The echolocation and hunting behavior of the bat - CiteSeerX

Journal of Comparative Physiology A

J Comp Physiol A (1987) 161:267-274

Sensory,

.~,,

and Behavioral Physiology

9 Springer-Verlag 1987

The echolocation and hunting behavior of the bat,

Pipistrellus kuhfi Hans-Ulrich Schnitzler 1, Elisabeth Kalko 1, Lee Miller 2, and Annemarie Surlykke 2 1 Lehrbereich Zoophysiologie, Universitfit Tfibingen, D-7400 T/ibingen, Federal Republic of Germany 2 Institute of Biology, Odense University, DK-5230 Odense, Denmark Accepted March 12, 1987

Summary. The echolocation and hunting behavior of Pipistrellus kuhli was studied in the field using multi-exposure photography synchronized with high-speed tape recordings. During the search phase, the bats used 8-12 ms signals with sweeps (sweep width 3-6 kHz) and pulse intervals near 100 ms or less often near 200 ms (Figs. I and 2). The bats seemed to have individual terminal frequencies that could lie between 35 and 40 kHz. The duty cycle of searching signals was about 8%. The flight speed of hunting bats was between 4.0 and 4.5 m/s. The bats reacted to insect prey at distances of about 70 to 120 cm. Given the flight speed, the detection distance was estimated to about 110 to 160 cm. Following detection the bat went into the approach phase where the FM sweep steepened (to about 60 kHz bandwidth) and the repetition rate increased (to about 30 Hz). The terminal phase or 'buzz', which indicates prey capture (or attempted capture), was composed of two sections. The first section contained signals similar to those in the approach phase except that the pulse duration decreased and the repetition rate increased. The second section was characterized by a sharp drop in the terminal frequency (to about 20 kHz) and by very short pulses (0.3 ms) at rates of up to 200 Hz (Figs. 1 and 3). Near the beginning of the buzz the bat prepared for capturing the prey by extending the wings and forming a tail pouch (Fig. 4). A pause of about 100 ms in sound emission after the buzz indicated a successful capture (Fig. 4). Pulse duration is discussed in relation to glint detection and detection distance. It is argued that the minimum detection distance can be estimated from the pulse duration as the distance where pulse-echo overlap is avoided. Abbreviations: CF constant frequency, F M frequency modu-

lated

Introduction

The echolocation behavior of bats hunting for insects has been described by several authors (for reviews see Novick 1977; Simmons et al. 1979; Schnitzler and Henson 1980; Pye 1980; Neuweiler 1983). All studies indicate that the three behavioral stages: search, approach and terminal phases used by Griffin et al. (1960) to describe the pulse patterns of vespertilionids hunting insects can be generalized for all bats. Studies on hunting behavior and capture techniques in bats suggest that the capture techniques described by Webster and Griffin (1962) such as mouth, tail membrane, wing scoop and somersault catches are representative of most bats. Most studies describe either echolocation behavior or hunting behavior of bats. Griffin et al. (1960), Webster (1963), Trappe (1982), Trappe and Schnitzler (1982), Vogler and Neuweiler (/983) studied both behaviors at the same time using synchronous recordings of the echolocation signals and filming or photographing hunting bats in the laboratory. Corresponding field studies are lacking. The development of a battery operated 6 flash strobe system synchronized with our high speed tape recorder made it possible for us to undertake field studies of foraging bats. The equipment is bulky and cannot be moved during a recording session. Therefore we had to find a bat species that reliably hunted in an open area where we had a chance to record the sounds and photograph the bats. Pipistrellus kuhli, one of the four European pipistrelle species, fulfilled all these conditions. In Yugoslavia we found a site where bats hunted for insects near a street light. We could make nearly 500 photographs from passing and hunting bats. The data of ten nights are presented in this paper. The echolocation and hunting behavior of this species has not yet been described.

H.-U. Schnitzler et al. : Echolocation and hunting behavior in Pipistrellus

268

A SPL

[

I

.

D,

tl

9

kHz 100 I OL-

,@ SPL

[

~,,,.

o

o

t

,

I

,J

I

I

t

t

9

kHz 100 F 0 z-

SPL

E

~- t

I

,

~

; : :::'7' .....

......i .............i ..........i ..........i

li[iiiiii i;

"7 ....

,

t

t

I

I

,,.z loo[

iiiiii,,ii ....................... i .............. i ...............i ................ i ...............................[ ................i ....

0 L.

~'

El

"

......................... ::::

I

I

0

100

I[I

I

I

200

300

_

400

t (mS)

I

....................................... I EI E:-;

:

:

;

;

I

................... ,..........:...........:...................... .........,.................... I I 0

I 10

| 20

t (ms)

Fig. 1 A. Hunting sequence of echolocation signals in Pipistrellus kuhli showing the search, approach and terminal phase or buzz. In this figure and others an arrow indicates the transition from search to approach phase. The lower four panels show details of a pulse from the search phase (1), the first pulse from the approach phase (2), two pulses from the first section of the terminal phase (3), and five pulses from the second section of the terminal phase (4). SPL relative Sound Pressure Level; kHz frequency

Methods Recording site and animals. The sound recordings and the photographs of P. kuhli hunting for insects were made under a street light in the village of Gornije Okrug near Trogir in the province Dalmatia, Yugoslavia. This recording site had the advantage that several bats regularly hunted near the street light where they could be photographed against the dark background of the open sky. Most recordings were made between 2300 h and 0100 h. To increase our chances of photographing hunting bats we sometimes tethered a moth on a string attached to a long pole. The fluttering or moving insect attracted bats into the study site. Unfortunately, we did not catch a bat. Nevertheless we are sure that we studied P. kuhli, because the bats produced the same signal type as a P. kuhli captured and identified near the study site the year before and because color photos of a flying bat showed the diagnostic white stripe along the caudal margin of the wing. Sound recording apparatus. Bat sounds were recorded on two channels of a battery powered magnetic tape recorder (Lennartz electronic 6000/607, speed 76.2 cm/s, 1/4 inch tape) following transduction and amplification using the microphone of a QMC Bat Detector (model $100) and an electrostatic microphone of a custom made detector (Miller and Andersen 1984). The

cutoff frequency of the whole system was above 100 kHz. The bat sounds were made audible with the divide by ten system of the custom made detector. These signals were stored on the third channel of the tape recorder. The synchronization pulses from the 6 flash strobe system were recorded on the fourth (FM) channel.

Analysis of sound recordings. For a detailed analysis of frequency structure, sound duration, repetition rate and duty cycle we used sonagram-like real time spectrograms, which were produced by the following method. Sound sequences were played into a Nicolet U A 500 real time spectrum analyzer (range 10 kHz, terminated at 2 kHz) normally at 1/64th of the recording speed. At this setting a 50 ms time window for spectral analysis is moved in 10 ms steps giving consecutive spectra which contain 20% new information. Considering the speed reduction factor, this corresponds to a real frequency range of 128 kHz, a frequency resolution of 1280 Hz and a time difference of 0.16 ms between consecutive spectra. The spectra were displayed on an oscilloscope screen with frequency on the y-axis and amplitude on the z-axis. We photographed the spectra on 60 mm wide photographic paper running usually at 5 mm/s (T6nnies oscilloscope camera). Using this method we obtained continuous, sonagram-like real time spectrograms with a dynamic range of about 40 dB. The synchronization pulses were

H.-U. Schnitzler et al. : Echolocation and hunting behavior in Pipistrellus

269

B 1" 200

.-~ 1 5 0

9

1OO

i/i/i/\

E

x~

--x

t

50

".-~[

ix/

200

--

150

"=-

100

K

==

50

Y

.

~

9 9 ,.........

, ~

.----~-.--___.

E 10

8

. 6

$

~./\./

._/./'J"

.... \.~._.

,I

0

/

\

2 I

I

I

I

I

I

I

t

I

500

1000

1500

lud

I

,8 f 16 "o 12

8

4

Fig. 1B. The signal parameters of the 2000

sequence shown in A

I 2500 t(ms)

displayed on the second beam of the oscilloscope thus producing time marks on the spectrograms which indicated trigger impulses to each of the 6 stroboscopic flashes.

Photographing of hunting bats. A Nikon F2 35 mm single lens reflex camera with a zoom lens (80-200 mm) was aimed at an area where bats were likely to hunt for insects. When a bat appeared in this area the shutter was opened foor 1 s and this triggered a sequence of 6 flashes (Metz Mecablitz 45CT3 with long lens adapter 45-33). The interval between the first and the second flash was 200 ms. All other intervals were set at 100 ms. For each flash a synchronization pulse was produced and recorded on the tape. For most of the photographs we used the following settings: Focus 8-10m, F-stop 8 or 5.6, Zoom lens 80 ram, flash duration 1 or 0.4 ms. We used Kodak Tri-X Pan black and white film (400 ASA) developed with Microphen (4 rain at 25 ~

Data basis. Our results are based on data from sonagrams of 58 sound sequences like those in Fig. 1A and Fig. 3. From these sequences 30 were synchronized with multiexposure photographs. The photographs were chosen from 351 successful photographs. For more accurate measurements of sound duration we displayed the oscillograms and the sonagrams on an expanded time scale like those shown in the insets 1-4 in

Fig. 1 A. Frequency measurements were made of the sonagrams. The accuracy of these measurements is influenced by many factors. The spectrum analysis allows a frequency resolution of only 1280 Hz. The dynamic amplitude range of the sonagrams never exceeds 40 dB, thus it cannot show signal elements below this range. The directionality of sound emission in the bats, the directionality of the microphones, the frequency dependent atmospheric attenuation, and the frequency response of our recording apparatus are responsible for frequency dependent amplitude changes which also influence the sweep ranges displayed in the sonagrams. The frequency measurements are therefore not precise. They should be regarded as approximations to the real values. A statistical treatment of the frequency data obtained from the sonagrams could lead to inappropriate conclusions due to the many uncertainties to sound recording in the field.

Results

Echolocation behavior of hunting bats The three behavioral categories, search, approach a n d t e r m i n a l p h a s e u s e d b y G r i f f i n et a l ( 1 9 6 0 )

270

H.-U. Schnitzler et al. : Echolocation and hunting behavior in Pipistrellus

to describe the pulse pattern of Myotis lucifugus (Vespertilionidae) hunting for insects can also be used to describe the echolocation behavior of hunting Pipistrellus kuhli. Figure 1 A and B give an example of a hunting sequence and an analysis of sound parameters. In search flight, high flying bats (probably above 8 m) produce shallow F M sweeps with a duration of 8-12 ms (sweep width 3-=6 kHz). When the bats approached the study site, the sound duration was shortened to 8-6 ms and an initial downward sweep was added to the shallow sweep which increased the sweep width to about 35 kHz. The terminal frequency of the search pulses is kept rather constant through most of the sequence and seems to be characteristic for each individual bat. For example, in Fig. 1 A the terminal frequency is maintained at 3 6 k H z with little variation throughout the search, approach and first part of the terminal phase. The terminal frequencies of the bats recorded were mostly between 35-40 kHz. The transition from the shallower to the steep sweep can occur rapidly, as shown in Fig. 31 or more gradually as seen in the first five pulses in panel 2 of Fig. 1 A. We think that searching pulses with greater sweep widths are always emitted as the bats come closer to obstacles (like the street lamp) or closer to the ground. The F M part improves the ability for the bat to localize the obstacles in its flight path. The time intervals between searching signals are either around 100 ms or around 200 ms (Figs. 1 and 2) and on the average, every fourth interval is long, suggesting that the bat skipped one cycle. Consequently there can be about 8 pulses per 1000 ms giving a duty cycle of from 6% to 10% (Fig. 1B). (The duty cycle is the fraction of time filled with sound.) The approach phase begins as we define it with the first pulse that is emitted in reaction to a detected target as indicated by a distinct reduction of the pulse interval relative to the intervals in the search phase (Fig. 1 A and B). The start of the approach phase is marked by an arrow in all figures except Fig. 2. During the approach phase, P. kuhli shortens the pulse interval from about 70 ms to about 35 ms while the pulse duration is shortened from about 7 ms to about 3 ms. The sweep width is increased and a second harmonic is often introduced. In the fully developed F M signal of the approach phase the first harmonic starts at about 95 kHz and sweeps down to the terminal frequency. The duty cycle is kept at about 8% throughout the whole approach phase. The terminalphase or buzz is a group of short pulses emitted at a high rate just prior to a successful or attempted insect catch. In P. kuhli the termi-

50

i

N:217 40

30

20

10

50

100

150

200

250

300

350

pulse interval (ms)

Fig. 2. Distribution of pulse intervals during searching sequences of Pipistrellus kuhli. Note that about a fourth of the pulse intervals are long and grouped around 200 ms

nal phase can be separated into two sections (Fig. 1 A panel 3). In the first section the sweep width and terminal frequency is similar to that of pulses in the approach phase. In the second section the terminal frequency drops to values lying mostly between 19 and 24 kHz with sweep widths of between 20 and 40 kHz. In both sections a second harmonic may be present. In the first section the pulse duration decreases from about 2 ms to about 1 ms and the pulse interval is reduced from about 18 ms to 6 ms. In the second section the pulse duration is constant at about 0.3 ms with pulse intervals between 5 ms and 6 ms. Approach and terminal phase are highly variable. Some sequences in Fig. 3 illustrate this point. The approach phase can have as few as three pulses and be as short as 100 ms or contain 22 pulses and exceed 1000 ms (Fig. 3 E). Usually a long interval separates the last pulse of the approach phase from the first pulse of the terminal phase (Fig. 3 A). However, a smooth transition from one phase to the next is also possible (Fig. 3 E). The first section of the terminal phase can be as short as 50 ms and contain as few as 7 pulses. On the other hand, a bat can interrupt the first section of the terminal phase (probably to take a breath) and extending it to as much as 31 pulses emitted in 375 ms (Fig. 3G). The second section of the terminal phase can be very long (about 200 ms) and contain as many as 34 signals (Fig. 3 H). It can be as short as 40 ms and contain 8 pulses (Fig. 3I) or be dropped completely (Fig. 3 J). The latter situation suggests that the prey was missed. The pause after the terminal phase can vary between 50 and 212 ms and in documented successful attacks the pause

H.-U. Schnitzler et al. : Echolocation and hunting behavior in Pipistrellus

271

kHz 100 f

~

0

t

i

9

100 f

~

k

I

i I l[llll~l,

,t

I

i

L $,

0

'~176i ' ,, ,

I

li~il

L f

1

i

I

t

I

IL

i I [

L

k

I

i

I i

i

I

l

=

i

i

i

i

t

J

i

i

f

I

i l t L 1Illil I lita%L~,,,,,,, i

100~

L

I II~

I I t liul%,,,,.

I

vl'

L

I

t

0

I

I

t lllItIIt0,~l,,,,,,,,

i

i

1

L

100~

I

1LIlIlllltllltl,l~,,, i

i

1

,

I

t Itll II~llll~llii~lllllll~, I

k

0

I

t

i

i

hiliHi

0

L

t

,m ......... ,.................

I

,

,t

i

t

,

I -800

I

I

b i b, i ..................

~1000

i I

i,,,,.

I

t

t ( LLLllLlllllllil(

L

-1200

; t

I

t

i

i

i

-600

-400

-200

0

200

400 t (ms)

Fig. 3. Variation m approach and terminal phases during the pursuit of insects by Pipistrellus kuhli

|

|

|

Fig. 4. P. kuhli pursuing and catching an insect at the study site.

The figure is redrawn from the negative and shows five positions of the bat. The sound sequence is given in the lower two panels.

Numbers indicate synchronizations between the sound sequence and

thestroboscopic flashes. The time

1oo~ kHz

between consecutive flashes is 100 ms except that between flashes 1 and 2, which is 200 ms. The insect (not identified) was probably caught just prior to flash 5. (Sounds from a second bat seen in the background have been removed from the sequence.)

OE t

kHz100 f Or-

l l t l till 1

@

|

I

(

((I |

I

|

I

|

I

I

0

was at least 100 ms (Fig. 4). A shorter pause suggests an unsuccessful attack. The duration o f approach and terminal phase varied between 1015 ms (Fig. 3 H) and 247 ms indicating large differences in the duration o f pursuits.

I

100

200 t(ms)

Flight speed By measuring the number o f body lengths a bat moved between two flashes we calculated the flight speed o f the bats. Based on a series o f 12 photo-

272

H.-U. Schnitzler et al. : Echolocation and hunting behavior in Pipistrellus

graphs and at an assumed body length of 45 mm we determined speeds of between 3.5 and 5.3 m/s with most values lying between 4.0 and 4.5 m/s for straight level flight. When turning, the bats slowed to speeds of 1.8 to 3.8 m/s. The photographs of insect catches demonstrate that in some pursuits the bats maintained a speed of up to 2 m/s throughout the whole terminal phase but in other cases, especially when they had to make sharp turns, the bats nearly came to a halt when they were near the insect.

Hunting behavior and capture technique In search flight the bats passed the illuminated area near the street light with straight flight paths and disappeared into the darkness. The flight height was mostly 2 to 3 m above the ground, but sometimes above the 5 m high street light. The reaction to an insect was sometimes indicated by a distinct turn suggesting that the bats must have a rather wide search cone. One of our photographs shows a bat detecting an insect that was flying far to the side of the bat's flight path so this bat had to make a sharp turn to reach it. Our data compares favorably with those of Griffin et al. (1960) who estimated that Myotis lucifugus is able to detect fruit flies within a spatial cone of about 120 ~ in front of the head. The distance between the bat and the insect at the point of first reaction is difficult to judge, but using three suitable photos we estimated that the first reaction to the insect occurred at a range of 70 and 120 cm. During the approach the bat's head was pointed at the insect. When coming close to the insect, the bat's body was tilted upwards, the wings were extended and the interfemoral membrane was brought foreward and cupped. The insect is probably caught in the pouch formed by the interfemoral membrane. The capture ended when the bat bended its head forward into the tail membrane probably to retrieve the insect. Afterwards the bat straightened out and returned to search flight with the body oriented in the direction of flight. Some of the photographs show that the bats are highly maneuvrable, making sharp turns during a pursuit. In on photograph we observed a 180 ~ turn nearly one the spot.

Correlation between echolocation and hunting behavior The comparison of sound sequences and pursuits shows that a turn towards the target and the beginning of the approach phase occur at about the same time. The simultaneous reaction in both be-

haviors suggests that the insect is detected just before the bat reacts, which means in the pulse preceding the beginning of the approach phase as we define it. The interval between this pulse and the beginning of the approach phase is about 80 ms long, the time required to cover a distance of 40 cm at a flight speed of 5 m/s. This distance has to be added to the estimated reaction distance to get the detection distance which would be between 110 and 160 cm. The beginning of the upward tilting of the bat's body when approaching the insect coincides with the beginning of the final buzz when the bat is 30 to 50 cm from the insect. The instant the bat switches to the second section of the final buzz, characterized by low frequency sounds, the bat has already reached the upright position with extended wings and cupped tail membrane. The distance to the insect at the end of the buzz could not be measured but we guess that the buzz ends when the bat is very close to the insect or when the insect is caught. In successful catches the duration of the pause after the terminal phase may indicate how fast the bat retrieves the insect from the tail pouch. A short pause can also indicate that the bat missed the insect and quickly returned to the searching phase.

Sound emission and wingbeat cycle Sound emission correlates with wingbeat cycle. During searching flight, bats usually produce one sound when the wings are near the top of the stroke. The pulse interval of either approximately 100 ms or 200 ms reflects the correlation of sound emission, respiratory cycle and wingbeat rate (Schnitzler 1971; Suthers et al. 1972). A 200 ms pause indicates a wingbeat without sound emission. From this correlation we predicted that the wingbeat rate was about 10 wingbeats/s. This result has been confirmed by our photographs. For example the bat in the background of Fig. 4 shows about the same wing position in each exposure at a flash rate of 10 Hz.

Moths at the study site During the 10 days spent at the study site we trapped 146 moths using an ultraviolet (black) fluorescent light. Most of the moths belonged to two families, Lymantridae (Lymantria dispar, n = 7 5 ) and Noctuidae (25 species in 22 genera, n=55). The most numerous noctuid was Noctua tirreniea (n=11). We caught only 5 geometrids and 3 arctiids. Both lymantrid and noctuid moths have ears

H.-U. Schnitzler et al. : Echolocation and hunting behavior in Pipistrellus

sensitive to ultrasonic frequencies. We saw numerous L. dispar flying around the lamp post, but never saw a clear reaction to hunting bats or to artificial bat sounds generated by a custom made portable system (developed at the Institute of Biology, Odense University by B.B. Andersen and L.A. Miller). Discussion

The nearly constant frequency pulses used in the search phase improve target detection since the sound energy is concentrated into the narrow bandwidth of individual neurons (Grinnell and Hagiwara 1972). The rather long pulse durations of 8-12 ms and the duty cycle of 6-10% also increase the chance for the bat to receive a 'glint' echo from the insect prey. A glint (increase in echo amplitude) is produced when the insect's wings are perpendicular to the sound path giving a strong echo (Schnitzler et al. 1983; Schnitzler 1986). For instance, if we assume that an insect (Diptera) has a wingbeat rate of 100 Hz, a bat with a duty cycle of 8 % would have a chance to perceive 8 glints/s, which means about one glint per echo. Detection distance could be increased when an echo contains a glint. If long pulses improve the chance to detect a glint and therefore increase the detection distance why does P. kuhli not increase the duration of search pulses ? We assume that longer pulses would have the disadvantage of overlapping with the returning echo and therefore reducing the chance of detection by masking effects. At detection distances of 110 to 160 cm pulse durations of more than 6,5 to 9.5 ms would produce overlaps between returning echoes and outgoing pulses. Our recordings showed that in the few cases studied the bats just prevented this overlap. This suggests that P. kuhli tends to avoid pulse-echo overlap in order to avoid masking of insect echoes by its own sound. Thus the duration of the search pulses may indicate the minimum distance beyond which P. kuhli searches for insects. With 12 ms long search pulses P. kuhli is searching for insects beyond a distance of 204 em, while at 8 ms the search range begins at beyond 136 cm. The above argument could apply to all other FM-bats using long shallow sweeps in the search phase. The detection distance may also explain why the bat can afford to skip a pulse cycle during the search phase. At a flight speed of about 5 m/s the long signal intervals of about 200 ms correspond to a flight distance of 100 cm. Under these conditions the detection distance should be more

273

than 100 cm; otherwise, the bat might miss insects during the long interval. The estimated detection distance of 110-160 cm corresponds well with this suggestion. In all stages of a pursuit the pulse intervals are sufficiently long that the echo returning from the insect does not overlap with the next sound. The sound duration decreases when the bat approaches the insect. The pulse duration is always short enough so that the emitted pulse is finished when the echo returns and no pulse-echo overlap occurs. For instance, at the beginning of an approach phase (distance 120 cm) the echo returns after 7 ms and the pulse duration is about 7 ms. At the beginning of the terminal phase (distance 50 cm) the echo returns after 3 ms and the pulse duration is about 2 ms. At the beginning of the second section of the terminal phase (distance around 10 cm) the echo returns after 0.7 ms and the pulse duration is about 0.3 ms. At a signal duration of 0.3 ms which is also found at the end of the second section of the terminal phase an overlap occurs at a distance of about 5 cm. This may indicate to the bat that the prey is very close and that the bat can stop emitting signals and prepare for the catch. The above compares well with results from Myotis lucifugus, which also prevents pulseecho overlap in the approach and terminal phase (Cahlander et al. 1964) and with the results from Rhinolophusferrumequinum, which prevent overlap of the F M portion of their CF-FM signals (Schnitzler 1968). The increase in signal bandwidth and the decrease in sound duration in the approach and in the first section of the terminal phase probably improves the accuracy of localization (reviewed in Schnitzler et al. 1985). However, the lowering of the pulse frequency in the second section of the terminal phase is difficult to explain. As this is found also in other bats (e.g. in Pipistrellus pipistrellus and Myotis daubentoni, unpublished results) it is probably not accidental. Acknowledgements. We thank Norbert Hirneisen for capturing and identifying insects near the study site and Klaus Heblich for developing the multi flash-sequencer. We also thank Hans Baagoe for calculating flight speeds of Pipistrellus kuhli based on our photographs and for fruitful discussions on the flight behavior of bats. The research was supported by the Deutsche Forschungsgemeinschaft (SFB 307) and by Grants from the Danish Natural Sciences Research Council.

References Cahlander DA, McCue JJG, Webster FA (/964) The determination of distance by echolocating bats. Nature 20/: 544-546

274

H.-U. Schnitzler et al. : Echolocation and hunting behavior in Pipistrellus

Griffin DR, Webster FA, Michael CR (1960) The echolocation of flying insects by bats. Anim Behav 8:141-154 Grinnell AD, Hagiwara S (1972) Adaptations of the auditory nervous system for echolocation studies of New Guinea bats. Z Vergl Physiol 76:41-81 Miller LA, Andersen BB (1984) Studying bat echolocation signals using ultrasonic detectors. Z Sfiugetierk 49:6-13 Neuweiler G (1983) Echolocation and adaptivity to ecological constraints. In: Huber F, Markl H (eds) Neuroethology and behavioral physiology. Springer, Berlin Heidelberg New York, pp 280-302 Novick A (1977) Acoustic orientation. In: Wimsatt WA (ed) Biology of bats, vol III. Academic Press, New York, pp 73287 Pye D (1980) Adaptiveness of echolocation signals in bats. TINS, pp 232-235 Schnitzler HU (1968) Die Ultraschall-Ortungslaute der Hufeisen-Flederm/iuse (Chiroptera-Rhinolophidae) in verschiedenen Orientierungssituationen. Z Vergl Physiol 57: 376408 Schnitzler HU (1971) Bats in the wind tunnel. Z Vergl Physiol 73 : 209-221 Schnitzler HU (1987) Echoes of fluttering insects: information for echolocating bats. In: Fenton MB, Racey PA, Rayner IMV (eds) Recent advances in the study of bats. Cambridge University Press, Cambridge, pp 226-243 Schnitzler HU, Henson OW Jr (1980) Performance of airborne animal sonar systems: I. Microchiroptera. In: Busnel RG, Fish JF (eds) Animal sonar systems, Plenum Press, New York, pp 109-181

Schnitzler HU, Menne D, Kober R, Heblich K (1983) The acoustical image of fluttering insects in echolocating bats. In: Huber F, Markl H (eds) Neuroethology and behavioral physiology. Roots and growing points. Springer, Berlin Heidelberg New York, pp 235-250 Schnitzler HU, Menne D, Hackbarth H (1985) Range determination by measuring time delays in echolocating bats. In: Michelson A (ed) Time resolution in auditory systems. Springer, Berlin Heidelberg New York, pp 180-204 Simmons JA, Fenton MB, O'Farrell MJ (1979) Echolocation and pursuit of prey by bats. Science 203 : 16-21 Suthers RA, Thomas SP, Suthers BJ (1972) Respiration, wing beat and ultrasonic pulse emission in an echolocating bat. J Exp Zool 56:37-48 Trappe M (1982) Verhalten und Echoortung der Grossen Hufeisennase beim lnsektenfang. Dissertation, Fakult/it fiir Biologic, Tiibingen Trappe M, Schnitzler HU (1982) Doppler shift compensation in insect-catching horseshoe bats. Naturwissenschaften 69:193-194 Vogler B, Neuweiler G (1983) Echolocation in the noctule (Nyctalus noctula) and horseshoe bat (Rhinolophus ferrumequinum). J Comp Physiol 152:421-432 Webster FA (1963) Active energy radiating systems: the bat and ultrasonic principles. II. Acoustical control of airborne interceptions by bats. Proc Int Congr Tech and Blindness. AFB, New York 1:4%135 Webster FA, Griffin DR (1962) The role of the flight membranes in insect capture in bats. Anim Behav 10: 332340