intraspecific communication, in addition to being correlates of body size or because they promote the detection of prey of different size classes. Effective communication by echolocation also contributes potentially to reproductive isolation and consequently speciation. We conclude by identifying future directions of research that we consider worthwhile pursuing to deepen our understanding of the role of bat echolocation for communication.
Species identity information in echolocation pulses
Many bat species emit distinctive echolocation pulses because pulse design is shaped by ecological factors, and most bats inhabit distinctive ecological niches. Echolocation also is specifically tailored to the current behavioural task a bat is facing. For example bats that fly in open spaces often emit pulses that are of long duration (because forward masking is not a problem—targets are distant and echoes return relatively late). Open space foragers also typically show long pulse intervals (there is ample time to process returning echoes before the subsequent pulse is emitted), and call at low frequencies (because echoes need to return from distant targets and low frequencies are less subject to attenuation) (Schnitzler and Kalko 2001; Schnitzler et al. 2003; Jones and Teeling 2006). Pulse frequency is also often negatively related to body size across species within families of bats (Jones 1999). Thus, ecology, allometric scaling and also phylogeny explain some of the variation of echolocation pulse parameters amongst the >900 extant echolocating bat species.
Although bats modify echolocation pulses according to whether they are commuting, searching for food or capturing it (task-specific pulse structure), echolocation pulses are often relatively reliable badges of species identity to researchers (e.g. Parsons and Jones 2000; Russo and Jones 2002), and could therefore, potentially also be used by bats to distinguish conspecifics from heterospecifics. Indeed, selection imposes task-specific constraints on echolocation pulse design that can result in echolocation pulses being more stereotyped in structure than social calls (Fenton 1994). Recent behavioural experiments suggest that horseshoe bats can distinguish their own species from sympatric congeners by listening to their echolocation pulses and even discriminate between pulses from different congeners (Schuchmann and Siemers 2010). Nevertheless, differences among taxa are more pronounced in some groups than in others. For example, echolocation pulses are often similar in sympatric species of the genus Myotis (Parsons and Jones 2000; Russo and Jones 2002), and can be problematic for species identification by humans performing acoustic surveys. Whether species identification is easier for bats than for humans can only be evaluated by carefully designed experiments.
While bats are not songbirds and the primary selection pressure on pulse design is on generation of ‘optimal’ echoes (Barclay 1999), the evidence summarized above shows that distinctive echolocation pulses allow great potential for species identification by bats. Most research on factors influencing intraspecific variation in echolocation pulses has focussed on species that produce long CF components in their pulses, because these dominant frequencies that are relatively stable (at least when the bats are perched and are not performing Doppler shift compensation). The few experimental studies that have been performed have largely used vespertilionid bats that produce frequency modulated (FM) pulses, however. We shall initially review whether sex, age, body condition and even individual identity can potentially be revealed by pulse characteristics, and in each case discuss whether it is in the signaller’s interest to have its identify revealed. When discussing intra- and interspecific, variation in pulse characteristics, it is always important to ask whether there are benefits to the signaller that result from being identifiable. Although benefits to receivers can readily be postulated, selection will act on the signalling individual, which may in fact benefit by disguising its identity.
Sex differences in echolocation pulses
Identifying an individual’s sex from its echolocation pulses could potentially be important for bats. Maternity colonies are often dominated by females, and females may benefit by locating and following same-sex individuals when returning to roosts for example. Signallers may benefit if flying with other females reduces predation risk. Sex differences in echolocation pulses may also benefit both signallers and receivers in mate location, and listening to echolocation pulses may potentially allow mate location at distances greater than is possible through olfaction. Although benefits of sex recognition and advertisement are likely, is there any evidence that sex can be encoded reliably by echolocation pulse characteristics?
Male and female bats emit echolocation pulses that have different mean frequencies in at least eight species, making it likely that individuals can recognise the sex of at least some individuals. Considerable overlap still occurs between pulse frequencies of the sexes in most of these species, however. Most studies have been performed on species that emit constant frequency (CF) components in echolocation pulses. Among these, females emit pulses of higher frequency than those of males in Pteronotus parnellii (Suga et al. 1987), Rhinolophus rouxi, R. hipposideros and Asellia tridens (reviewed in Jones 1995). More recent studies have shown that female R. blasii (Siemers et al. 2005), R. pumilus (Yoshino et al. 2006) and R. monoceros (Chen et al. 2009) also emit pulses of higher frequency. In at least six species of rhinolophids, frequency does not differ between sexes (reviewed in Siemers et al. 2005). Although females emit higher frequency pulses than males in at least five rhinolophid species, male Hipposideros speoris and H. ruber (Guillén et al. 2000) emit pulses that are typically higher in frequency than those of females (reviewed in Jones 1995). Sexual differences in echolocation pulse frequency are not related to sexual size dimorphism in a consistent way, although in most rhinolophids females are the larger sex (Jones 1995). In a study on R. rouxi, Gerhard Neuweiler provided the only documented case where males and females show virtually no overlap in pulse frequency, and frequency can then encode sex of the signaller with certainty (Neuweiler et al. 1987).
As yet, there is only a single behavioural study that tested whether bats can distinguish sex of conspecifics by their echolocation pulses. During and after playback of echolocation pulses from an unfamiliar bat the vocalization rates of the test bats, Eptesicus fuscus, a species that uses FM pulses, changed depending on the sex of the bat whose pulses were used as a playback stimulus (Kazial and Masters 2004). The test pulses were recorded from bats in confined spaces and in accordance with the findings from these playback experiments, Grilliot et al. (2009) recently showed sex differences in ultrasonic vocalizations of the same species in a confined space. However, Grilliot et al. (2009) found no sex differences in pulse features from bats attached to a zip-line in open airspace.
Age differences in echolocation pulses
Although receivers could potentially benefit from identifying the age of conspecifics (e.g. by distinguishing potential mates (individuals of reproductive age) from non-reproductive individuals; or by identifying subordinates in competitive interactions), signallers may wish to disguise competitive potential, so selection for age-related acoustic badges may not occur unless there are clear advantages to the signaller (e.g. by reducing aggressive behaviour initiated by dominant individuals).
In all species where age differences in pulse frequency have been identified, juveniles emit lower frequencies than adult bats (Asellia tridens, Rhinolophus hipposideros, R. ferrumequinum, Myotis daubentonii, M. lucifugus, reviewed in Jones (1995); more recent data from R. euryale, R. mehelyi (Russo et al. 2001), R. blasii (Siemers et al. 2005), R. pumilus (Yoshino et al. 2006) and R. monoceros (Chen et al. 2009). Again, however, overlap between age classes can be extensive, and age categorisation can only be done in a probabilistic manner. Indeed, in R. ferrumequinum, echolocation pulse frequency increases during the first 2 years of life, stabilises, and then falls off again in old age (>10 years, Jones and Ransome 1993).
As a general pattern in both bats that use CF pulses and bats that use FM pulses, echolocation pulses increase gradually in frequency during early vocal ontogeny (Moss et al. 1997; Vater et al. 2003; Liu et al. 2007). This process is often not complete when young bats become volant a few weeks after birth, and might thus largely explain the juvenile-adult difference in echolocation pulse parameters. Research into the morphological and physiological mechanisms responsible for age-related changes in pulse frequency would be revealing. A behavioural test as to whether bats can discriminate age class from echolocation pulses remains to be conducted.
Correlates of body size or body condition in echolocation pulses
Body mass and condition may be signals of potential competitive ability or mate quality, and such signals may be under strong selection to be robust to deceit (Maynard Smith and Harper 2004). Is there evidence that echolocation pulse characteristics can be honest signals of competitive ability or quality in echolocating bats? Indirect evidence may come from correlations between likely indices of quality, such as body mass or condition (i.e. mass corrected for size), and pulse features.
CF frequency is highest in bats of intermediate size in Asellia tridens (Jones et al. 1993). In Myotis adverus (=macropus; Cooper et al. 2001), a species that emits FM pulses, larger males emit pulses with lower dominant frequencies. A similar trend is seen in first-year M. daubentonii (Jones and Kokurewicz 1994). Conversely, CF frequency is positively related to body condition (mass/forearm length) in Hipposideros fulvus (Jones et al. 1994), Hipposideros ruber (Guillén et al. 2000) and R. mehelyi (Siemers et al. 2005).
A clear link between body condition or size and echolocation pulse frequency seems to be the exception rather than the rule, however. In most species studied, no such relationships are apparent (Jones 1995; Siemers et al. 2005) and in general there is little evidence that echolocation pulse frequency can reveal an individual’s potential competitive ability or mate quality reliably. Moreover, in Pteronotus parnelli pulse frequency and associated neural tuning shift concomitantly in relation to body temperature (Huffman and Henson 1993a, b), confounding any potential that frequency may have to signal or assess an individual’s condition reliably. It has not yet been tested experimentally if and to what degree bats can extract information on body condition and hence potential mate quality from echolocation pulses.
Social group signatures in echolocation pulses
Bats may benefit by encoding information about their social group in echolocation pulses if for example relatives or roost mates show reciprocal altruism with one another (Wilkinson 1984), share information about foraging sites (Wilkinson 1992), or if calls co-ordinate group foraging. Many animals produce group-specific vocalizations (e.g. chimpanzees, Crockford et al. 2004). Greater spear-nosed bats Phyllostomus hastatus produce social calls that are group-specific to co-ordinate group foraging, and can identify individuals from specific social groups by call features alone (Boughman 1997; Boughman and Wilkinson 1998). However, social calls—unlike echolocation pulses —are not so tightly constrained by perceptual needs such as optimal echo generation, and thus can encode high levels of individual variation (Fenton 1994; Siemers 2006). Social calls can also be influenced markedly by social learning from group mates (Esser 1994). Nevertheless, echolocation pulse features varied significantly among maternity roosts of Myotis lucifugus (Pearl and Fenton 1996), and such differences do not appear to be merely the consequences of different amounts of clutter outside roosts (Jameson and Hare 2009). Voigt-Heucke et al. (2010) found a nonsignificant trend towards group signatures in echolocation pulses of the lesser bulldog bat (Noctilio albiventris).When presented with playbacks of echolocation pulses, N. albiventris reacted with a complex repertoire of social behaviours. Stronger reactions were shown towards echolocation pulses of unfamiliar conspecifics, but much less so to signals of familiar conspecifics from the same social group, than towards heterospecifics and white noise (Voigt-Heucke et al. 2010). This new study provides the first experimental indication of a potential group signature in bat echolocation.
In captive Hipposideros terasensis, changes in echolocation pulse frequency were observed when individuals were added to or removed from a colony (Hiryu et al. 2006). The pulse frequency of each bat showed a long-term gradual change throughout the year, and all bats in the colony increased or decreased frequency in the same direction as a group independent of season. The authors suggest that the audio–vocal feedback for conspecific pulses is involved in the short- or long-term intra-individual variation in pulse frequency. Such social learning of echolocation pulse frequency is consistent with findings that frequency is partly learnt from mothers in greater horseshoe bats (Jones and Ransome 1993). Hipposideros larvatus colonies seem to maintain ‘private bandwidths’ of call frequencies that may facilitate communication among bats from the same colony (Jiang et al. 2010).
Geographic variation in echolocation pulses
Animals often modify their behaviour in relation to familiarity with conspecifics. In some cases, individuals that are not familiar neighbours may be considered as intruders and become the targets of aggressive behaviour, the ‘dear enemy’ phenomenon (Temeles 1994). Conversely, non-local individuals may be favoured as mates, especially if this results in both partners avoiding inbreeding. Potential costs and benefits exist for signallers if call characters reveal geographic identity, and whether area-specific vocalisations (dialects) are favoured by selection will depend on whether the benefits outweigh the costs.
Understanding geographic variation in echolocation signals is difficult because factors such as geographic differences in habitat structure may influence pulses recorded, and differences in perceptual challenges faced by bats in different habitats may mask inherent geographic differences in pulse features within species. Nevertheless, studies on horseshoe bats with relatively stable signals at rest are revealing geographic differences in pulse design. Rhinolophus monoceros shows clinal variation in echolocation pulse frequency in Taiwan, with frequency decreasing with latitude until 24°N, and then increasing further north. A parallel cline in body size was not apparent (Chen et al. 2009). Geographic variation in pulse frequency has also been documented in R. pumilus (Yoshino et al. 2006, 2008), R. ferrumequinum (Rossiter et al. 2007; Flanders et al. 2009), Tadarida brasiliensis (Gillam and McCracken 2007) and Rhinonicteris aurantia (Armstrong and Coles 2007). Whether local adaptation (e.g. adaptations related to how variation in local temperature and humidity affect attenuation) drives frequency variation is still often unclear, although echolocation pulse frequency is related inversely to humidity in Hipposideros ruber (Guillén et al. 2000).
Individual signatures in echolocation pulses
Given that factors such as sex, age, body size and geography can affect pulse characteristics (notably frequency) the potential for individuals having distinctive echolocation pulses certainly exists. Such distinctiveness would be valuable, because it reduces the potential of a bat being confused or jammed by echoes from the pulses of a nearby conspecific. Moreover, reliable individual recognition from echolocation pulses over a distance would likely be beneficial as basis for reciprocal interactions and cooperation within social groups (Siemers and Kerth 2006). So, do bats have individually distinctive echolocation pulses that allow them to identify individuals by acoustic cues alone? Pteronotus parnellii individuals have relatively stable resting CF values, and associated ‘personalised’ tuning of neurons in the auditory cortex (Suga et al. 1987). In many species, intra-individual pulse structure varies substantially in relation to perceptual challenges, and this extensive variability may make it difficult for bats to have consistently individualistic vocal signatures (Siemers and Kerth 2006). Nonetheless, differences in, for example the narrowband tail frequencies in the search phase pulses of molossids may allow bats to identify echoes from their own pulses when small numbers of conspecifics are present (Ulanovsky et al. 2004). Whether bats are able to recognise echoes of their own pulses when flying in large groups remains unknown.
Despite the task-specificity of echolocation pulse structure (e.g. variation caused by flying in open versus cluttered space), there is evidence of individual-specific signatures in echolocation pulses, both from the field (Brigham et al. 1989; Obrist 1995; Fenton 2003; Fenton et al. 2004; Siemers and Kerth 2006; Kazial et al. 2008a) and in the laboratory (Masters et al. 1991, 1995; Burnett et al. 2001; Kazial et al. 2001; Yovel et al. 2009). However, it is questionable whether the parameter space offers enough variability to encode reliable individual signatures among the dozens or even thousands of animals that are found in the colonies of most bat species (Siemers and Kerth 2006).
Classical laboratory observations by Gerhard Neuweiler’s mentor, Möhres (1967) provided the first indication on a behavioural level that bats (Rhinolophus ferrumequinum in this case) might be able to distinguish between individuals from listening to their echolocation pulses. Recent playback studies now provide experimental evidence for individual recognition by listening to echolocation pulses in Myotis lucifugus (Kazial et al. 2008b) and M. myotis (Yovel et al. 2009). Yovel et al. (2009) modelled the bats’ behavioural decisions using nonlinear statistical classifiers. The results suggest that M. myotis bats apply a prototype classification, i.e. they learn the average pulse characteristics of individuals and use these as a reference for individual recognition. Both bats and computer model performed well at classifying pulses of few individuals that were all recorded from flight in confined space. Whether classification performance remains high for natural colonies of several hundred bats and for pulses from both confined and open spaces remains to be tested in future studies. An important outcome from the Yovel et al. (2009) study is that formant-related features, i.e. the distribution of emitted energy over the frequency spectrum and thus ultimately the voice of a bat, are promising candidates for individual signatures that may be stable after taking account of task-specific pulse variation.
Can bats extract information on con- or heterospecific behaviour from echolocation pulses?
Communication and the evolution of species-specific echolocation pulses
Pulse frequency scales negatively with body mass across species, and so larger bats tend to emit lower frequencies (Jones 1999). Hence it could be that ecomorphology (especially body size, or a correlate of body size) drives resource partitioning in bat communities, and pulse frequency patterns arise via a coupling of frequency with size.
Pulse frequency influences echo strength from targets of a given size. This is because the strength of backscatter reaches a maximum when the wavelength of the sound equals or is smaller than object size; for flying insects, typically wing length is the crucial size measure when it comes to echo reflection (Houston et al. 2003). Thus bats that call at higher frequencies (i.e. with shorter wavelengths) are predicted to better detect smaller prey, and pulse frequency may be effective in dietary niche partitioning. Pulse frequency introduces sensory biases in bat perception, so that the same world appears different to different species: with increasing pulse frequency small prey become more perceivable, and large prey relatively less perceivable (because of atmospheric attenuation) to echolocating bats (Safi and Siemers 2010). Although target strength differences may be important in promoting dietary niche separation between species with rather different echolocation pulse frequencies, differences in target strength are negligible for bats emitting frequencies of only a few kilohertz difference, such as cryptic pipistrelles P. pipistrellus and P. pygmaeus (Jones and Barlow 2003). Wavelength differences decrease as frequency increases, and are unlikely to influence prey detection in bats that emit high frequencies for echolocation, such as most rhinolophids and hipposiderids.
However, some bats deviate from allometric expectations, and may emit higher or lower frequencies than expected for their body size. One example of this is Rhinolophus clivosus from southern Africa, which calls at about 92 kHz, though a bat of its size is predicted to call at 56 kHz by allometric scaling (Jacobs et al. 2007). It thus avoids overlap of frequency bands, with sympatric congeners, while the prey sizes exploited by these species do overlap (Jacobs et al. 2007). Therefore, food resource partitioning cannot explain the frequency band separation in these African rhinolophids. In the Asian Rhinolophus macrotis dietary partitioning might be responsible for the deviation of echolocation pulse frequency from allometry, however (Shi et al. 2009). R. macrotis calls at lower frequencies than predicted by allometry, and feeds on larger insects and a wider range of prey sizes compared with the sympatric R. lepidus, which is morphologically similar but which calls at frequencies that fit with allometric predictions (Shi et al. 2009). Alternative hypotheses for divergence in pulse frequency were not fully explored, however. Many cryptic species of bats show minor differences in morphology, but emit echolocation pulses that occupy different frequency bands (e.g. Jones and Van Parijs 1993; Kingston et al. 2001; Thabah et al. 2006). The patterns found in bat communities are complex and there are certainly several selective forces in play. Phylogenetic inertia, allometric scaling and ecological adaptation certainly have their roles, but probably are not sufficient to explain frequency partitioning.
A compelling additional explanation is that the evolution of frequency differences evolved to facilitate intraspecific communication. Having a private bandwidth for communication would explain why frequency differences between species are often abruptly defined. It is interesting to speculate as to how such frequency differences may evolve. Although there are advantages to receivers in being able to locate signallers, what are the advantages to the signaller? There may be advantages to being located and identified, for example in becoming less susceptible to predation when in a group, or if benefits associated with communal roosting accrue. At the same time, there are potential costs to the signaller in being located, especially if competition for scarce prey is increased. Both signaller and receiver benefit from unambiguous recognition of mates, because time otherwise spent in unnecessary mating attempts is saved.
If communication and thus competition for frequency space drives bat species into occupying different frequency bands, then testable predictions can be generated. We can predict that echolocation pulses should become more divergent when similar species are sympatric than when allopatric. Such acoustic character displacement has been documented for the acoustic mating signals emitted by insects (e.g. Shaw 2000), and amphibians (e.g. Gerlinde and Gerhardt 2003).
The use of echolocation pulses for identifying conspecifics has implications for speciation. Echolocation signals may function in promoting reproductive isolation. An extreme example of this might occur in some taxa of horseshoe bats, which may have diverged by ‘harmonic hopping’—shifting the dominant frequency of their echolocation pulses into the higher harmonics of a harmonic series (Kingston and Rossiter 2004). The use of radically different frequencies may allow the exploitation of new dietary niches but also may contribute to assortative mating and reproductive isolation if the echolocation pulses are used in communication. In general, the use of echolocation pulses for communication is perhaps of great importance in locating suitable mating partners and in the maintenance of reproductive isolation among bat species.
Although echolocation pulses evolved primarily for orientation and foraging, they provide public information which can readily be used by other bats from the same, or even from different, species. Growing evidence supports the hypothesis that echolocation pulses can encode information about a bat’s identity, at a range of levels varying from geographical location, through to colony, sex, body size and age. The reliability of such information varies considerably among species, however, and consistent patterns across species are unusual. There is now growing evidence to support the hypothesis that variation in pulse frequency among species within guilds is favoured by natural selection so that ambiguity during intraspecific communication is minimised. Although acoustic character displacement occurred in at least one case study, its generality needs further investigations comparing sympatric and allopatric populations of species emitting similar pulses, and identifying the sources of sympatric populations by using genetic studies. Other topics for future study include further playback studies to determine the nature of discriminations that bats can achieve through information in echolocation pulses, and elucidation of which pulse characteristics are important in discrimination. More rigorously controlled laboratory studies are needed here, but ultimately playback experiments in the field will be necessary to unravel the full significance that echolocation pulses bear for communication both within bat species and across species boundaries. We still need a better understanding if and how individual recognition can operate in large colonies and across the tremendous variation imposed by task-specific calling. Another important area for future research will be to uncover relationships between genetic population structure and the evolution of regional vocal dialects.
We thank Leonie Baier for preparing Fig. 1, Maike Schuchmann, Irek Ruczynski and two anonymous reviewers for comments.
- Bradbury JW, Vehrencamp SL (1998) Principles of animal communication. Sinauer Associates Inc., SunderlandGoogle Scholar
- Burnett SC, Kazial KA, Masters WM (2001) Discriminating individual big brown bat (Eptesicus fuscus) sonar vocalizations in different recording situations. Bioacoustics 11:189–210Google Scholar
- Fenton MB (1985) Communication in the Chiroptera. Indiana University Press, BloomingtonGoogle Scholar
- Gerlinde H, Gerhardt C (2003) Reproductive character displacement in the acoustic communication of green tree frogs (Hyla cinerea). Evolution 57:894–904Google Scholar
- Heller K-G, von Helversen O (1989) Resource partitioning of sonar frequency bands in rhinolophoid bats. Oecologia 80:178–186Google Scholar
- Houston RD, Boonman A, Jones G (2003) Do echolocation signals restrict bats’ choice of prey? In: Thomas JA, Moss CF, Vater M (eds) Echolocation in bats and dolphins. University of Chicago Press, Chicago, pp 339–345Google Scholar
- Jiang T, Liu R, Metzner W, You Y, Li S, Liu S, Feng J (2010) Geographical and individual variation in echolocation calls of the intermediate leaf-nosed bat. Hipposideros larvatus. Ethology 116:691–713Google Scholar
- Jones (1995) Variation in bat echolocation: implications for resource partitioning and communication. Le Rhinolophe 11:53–59Google Scholar
- Jones G, Barlow KE (2003) Cryptic species of echolocating bats. In: Thomas JA, Moss CF, Vater M (eds) Echolocation in bats and dolphins. University of Chicago Press, Chicago, pp 345–349Google Scholar
- Jones G, Sripathi K, Waters DA, Marimuthu G (1994) Individual variation in the echolocation calls of three sympatric Indian hipposiderid bats, and an experimental attempt to jam bat echolocation. Folia Zool 43:347–362Google Scholar
- Liu Y, Feng J, Jiang YL, Wu L, Sun KP (2007) Vocalization development of greater horseshoe bat, Rhinolophus ferrumequinum (Rhinolophidae, Chiroptera). Folia Zool 56:126–136Google Scholar
- Maynard Smith J, Harper D (2004) Animal signals. Oxford University press, OxfordGoogle Scholar
- Möhres FP (1967) Communicative characters of sonar signals in bats. In: Busnel R-G (ed) Animal sonar systems: biology and bionics. NATO Advanced Study Institute, Frascati, pp 939–945Google Scholar
- Ruczynski I, Kalko EKV, Siemers BM (2009) Calls in the forest: a comparative approach to how bats find tree cavities. Ethology 115:167–177Google Scholar
- Siemers BM (2006) Bats: communication by ultrasound. In: Brown K (ed) The encyclopedia of language and linguistics, 2nd edn. Elsevier, Amsterdam, pp 699–704Google Scholar
- Suga N, Niwa H, Taniguchi I, Margoliash D (1987) The personalized auditory cortex of the mustached bat: adaptation for echolocation. J Neurophys 58:643–654Google Scholar
- Thabah A, Rossiter SJ, Kingston T, Zhang S, Parsons S, Mya My K, Zubaid A, Jones G (2006) Genetic divergence and echolocation call frequency in cryptic species of Hipposideros larvatus sensu lato (Chiroptera: Hipposideridae) from the Indo-Malayan region. Biol J Linn Soc 88:119–130CrossRefGoogle Scholar
- Voigt-Heucke SL, Taborsky M, Dechmann DKN (2010) A dual function of echolocation: bats use echolocation calls to identify familiar and unfamiliar individuals. Anim Behav. doi:10.1016/j.anbehav.2010.03.025
- Yoshino H, Matsamura S, Kinjo K, Tamura H, Ota H, Izawa M (2006) Geographical variation in echolocation call and body size of the Okinawan least horseshoe bat, Rhinolophus pumilus (Mammalia: Rhinolophidae), on Okinawa-jima Island, Ryukyu Archipelago, Japan. Zool Sci 23:661–667PubMedCrossRefGoogle Scholar