Integrative and Comparative Biology Advance Access published online on May 14, 2008
Integrative and Comparative Biology, doi:10.1093/icb/icn035
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
Understanding signal design during the pursuit of aerial insects by echolocating bats: tools and applications
*School of Biological Sciences, University of Bristol, Woodland Road, Bristol BS8 1UG; Department of Electronic and Electrical Engineering, University College London, London WC1E 7JE, UK
Correspondence: 1E-mail: Gareth.Jones{at}bristol.ac.uk
 |
SYNOPSIS |
|---|
Bats are among the few predators that can exploit the large quantities of aerial insects active at night. They do this by using echolocation to detect, localize, and classify targets in the dark. Echolocation calls are shaped by natural selection to match ecological challenges. For example, bats flying in open habitats typically emit calls of long duration, with long pulse intervals, shallow frequency modulation, and containing low frequencies—all these are adaptations for long-range detection. As obstacles or prey are approached, call structure changes in predictable ways for several reasons: calls become shorter, thereby reducing overlap between pulse and echo, and calls change in shape in ways that minimize localization errors. At the same time, such changes are believed to support recognition of objects. Echolocation and flight are closely synchronized: we have monitored both features simultaneously by using stereo photogrammetry and videogrammetry, and by acoustic tracking of flight paths. These methods have allowed us to quantify the intensity of signals used by free-living bats, and illustrate systematic changes in signal design in relation to obstacle proximity. We show how signals emitted by aerial feeding bats can be among the most intense airborne sounds in nature. Wideband ambiguity functions developed in the processing of signals produce two-dimensional functions showing trade-offs between resolution of time and velocity, and illustrate costs and benefits associated with Doppler sensitivity and range resolution in echolocation. Remarkably, bats that emit broadband calls can adjust signal design so that Doppler-related overestimation of range compensates for underestimation of range caused by the bat's movement in flight. We show the potential of our methods for understanding interactions between echolocating bats and those prey that have evolved ears that detect bat calls.
|
Introduction |
|---|
Insects that fly at night are potentially a major food resource for predatory animals. Although the total abundance of aerial insects often decreases during the night when temperatures are low, swarming dipteran flies may reach highest abundance around dusk, and many moths (Lepidoptera) are strictly nocturnal (Rydell et al. 1996
). There are few groups of predators that have evolved adaptations for exploiting nocturnal insects as food. Some birds, such as nightjars and some related groups (order Caprimulgiformes, ca. 100 species) have highly developed eyesight that allows them to hunt moths and other large flying insects at night (Sierro et al. 2001). Although large prey may be pursued individually, smaller insects may be captured by trawling while the birds fly with their widely spread gapes open (Martin 1990). The main predators on nocturnal aerial insects are echolocating bats. Vision is of limited use for detecting and localizing small insect prey in the dark. In contrast, echolocation, especially involving the use of ultrasound with small wavelengths, provides the potential for exploiting this rich food source that cannot easily be exploited by visual predators. For a full understanding of aeroecology it is therefore important to consider the adaptations and roles of echolocating bats in ecosystems.
In this article we focus on adaptations: we analyze how the echolocation calls of aerial feeding bats have been shaped by natural selection in ways that facilitate the detection, localization, and sometimes the classification of aerial insects. We show how new advances in acoustic and optical tracking can be used to reconstruct the 3-D positions of echolocating bats, and hence relate signal design to features such as flight speed and proximity to obstacles. Tracking methods have allowed measurements of the remarkable intensity of signals emitted by aerial feeding bats which in turn allows estimating their maximum ranges of echolocation. The details of signal design for in-flight echolocation can be better understood by the application of methods traditionally used in radar and sonar engineering: we illustrate how Wideband ambiguity functions (WAFs) can be used to illustrate how Doppler susceptibility and localization performance depend on signal design. We attempt to explain these methods in a manner intelligible to a wide audience of biologists. We argue that bats can minimize Doppler-dependent errors in localization by adjusting signal design in relation to the proximity of obstacles. Finally, we suggest some future perspectives for the use of three-dimensional localization methods in understanding flight and echolocation behavior in bats. Although bats have evolved sophisticated echolocation that permits exploitation of nocturnal insects, many insects have evolved hearing that is sensitive to ultrasound, thereby allowing them to evade predators (Jones and Rydell 2005). After detecting echolocation calls, many insects undertake elaborate evasive flight maneuvers (Miller and Surlykke 2001). Methods that permit the reconstruction of flight paths allow the potential to better understand the evasive behavior of prey in relation to the intensity of calls emitted by predatory bats, and to produce quantitative descriptions of encounters between bats and their tympanate prey. Throughout our analysis, we illustrate concepts by focusing on the echolocation calls of Eptesicus bottae, an aerial insectivorous species of vespertilionid bat that we studied in the Negev Desert in Israel (Holderied et al. 2005). All methods used have been described previously by Holderied et al. (2005, 2006).
|
Methods for quantifying flight and echolocation behavior |
|---|
TOP
SYNOPSISIntroduction Methods for quantifying flight... Signal design and ecologySignal intensityApplications of processing...Distance of focus (DOF)Future perspectivesAcknowledgmentsReferences |
|---|
To understand how bats change signal design in relation to ecology, ideally it is necessary to determine the position of the bat accurately and to relate the bat's position to surrounding obstacles, while synchronously recording its calling behavior. Successive 3-D reconstructions of position also allow calculation of the bat's flight speed, and provide information on aspects of flight performance such as manoeuvrability, acceleration, and speed. Reconstruction is normally achieved either by optical or acoustical methods (Holderied and Jones in press).
Optical reconstruction can be achieved by stereogrammetry (using information from differences in the viewing angle of two cameras) using still or video images. Although high frame rates can be used to study kinematics in the laboratory with high-speed video cameras (up to 240 Hz—e.g., Ghose and Moss 2006; Moss et al. 2006), lower frame rates (25–60 Hz) are usually sufficient for field studies. Optical 3-D localization of moving targets requires at least two images captured by two cameras at different positions at exactly the same time. Based on the relative positions and orientations of both cameras, the 3-D position of the target at the time the pictures were taken can be inferred by triangulation. Methods include the use of multiple flashguns and still cameras (Jones and Rayner 1988), and the technique is especially powerful when flashgun emissions can be recorded simultaneously with echolocation calls (Kalko and Schnitzler 1993; Britton et al. 1997) so that flight and echolocation behaviors can be linked. Recent advances in digital memory and fast-sampling A/D-cards have increased the portability of equipment for studies of 3-D reconstruction with simultaneous recording of echolocation calls. Video cameras can also be used for optical reconstruction (e.g., Holderied et al. 2005, Fig. 1). CCD video cameras have also been used to visualize flight behavior with stroboscopic infrared illumination used to better "freeze" the motion of the bats (Siemers and Schnitzler 2000). Thermal imaging cameras give great potential for optical tracking in the field without using additional sources of illumination (Hristov et al. 2008). Indeed, thermal imaging and microphone array methods can be combined to reconstruct flight paths (Eastman and Simmons 2005).
|
Acoustical tracking is used to calculate the 3-D position of bats from differences in time of arrival of calls recorded at microphone arrays (Holderied and Jones in press). Although arrays with as few as three microphones can give adequate information to reconstruct flight paths in two dimensions, more sophisticated arrays are necessary for precise 3-D reconstruction. Holderied and von Helversen (2003
) used two arrays of four microphones, each arranged in a symmetric star pattern (Fig. 2), to localize flying bats from their calls, and measured call intensities from 11 aerial feeding species in nature (Fig. 3).
|
|
Which tracking method is used will depend on context. Acoustical tracking methods cover a larger volume of space, facilitate linking flight position to echolocation behavior, allow easy assignation of calls to individuals when several bats are flying together, and does not require additional lighting that might have an impact on the bat's behavior. However, cross-correlations that are necessary for calculating differences in time of arrival are difficult with narrowband signals, and calls of any type need to be of sufficient quality for cross-correlations to be performed at all. Optical tracking can give greater temporal resolution and allows visualization of other objects of interest, such as background clutter and prey. The pros and cons of reconstruction techniques are discussed in more depth by Holderied and Jones (in press). Ultimately, the choice of tracking method will depend on the question that is being addressed, and no method is ideal for all situations.
|
Signal design and ecology |
|---|
Signals emitted by bats are shaped both by environmental factors and by phylogenetic constraints (Jones and Teeling 2006
). Although echolocation calls reveal some exquisite examples of convergent evolution [e.g., the independent evolution of long constant-frequency calls and Doppler-shift compensation in horseshoe bats of the Old World (Rhinolophidae) and Pteronotus parnellii (Mormoopidae) in the New World], some families and genera show some inertia in signal flexibility that appears to be constrained by their evolutionary history. For example, all bats in the Rhinolophidae emit long constant-frequency signals beginning and ending with brief broadband sweeps, and many bat species in the genus Myotis utilize a limited repertoire of broadband sweeps in echolocation: these limitations may be imposed by limited genetic and adaptive potential within clades.
Despite these phylogenetic constraints, patterns of convergent evolution highlight the importance of environmental features in determining signal design. Bat habitats can be defined by the tasks of echolocation, because particular habitats tend to be inhabited by bat species that share common features in call design (Simmons et al. 1979; Schnitzler and Kalko 2001; Schnitzler et al. 2003). Some authors have attempted to define habitats in terms of echolocation behavior: for example, Vespertilio murinus changes signal parameters in relation to distance to background when flying 5–6 m from large objects, but does not change call parameters beyond this range, suggesting that it is not reacting to large targets beyond this threshold of range. Schaub and Schnitzler (2007) argued that this transition in echolocation behavior marks the distinction between "edge" and "open space", although obviously such distinctions will be species-specific because different species have different call designs, and they may also depend on the reflective strength of targets. Aerial feeding bats typically occupy open-space habitats where they hunt for prey far from vegetation and other clutter that creates echoes other than those from the target of interest. Bats can also capture aerial prey in more cluttered situations. For example, some bat species (e.g., pipistrelles Pipistrellus spp.) often forage along the edges of woodlands (Kalko and Schnitzler 1993) and other species may catch aerial prey within heavily vegetated areas (e.g., Myotis nattereri: Siemers and Schnitzler 2000). In general, bats use specific signal designs in given ecological situations, and also change signal design as they move from one habitat to another.
Bats that feed on aerial insects in open spaces catch prey in flight (aerial hawking), and while searching for prey emit calls that are relatively long in duration, have long pulse intervals, incorporate relatively low frequencies, and are often narrow in bandwidth. A typical example is E. bottae (Vespertilionidae) (Figs 4 and 5, left column shows a typical search call), although similar features are shown by aerial feeding bats in the families Molossidae, Rhinopomatidae, and Emballonuridae (Schnitzler and Kalko 2001). Bats foraging in open spaces operate at low duty cycles (i.e., the signal is "on" typically for < 20% of the time), and they detect echoes in the gaps between pulses. When searching for distant targets, pulses can be long because it takes considerable time for their echoes to return, and so masking of the returning echo by the emitted pulse is avoided. Long pulse intervals represent long "listening gaps" for receiving echoes from distant targets, and bats sometimes skip calling for one or more wing beats to extend the pulse interval and hence the listening period (Holderied and von Helversen 2003). Emitted frequencies are usually the lowest found in each species because low frequencies are less subject to excess atmospheric attenuation and travel further in air than do higher frequencies (Lawrence and Simmons 1982). Narrowband pulses are well designed for detection of distant targets (Boonman and Ostwald 2007) and also for classification of targets because they can encode the wingbeats of insects (Schnitzler and Kalko 2001; Schnitzler et al. 2003). Hence, signal design in aerial feeding bats that search open spaces for insects makes adaptive sense from the perspective of sensory ecology.
As the bats approach targets, signal design changes in a predictable manner (Fig. 4). Call sequences during prey capture were initially described by Griffin et al. (1960). Some of the first attempts to understand echolocation behavior during the pursuit of aerial insects from a comparative perspective were made by Simmons et al. (1975, 1979). Calls become shorter as echoes return earlier, such that overlap between pulse and echo is still generally avoided (Kalko and Schnitzler 1993). Pulse intervals become shorter as targets are approached, and the return time of the echo decreases. Calls become more broadband as the task changes from detection to localization, because broadband signals activate more neuronal filters, thereby improving determination of range (Moss and Schnitzler 1995; Boonman and Ostwald 2007). If the bat is attempting to capture an aerial insect, the changes described above result in a feeding (or terminal) buzz (Fig. 4) in which repetition rate of the pulses can reach as high as 200 Hz. Although earlier work concentrated on how bats change pulse duration of the pulse in ways that avoid forward masking of echoes by emitted pulses, we show below how subtleties in the ways that frequency changes over time (call "shape") can also be adaptive.
|
We now consider two areas in which methods of 3-D reconstruction have greatly improved our understanding of the use of echolocation by aerial feeding bats: measuring call intensity and understanding how bats vary call "shape" in relation to the proximity of objects.
|
Signal intensity |
|---|
Some of the most exciting recent studies on bat echolocation have used 3-D reconstruction to measure the intensities of calls produced by free-living bats. Calls recorded from aerial feeding bats have been among the most intense airborne signals recorded from any animal. Holderied and von Helversen (2003
) measured SLs higher than 125 dB peak equivalent sound pressure level (peSPL) at 10 cm from aerial feeding bats, with measurements of 133 dB at 10 cm peSPL for rapidly-flying species in uncluttered space (Fig. 3). The E. bottae (Fig. 4), an aerial feeding desert bat, emits signals as intense as 133 dB peSPL at 10 cm, and may be able to detect large targets such as big moths and conspecifics as far as 20 m distant (although small insects are detected at much closer range) (Holderied et al. 2005) (Fig. 6). Holderied and von Helversen (2003) used call intensity to calculate maximum detection distances for objects varying in size, and proposed that bats match their maximum detection distances for insects and larger flying prey to their preferred gap between calls, which is determined by the period of their wing beats. Ideally, all insect echoes arrive at the bat's ears in the window between the end of call emission and before the emission of the subsequent call, because then bats can unambiguously assign a specific call to an echo. More distant targets may lie outside of the wingbeat window, and bats can overcome problems of incorrect assignment of echoes to calls by alternating call design (Jung et al. 2007), or by extending the length of the wingbeat window by skipping calls, a strategy often used by bats that fly in open spaces (Holderied and von Helversen 2003). |
Applications of processing methods of radar signals |
|---|
As described above, bats vary the duration, pulse interval, and bandwidth of calls in relation to the proximity of obstacles in predictable ways. Subtle changes also occur in the ways that call frequency changes over time as bats approach targets. Our understanding of why bats change the "shape" of calls can be enhanced by applying methods developed by engineers for the analysis of signal performance in radar and sonar.
One of the major issues facing pulse design for flying animals is the Doppler tolerance of the signal, and how Doppler tolerance might relate to localization of objects including prey. Doppler effects experienced by flying bats compress received signals and elevate echo frequencies. With flight speeds typically around 3–8 m/s, echoes of bats’ calls will be subjected to Doppler shifts of 1.8–4.8% in frequency (Boonman et al. 2003). Such Doppler effects are relevant to bats that emit constant frequency signals as well as to bats emitting broadband echolocation calls. WAFs have been used by radar and sonar engineers to better understand the performance of broadband echolocation signals used by bats. Some of the earliest applications of radar/sonar theory for understanding signal design both across and within species were by Cahlander (1966), with subsequent major contributions including those of Altes and Titlebaum (1970), Altes (1984), and Simmons and Stein (1980).
Radar engineers use WAFs to determine how particular waveforms may be best suited for specific applications to radar. WAFs compute resolution in time delay (range) and Doppler factor (velocity) simultaneously and are used in radar and sonar signal processing to show these key characteristics of sensor performance as a function of system (signal) specification. A WAF is generated by cross-correlating two copies of a transmitted signal. One copy of the signal is then given a small Doppler shift and the process is repeated for a range of Doppler shifts (i.e., flight speeds) spanning the range of interest. The resulting cross-correlation functions are plotted as a series of horizontal slices in the WAF. WAFs typically show Doppler effects (i.e., flight speeds) on the y-axis and ranging performance (i.e., delay) on the x-axis. The color of the function at a given point on the plot represents the strength of the receiver output at a particular relative speed and delay. Five examples of WAFs of echolocation calls from E. bottae covering the transition from detection in open space to localization during a terminal buzz are shown in Fig. 5 (3rd row of panels). We have presented the WAFs to include both positive and negative relative flight speeds to facilitate comparisons with radar ambiguity plots that consider targets moving towards and away from the transmitter. For a flying bat the positive relative speeds are by far the most relevant.
|
WAFs are extremely useful tools because they reveal several important features of a signal: on one hand, they show the strength of the received signal and therefore are an indicator of maximum detection. On the other hand, they visualize the two Doppler-related errors in ranging performance (Holderied et al. 2006
). The first of these Doppler-related errors affects the accuracy of the range measurement. This ranging error affects localization. For downward sweeps and positive Doppler shifts it causes an overestimation of the actual range. This error increases with flight speed and is therefore called range-Doppler coupling. In the WAF, this is represented by the steepness of the contour. The steeper the contour the smaller is the range-Doppler coupling. In contrast, the other Doppler-related error affects the ranging acuity, which can be thought of as the precision or resolution of the ranging. A reduced ranging acuity is not only detrimental to localization but also to resolving details in complex echoes and thus affects recognition of objects. Acuity is negatively related to the width of the envelope of the cross-correlation function envelope and in the WAF it is depicted as the horizontal width of the contour. This width often increases with flight speed, indicating reduced acuity in ranging.
A signal that minimizes range-Doppler coupling errors would be narrow along the range axis, which shows that there is minimal change in range with change in flight speed (no range-Doppler coupling). In addition, a signal that is narrow in range at all flight speeds is an indicator of high-ranging acuity and is termed "Doppler-tolerant" in radar research. A doubly Doppler-tolerant signal will thus have a vertical narrow contour indicating that ranging is little affected by velocity in either its accuracy or acuity. The signal design that is least sensitive to Doppler errors in ranging acuity (but not in ranging accuracy!) is one that shows hyperbolic frequency modulation (Kroszczynski 1969). Conversely, a signal that best detects (i.e., most strongly responds to) Doppler shifts would show a horizontal contour with minimal ambiguity about velocity. Long, constant-frequency signals have excellent potential for detecting Doppler shifts and can have such contour plots. However, the ability of constant-frequency signals to measure range reliably is poor.
We will now show how this approach can be used to quantify Doppler tolerance and localization performance in a range of bats’ call designs. Of special interest is the change in call design used by bats as they approach targets. During these "feeding buzzes" the bat might change its call design from one that gives strong receiver responses for easy detection and at the same time allow recognition of objects based on prey flutter, to one that optimizes localization. This is because the task changes from detection to localization during a feeding event, and because with closer approach accurate ranging becomes more important so calls need to be less affected by Doppler errors.
The third row of Fig. 5 exemplifies changes in the WAF of five calls ranging from search phase, approach phase, to feeding buzz (from left to right). The following trends can be seen: first, a decrease in the maximum value in the WAF, which shows that calls of the search and approach phases elicit higher responses in the receiver and are therefore better suited for detection than are feeding buzzes. Note that for this analysis amplitudes of the call have been normalized, so this effect is not due to the lower intensities of buzz calls but rather due to their reduced duration. Second, range-Doppler coupling (indicated by the inverse steepness of the WAF contour) is higher in calls in the search and approach phases and gradually decreases towards the feeding buzz. True buzz calls have a very small range-Doppler coupling and are thus largely tolerant to this sort of ranging error. Third, ranging acuity (measured as the width of the WAF contour) is also decreasing during the approach. The narrowest contours are found in the buzz calls. Thus, as the bat approached its prey, there was a steady increase both in ranging accuracy [from 2.6 ms (45 cm) to 17 µs (3 mm)] and ranging acuity [between 1969 µs (33.8 cm) and 71 µs (1.2 cm); both calculated for a speed of 6 m/s]. The advantage of the WAF is that it shows that this is true for all potential relative flight speeds between bat and moth up to at least 10 m/s. The use of WAFs has been fundamental in understanding how ranging performance in echolocation is dependent on Doppler-tolerance, which in turn is determined by call shape.
|
Distance of focus (DOF) |
|---|
Although analyses using WAFs show how Doppler shifts may cause an overestimation of the range to the target due to the inevitable range-Doppler coupling, bats also experience another ranging error during flight. A bat approaches a target in the time between calling and receiving the echo. Because the distance that the bat flies reduces the distance that sound travels, the delay of the echo is shortened. Consequently, when the echo is received, the target's range is closer by half the distance flown than it was at the time of calling. This underestimation of the range to the target becomes more marked with increasing flight speed, and with distance to the target.
The overall ranging error a flying bat experiences is the superposition of both these errors and they might mutually cancel. In theory, bats could adjust signal design so that the Doppler-related range overestimation exactly compensates for the range underestimation caused by the bat's movement in flight. The range at which the errors mutually cancel has been termed the DOF (Boonman et al. 2003; Holderied et al. 2006). Our analyses assume that the bat relates target distance to the position it was at when it emitted the call. If the bat relates the distance to the position it has when it receives the echo, it would further overestimate range, thus compounding the Doppler-related error.
The DOF forms a sphere around the bat, (Fig. 5; fourth row of panels), and its diameter critically depends on signal design. Steeper, more broadband calls have a short DOF, while more narrowband calls of longer duration have a greater DOF. The DOF strategy is computationally efficient: the bat needs to be aware of which signal type focuses to which distance, and roughly match this to the range determined by the preceding call. In this way, the bat can achieve high ranging accuracy that compensates for inevitable ranging errors related to its flight speed.
Field studies using acoustic tracking show that whiskered bats Myotis mystacinus modify call design in accordance with predictions of the DOF theory. Calls that resulted in more distant DOF values were emitted when bats flew further from a hedge. The bats adjusted call design gradually in a range-dependent manner to minimize localization errors at the distance of the target of interest. In general, there was a very close match between the predicted DOF and the distance to the target. The DOF underestimated the distance to the hedge slightly. Such an underestimation makes the bat keep a greater distance than minimally necessary and might therefore decrease the risk of collision (Holderied et al. 2006). However, we would not expect a bat to estimate range incorrectly when capturing a prey item.
In Fig. 6, we present the first DOF analysis of a free-ranging bat attacking a moth. In parallel with a decrease in SLs and duration of calls and of calling intervals, there are also systematic changes in the DOF. Starting at about 15 m, DOF gradually decreases as the bat approaches the moth, finally being close to the actual bat-moth distance during the feeding buzz. The overall ranging error will thus be efficiently reduced during the attack from an underestimation of 21 cm during early approach to a probably inconsequential overestimation of 0.2–0.4 cm at the feeding buzz. While there is no close match between DOF and distance between bat and moth at greater distances (when the bat also needs to be aware of the proximity of more distant elements of the landscape), the general changes in signal design are clearly in agreement with the predictions of the DOF analysis. At least in the final stages of the approach there also is a close quantitative match.
|
Earlier (and not mutually exclusive) hypotheses consider how bats change pulse duration, pulse repetition rate, and pulse interval during pursuit (Kalko and Schnitzler 1993
), but only through the application of WAF and DOF approaches have we understood why the shape and bandwidth of the signal may change, and how echolocative performance is strongly affected by flight speed. Moreover, DOF theory generates quantitative predictions about how bats should adjust signal design (in particular bandwidth) if they aim to minimize errors in localization. |
Future perspectives |
|---|
Although bats have evolved sophisticated echolocation signals that allow them to feed on aerial insects in the dark, many insects have evolved defenses that reduce their risk of being captured by echolocating bats. Hearing has evolved independently at least 19 times in insects, and strategies of avoiding bats are now known, or potentially likely, in five orders of insects: moths and nocturnal butterflies (Lepidoptera), crickets and locusts (Orthoptera), praying mantids (Dictyoptera), beetles (Coleoptera), and flies (Diptera) (Miller and Surlykke 2001
; Jones and Rydell 2005). Optical tracking gives great potential for quantifying the reactions of tympanate insects to echolocating bats, and to determine whether avoidance strategies relate to intensity of the signal. Some preliminary results for E. bottae interacting with a moth (Fig. 7) show a moth reacting at only 3–4 m by flying higher. The moth then performed an abrupt fast downward loop just before the attacking bat would have caught it. The moth returned to its initial height, flying in the opposite direction from the bat. On this occasion the moth outmaneuvered the bat with minimum effort in a near-bat reaction involving two distinct stages.
|
Calling behavior of the bat and flight behavior of bat and moth are linked in Figs 5–7
. This allows evaluation of the timing of the evasive response with respect to the perceived echolocation behavior. The increase in height of flying is initiated when the bat is 3 m away and still in the earlier stages of the approach. The swift downward evasive flight, however, sets in only about 200 ms before the bat starts the final buzz. Moth flight speed is highest and distance from bat to moth is closest right in the middle of the feeding buzz. The outmaneuvered bat aborts the buzz and gradually returns to calls that are more like those in the approach phase. The bat did not attempt an immediate second attack on this moth. Interestingly, this very efficient evasive maneuver of the moth shows striking similarities with an aircraft combat maneuver known as High Yo-Yo (Shaw 1985). The use of optical and acoustical tracking will be important methods in the future of research on bat echolocation. Their potential for giving insight into call intensity and signal design is already apparent for bats in the family Vespertilionidae. It will be interesting to determine the relevance of DOF in understanding signal structure in relation to proximity of the target in other bat families that use a diverse range of signal types. Tracking methods allow calls to be located accurately in 3-D space, and changes in call design can be related to factors such as flight speed and the proximity of obstacles. Visualizations of the reactions of insects to echolocating bats offer great potential for understanding how escape mechanisms have evolved in insects in a fascinating predator-prey interaction.
|
Acknowledgments |
|---|
We are grateful to TH Kunz and NI Hristov for inviting GJ to participate in this symposium. We also wish to thank the Society for Integrative and Comparative Biology for waiving registration fees and for their support in hosting the symposium. Field work in Israel was performed with assistance from Carmi Korine, Berry Pinshow, and Brock Fenton. We thank the Air Force Office of Scientific Research, through a grant to Boston University (FA9550-7-1-0449 to TH Kunz), for providing partial travel support to GJ to participate in the Aeroecology symposium. We also wish to thank BBSRC (grant BB/F002386/1) for funding our research. Studies on radar applications were also funded by SEAS DTC.
|
FOOTNOTES |
|---|
From the symposium "Aeroecology" presented at the annual meeting of the Society for Integrative and Comparative Biology, January 2–6, 2008, at San Antonio, Texas.
|
References |
|---|
Altes RA. Echolocation as seen from the viewpoint of radar/sonar theory. In: Localization and orientation in biology and engineering—Varjú D, Schnitzler H-U, eds. (1984) Heidelberg (Germany): Springer. 234–44.
Altes RA, Titlebaum EL. Bat signals as optimally Doppler tolerant waveforms. J Acoust Soc Am (1970) 48::1014–20.[CrossRef][Web of Science]
Boonman AM, Ostwald J. A modeling approach to explain pulse design in bats. Biol Cybern (2007) 97::159–72.[CrossRef][Web of Science][Medline]
Boonman AM, Parsons S, Jones G. The influence of flight speed on the ranging performance of bats using frequency modulated echolocation pulses. J Acoust Soc Am (2003) 113::617–28.[CrossRef][Web of Science][Medline]
Britton ARC, Jones G, Rayner JMV, Boonman AM, Verboom B. Flight performance, echolocation and foraging behaviour in pond bats, Myotis dasycneme (Chiroptera: Vespertilionidae). J Zool Lond (1997) 241::503–22.
Cahlander D. Discussion. In: Les Systèmes Sonar Animaux. Biologie et Bionique—Busnel R-G, ed. (1966) II:. Jouy-en-Josas, France: INRA-CNRZ. 1052–81.
Eastman KM, Simmons JA. A method of flight path and chirp pattern reconstruction for multiple flying bats. Acoust Res Lett Online (2005) 6::257–62.[CrossRef][Web of Science]
Ghose K, Moss CF. Steering by hearing: a bat's acoustic gaze is linked to its flight motor output by a delayed, adaptive linear law. J Neurosci (2006) 26::1704–10.
Griffin DR, Webster FA, Michael CR. The echolocation of flying insects by bats. Anim Behav (1960) 8::141–54.[CrossRef]
Holderied MW, Jones G. Flight dynamics of bats. In: Ecological and behavioral methods for the study of bats—Kunz TH, Parsons S, eds. Baltimore (MD): Johns Hopkins Press. (in press).
Holderied MW, Jones G, von Helversen O. Flight and echolocation behaviour of whiskered bats commuting along a hedgerow: range-dependent sonar signal design, Doppler tolerance and evidence for ‘acoustic focussing’. J Exp Biol (2006) 209::1816–26.
Holderied MW, Korine C, Fenton MB, Parsons S, Robson S, Jones G. Echolocation call intensity in the aerial hawking bat Eptesicus bottae (Vespertilionidae) studied using stereo videogrammetry. J Exp Biol (2005) 208::1321–7.
Holderied MW, von Helversen O. Echolocation range and wingbeat period match in aerial-hawking bats. Proc R Soc Lond B (2003) 270::2293–9.[Medline]
Hristov NL, Betke M, Kunz TH. Applications of thermal infrared imaging for research in aeroecology. In: Integr Comp Biol (2008) doi:10.1093/icb/icn037.
Jones G, Rayner JMV. Flight performance, foraging tactics and echolocation in free-living Daubenton's bats Myotis daubentoni (Chiroptera: Vespertilionidae). J Zool Lond (1988) 215::113–32.
Jones G, Rydell J. Attack and defense: interactions between echolocating bats and their insect prey. In: Bat ecology—Kunz TH, Fenton MB, eds. (2005) Chicago (IL): Chicago University Press. 301–45.
Jones G, Teeling EC. The evolution of echolocation in bats. Trends Ecol Evol (2006) 21::149–56.[CrossRef][Medline]
Jung K, Kalko EVV, von Helversen O. Echolocation calls in Central American emballonurid bats: signal design and call frequency alternation. J Zool Lond (2007) 272::125–37.[CrossRef]
Kalko EKV, Schnitzler H-U. Plasticity in echolocation signals of European pipistrelle bats in search flight – implications for habitat use and prey detection. Behav Ecol Sociobiol (1993) 33::415–28.[Web of Science]
Kroszczynski JJ. Pulse compression by means of linear period modulation. Proc IEEE (1969) 57::1260–6.
Lawrence BD, Simmons JA. Measurements of atmospheric attenuation at ultrasonic frequencies and the significance for echolocation by bats. J Acoust Soc Am (1982) 71::585–90.[CrossRef][Web of Science][Medline]
Martin G. Birds by night (1990) London: Poyser.
Miller L, Surlykke A. How some insects detect and avoid being eaten by bats: tactics and countertactics of prey and predator. Bioscience (2001) 51::570–81.[CrossRef][Web of Science]
Moss CF, Bohn K, Gilkenson H, Surlykke A. Active listening for spatial orientation in a complex auditory scene. In: PLoS Biol 4:e79 (2006).
Moss CF, Schnitzler H-U. Behavioral studies of auditory information processing. In: Hearing by bats—Popper AN, Fay RR, eds. (1995) New York: Springer. 87–145.
Rydell J, Entwistle A, Racey PA. Timing of foraging flights of three species of bats in relation to insect activity and predation risk. Oikos (1996) 76::243–52.[CrossRef][Web of Science]
Schaub A, Schnitzler H-U. Echolocation behavior of the bat Vespertilio murinus reveals the border between the habitat types "edge" and "open space". Behav Ecol Sociobiol (2007) 193::1185–94.
Schnitzler H-U, Kalko EKV. Echolocation by insect-eating bats. Bioscience (2001) 51::557–69.[CrossRef][Web of Science]
Schnitzler H-U, Moss CF, Denzinger A. From spatial orientation to food acquisition in echolocating bats. Trends Ecol Evol (2003) 18::386–94.[CrossRef]
Shaw RL. Fighter combat: tactics and maneuvering (1985) Annapolis (MD): United States Naval Institute Press.
Siemers BM, Schnitzler H-U. Natterer's bat (Myotis nattereri Kuhl, 1818) hawks for prey close to vegetation using echolocation signals of very broad bandwidth. Behav Ecol Sociobiol (2000) 47::400–12.[CrossRef][Web of Science]
Sierro A, Arlettaz R, Naef-Daenzer B, Strebel S, Zbinden N. Habitat use and foraging ecology of the nightjar (Caprimulgus europaeus) in the Swiss Alps: towards a conservation scheme. Biol Conserv (2001) 98::325–31.[CrossRef]
Simmons JA, Fenton MB, O’Farrell MJ. Echolocation and pursuit of prey by bats. Science (1979) 203::16–21.
Simmons JA, Howell DJ, Suga N. Information content of bat sonar echoes. Am Sci (1975) 63::204–15.[Web of Science][Medline]
Simmons JA, Stein RA. Acoustic imaging in bat sonar: echolocation signals and the evolution of echolocation. J Comp Physiol (1980) 135::61–84.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||