The brain-stem auditory-evoked response in the big brown ... brain-stem... · The brain-stem...

10
The brain-stem auditory-evoked response in the big brown bat (Eptesicus fuscus) to clicks and frequency-modulated sweeps Robert Burkard Departments of Communication Disorders and Otolaryngology, Boston University, Boston, Massachusetts 02215 Gynthia F. Moss Department of Psychology, Harvard University, Cambridge, Massachusetts 02138 (Received 29 June 1993;accepted for publication 3 February 1994) Three experiments were performed to evaluate the effects of stimulus level on the brain-stem auditory-evoked response (BAER) in the big brown bat (Eptesicus fuscus), a species that uses frequency-modulated (FM) sonar sounds for echolocation. In experiment 1, theeffects of click level ontheBAER wereinvestigated. Clicks werepresented at levels of 30 to 90 dB pSPL in 10-dBsteps. Eachanimalresponded reliablyto clicksat levelsof 50 dB pSPL and above, showing a BAER containing four peaks in the first 3-4 ms from click onset (waves i-iv). Withincreasing click level, BAER peakamplitude increased andpeaklatency decreased. A decrease in the i-iv interval also occurred with increasing click level. In experiment 2, stimuli were 1-ms linear FM sweeps, decreasing in frequency from 100 to 20 kHz. Stimulus levelsranged from 20 to 90 dB pSPL. BAERs to FM sweeps were observed in all animals for levelsof 40 dB pSPL and above. These responses were similarto the click-evoked BAER in waveform morphology, with the notable exception of an additional peak observed at thehigher levels of FM sweeps. This peak (wave ia) occurred priorto the firstwave seen at lowerlevels (waveib). As the level of the FM sweep increased, there wasa decrease in peak latency andan increase in peak amplitude. Similarity in the magnitude and behavior of thei-iv andib-iv intervals suggests that waveib to FM sweeps is the homolog of thewavei response to click stimuli. Experiment 3 tested the hypothesis thatwaveia represented activity emanating from more basal cochlear regions than wave ib.FM sweeps (100-20 kHz) werepresented at 90 dB pSPL,and broadband noise wasraised in leveluntil the BAER was eliminated. This "masked threshold" occurred at 85 dB SPL of noise. At masked threshold, the broadband noise was steeply high-pass filtered at five cutoff frequencies ranging from20 to 80 kHz. Generally, wave ia was eliminated for masker cutoff frequencies of 56.6 kHz and below, whilewave ib was typically observed for masker cutoffsdown to 28.3 kHz. The results of thesethree experiments arecompared and contrasted with data from other mammalian BAER studies. PACS numbers: 43.64.Qh, 43.64.Ri, 43.64.Tk INTRODUCTION The echolocating bat,Eptesicus fuscus, actively interro- gates its environment by emitting ultrasonicfrequency- modulated (FM) vocal signals and listening to the echoes reflected off targets in the pathof the sound beam.From the featuresof the returningechoes, the bat extractsspatial acoustic information, guiding orientation in the environment andthe capture of prey (Griffin, 1958). Particularly impor- tant to the bat's success in intercepting prey is a reliable estimate of target range, whichthe bat obtains from the time interval between its sonar emissions andthe returning echoes (Simmons, 1973).The bat's perception of target range is sus- ceptibleto amplitude-induced latencyshiftsin neural dis- charge, and psychophysical data from thisbat show a change in neuraldischarge latency of 13-18/as for eachdB change in sound level (Simmons et al., 1990a,b). Simmons et al. (1990a) studied amplitude-induced la- tency shiftsin the bat's auditory system, recording brain- stemauditory-evoked responses (BAERs) from a tungsten electrode placed in theinferior colliculus of Eptesicus fuscus. The stimuli used in this experiment were FM sweeps, simu- lating the bat'sbiosonar sounds. At a sound level approxi- mately 15 dB above threshold, they reporta slope of -13 /as/dBfor the N1 response and -18 /as/dBfor the N4 re- sponse, values thatareconsistent with thebehavioral data on time-intensity trading for the perception of target range (Sim- mons et al., 1990a,b). The different slopes of the N1 and N4 responses re- ported by Simmons et al. (1990a) result in a change in the time interval between the two waves with stimulus level, a finding that contrasts with BAER datain other mammalian species. In other mammals, theslope of thelatency/intensity function hasbeenmeasured using click stimuliand far-field recording (e.g., Burkard and Voigt, 1989), and the typical mammalian BAER latency/intensity function shows similar slopes for the early and later peaks. Thus the interval be- tween the BAER waves recorded from other mammals is relativelyconstant across sound level. 801 J. Acoust. Soc.Am. 96 (2), Pt. 1, August 1994 0001-4966/94/96(2)/801/10/$6.00 ¸ 1994Acoustical Society of America 801 Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 128.220.159.2 On: Wed, 11 May 2016 20:15:42

Transcript of The brain-stem auditory-evoked response in the big brown ... brain-stem... · The brain-stem...

Page 1: The brain-stem auditory-evoked response in the big brown ... brain-stem... · The brain-stem auditory-evoked response in the big brown bat (Eptesicus fuscus) to clicks and frequency-modulated

The brain-stem auditory-evoked response in the big brown bat (Eptesicus fuscus) to clicks and frequency-modulated sweeps

Robert Burkard

Departments of Communication Disorders and Otolaryngology, Boston University, Boston, Massachusetts 02215

Gynthia F. Moss Department of Psychology, Harvard University, Cambridge, Massachusetts 02138

(Received 29 June 1993; accepted for publication 3 February 1994)

Three experiments were performed to evaluate the effects of stimulus level on the brain-stem auditory-evoked response (BAER) in the big brown bat (Eptesicus fuscus), a species that uses frequency-modulated (FM) sonar sounds for echolocation. In experiment 1, the effects of click level on the BAER were investigated. Clicks were presented at levels of 30 to 90 dB pSPL in 10-dB steps. Each animal responded reliably to clicks at levels of 50 dB pSPL and above, showing a BAER containing four peaks in the first 3-4 ms from click onset (waves i-iv). With increasing click level, BAER peak amplitude increased and peak latency decreased. A decrease in the i-iv interval also occurred with increasing click level. In experiment 2, stimuli were 1-ms linear FM sweeps, decreasing in frequency from 100 to 20 kHz. Stimulus levels ranged from 20 to 90 dB pSPL. BAERs to FM sweeps were observed in all animals for levels of 40 dB pSPL and above. These responses were similar to the click-evoked BAER in waveform morphology, with the notable exception of an additional peak observed at the higher levels of FM sweeps. This peak (wave ia) occurred prior to the first wave seen at lower levels (wave ib). As the level of the FM sweep increased, there was a decrease in peak latency and an increase in peak amplitude. Similarity in the magnitude and behavior of the i-iv and ib-iv intervals suggests that wave ib to FM sweeps is the homolog of the wave i response to click stimuli. Experiment 3 tested the hypothesis that wave ia represented activity emanating from more basal cochlear regions than wave ib. FM sweeps (100-20 kHz) were presented at 90 dB pSPL, and broadband noise was raised in level until the BAER was eliminated. This "masked threshold" occurred at 85 dB SPL of noise. At masked threshold, the

broadband noise was steeply high-pass filtered at five cutoff frequencies ranging from 20 to 80 kHz. Generally, wave ia was eliminated for masker cutoff frequencies of 56.6 kHz and below, while wave ib was typically observed for masker cutoffs down to 28.3 kHz. The results of these three experiments are compared and contrasted with data from other mammalian BAER studies.

PACS numbers: 43.64.Qh, 43.64.Ri, 43.64.Tk

INTRODUCTION

The echolocating bat, Eptesicus fuscus, actively interro- gates its environment by emitting ultrasonic frequency- modulated (FM) vocal signals and listening to the echoes reflected off targets in the path of the sound beam. From the features of the returning echoes, the bat extracts spatial acoustic information, guiding orientation in the environment and the capture of prey (Griffin, 1958). Particularly impor- tant to the bat's success in intercepting prey is a reliable estimate of target range, which the bat obtains from the time interval between its sonar emissions and the returning echoes (Simmons, 1973). The bat's perception of target range is sus- ceptible to amplitude-induced latency shifts in neural dis- charge, and psychophysical data from this bat show a change in neural discharge latency of 13-18/as for each dB change in sound level (Simmons et al., 1990a,b).

Simmons et al. (1990a) studied amplitude-induced la- tency shifts in the bat's auditory system, recording brain- stem auditory-evoked responses (BAERs) from a tungsten

electrode placed in the inferior colliculus of Eptesicus fuscus. The stimuli used in this experiment were FM sweeps, simu- lating the bat's biosonar sounds. At a sound level approxi- mately 15 dB above threshold, they report a slope of -13 /as/dB for the N1 response and -18 /as/dB for the N4 re- sponse, values that are consistent with the behavioral data on time-intensity trading for the perception of target range (Sim- mons et al., 1990a,b).

The different slopes of the N1 and N4 responses re- ported by Simmons et al. (1990a) result in a change in the time interval between the two waves with stimulus level, a

finding that contrasts with BAER data in other mammalian species. In other mammals, the slope of the latency/intensity function has been measured using click stimuli and far-field recording (e.g., Burkard and Voigt, 1989), and the typical mammalian BAER latency/intensity function shows similar slopes for the early and later peaks. Thus the interval be- tween the BAER waves recorded from other mammals is

relatively constant across sound level.

801 J. Acoust. Soc. Am. 96 (2), Pt. 1, August 1994 0001-4966/94/96(2)/801/10/$6.00 ¸ 1994 Acoustical Society of America 801

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 128.220.159.2 On: Wed, 11 May 2016 20:15:42

Page 2: The brain-stem auditory-evoked response in the big brown ... brain-stem... · The brain-stem auditory-evoked response in the big brown bat (Eptesicus fuscus) to clicks and frequency-modulated

The level dependence of the interval between the early and later BAER peaks in the bat may reflect some special- ization of auditory processing particular to echolocation be- havior. Alternatively, this result may be attributed to the in- vasive recording procedure, the broadband spectral characteristics of the sounds (20-100 kHz) and/or the FM stimuli used in the Simmons et al. (1990a) study, all distinct from the conventional far-field recording procedures and the click stimuli used in other experiments. In the present study, we examined these possibilities by obtaining latency/ intensity functions from far-field BAERs in the echolocating bat, Eptesicus fuscus, using both clicks and FM sweeps.

I. GENERAL METHODS

Brain-stem auditory responses (BAERs) were recorded from adult big brown bats (Eptesicus fuscus), which ranged in weight from 12.5-23 g, in three separate experiments. The animals Were anesthetized with an intramuscular injection of Ketamine and Xylazine (35 mg/kg) and placed on a homo- thermic blanket set to 38 øC.

For all experiments, electrical activity was recorded with Grass needle electrodes placed at skull midline (noninvert- ing), posterior to both pinnae (linked: inverting), and in the midline of the back (common). This activity was amplified, filtered (0.1-10 kHz) and signal averaged by a Nicolet Com- pact 4, with a 10.24 ms time epoch and 50-kHz sampling frequency. Each average was composed of 500 artifact-free sweeps, and for each condition, two averages were taken.

Acoustic stimuli were broadcast through a high- frequency transducer (Ultrasound Advice model USLS) placed 17 cm in front of the bat. Sound pressure level (SPL) was controlled via a Tucker-Davis Electronics attenuator

1

(model PA3) and monitored with • in. Bruel & Kjaer micro- phone (model 4138) which was located 2-3 cm above the bat's skull midline. The microphone was powered by a Lar- son Davis preamplifier/power supply (2200 C). Time-domain and frequency-domain representations of the acoustic stimuli were obtained by recording the signals with a LeCroy model 9310 digital oscilloscope. Spectra were obtained by passing the time-domain signal through a uniform window and per- forming a fast Fourier transform.

Dependent variables were the latency and peak-to- following-trough amplitude of waves i, ii, and iv. All peak latency data reported here correct for the 0.5-ms acoustic delay created by the 17-cm distance between speaker and bat.

II. EXPERIMENT I

A. Introduction

Clicks have been popular stimuli for the study of the BAER, due to their brief duration and broadband spectrum. Clicks have been used to elicit BAERs in a variety of mam- malian species, including humans, rats, mice, gerbils, cats, and monkeys (Burkard and Hecox, 1983; Burkard et al., 1990; Henry, 1979; Burkard and Voigt, 1989; Huang and Buchwald, 1978; Kraus et al., 1985). This experiment exam- ined the effects of click level on the BAER of the big brown bat.

-18.4

.-• -36.8

-55.2

-73.6 20 40 60 80 100

Frequency (kHz)

FIG. 1. The acoustic spectrum for the click stimuli used in experiment 1 is shown. The time-domain waveform is shown in the inset. For the inset, the

time base is 0.02 ms per division.

B. Methods

BAERs were recorded from six bats, and the order of stimulus level was randomized for each animal. Stimuli con-

sisted of 10-/as duration electrical pulses (Grass S88) pre- sented at a rate of 27 Hz at levels of 30 to 90 dB pSPL in 10-dB increments. The broadcast clicks had an initial con-

densation phase. The time-domain waveform and spectrum of the acoustic click stimulus used in experiment 1 are shown in Fig. 1. This stimulus shows a broad maximum, with the 10-dB down points extending from approximately 10 to 70 kHz. The rolloff at the high frequencies represents low-pass properties of the transducer, the electronic switch and the attenuators.

C. Results

Figure 2 shows BAERs from one representative bat for clicks ranging from 30 to 90 dB pSPL. The first four waves of the response occur within 5 ms of click onset. Wave iii tended to occur on the downward slope of wave ii, and was not always observed, especially at lower click levels. We therefore did not evaluate the behavior of wave iii in this

investigation. Wave iv occasionally showed two distinct peaks, and in these instances we chose the earliest peak in the complex.

Mean wave i, ii, and iv latencies, and standard devia- tions, are plotted across click level in Fig. 3 and show, as expected, a decrease in the observed peak latencies with in- creasing click level. Waves i, ii, and iv were observed in all six bats for click levels of 50 dB pSPL and above, with most animals showing an identifiable response to 40 dB pSPL clicks. Linear regression analyses were used to quantify the slope of the latency/intensity function to click stimuli. Re- sults of these analyses are presented in Table I. The first row in this table shows the slope of the linear regression line fit to the latency data obtained with 50-90 dB pSPL clicks. Sub- sequent rows show piecewise linear regression slopes across the 10-dB steps in click level. Several points are noteworthy. First, the slope of the latency/intensity function is shallower

802 J. Acoust. Soc. Am., Vol. 96, No. 2, Pt. 1, August 1994 R. Burkard and C. F. Moss: Bat BAERs to clicks and FM sweeps 802

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 128.220.159.2 On: Wed, 11 May 2016 20:15:42

Page 3: The brain-stem auditory-evoked response in the big brown ... brain-stem... · The brain-stem auditory-evoked response in the big brown bat (Eptesicus fuscus) to clicks and frequency-modulated

dB

pSPL

go

8o

7o

6o

5o

4o

3o

CLICK i

lii >• BATS /•... iv

- iJ •,•• .......

2 4 6 8 10

Time (ms)

FIG. 2. BAERs to a click intensity series from one bat are shown. For each click level, two responses are superimposed.

at higher click levels for all three peaks. Second, the overall slope of the latency/intensity function (50-90 dB pSPL) ap- pears to be progressively steeper for the later BAER peaks, with a slope of -8, -13, and -18 •s/dB for waves i, ii, and iv, respectively. The piecewise regression slopes show that the steeper slopes of the later BAER peaks are the result of latency shifts occurring for click levels at and below 70 dB pSPL. The effect of click level on peak latency was evalu-

TABLE I. Latency/intensity function slopes (/zs/dB) for clicks. These linear- regression slopes are based on the peak latencies of six bats.

Click levels Wave i Wave ii Wave iv

50-90 dB pSPL -7.63 - 12.78 - 17.68 50-60 dB pSPL -10.50 -21.17 -31.33 60-70 dB pSPL - 11.67 -20.83 -32.00 70-80 dB pSPL -5.00 -6.00 -4.83 80-90 dB pSPL - 2.67 - 2.5 - 1.83

ated statistically with three one-way repeated measure ANOVAs (one for each peak), and all were significant (p

Mean i-iv intervals, and standard deviations, are plotted across click level in Fig. 4. There is a clear decrease in the i-iv interval for increasing click level from 50 to 70 dB pSPL, with little change in the i-iv interval for increases in click level from 70 to 90 dB pSPL. The change in i-iv in- terval across click level reflects the steeper latency/intensity function for wave iv, as compared to wave i (see Table I). To evaluate whether the i-iv interval decrease with increasing click level was statistically significant, a one-way repeated- measure ANOVA was performed. There was a significant effect of click level on the i-iv interval (p<0.01).

The mean amplitude of waves i, ii, and iv are plotted across click level in Fig. 5. There is an increase in the am- plitude of waves i, ii, and iv with increasing click level. However, the slopes of the wave i and wave ii amplitude/ intensity functions are clearly steeper than that seen for wave iv. At 50 dB pSPL, mean amplitude of waves i, ii, and iv are 1.69-2.94 •V. At 90 dB pSPL, the mean amplitude of waves i, ii, and iv are 24.05, 21.67, and 6.7 •V, respectively. The effect of click level on peak amplitude was evaluated with three one-way repeated-measure ANOVAs (one for each peak), and all were significant (p<0.01).

ß J 1.5

Click

ß i a ii o iv

T

dB pSPL

FIG. 3. Mean BAER peak latencies are plotted across click pSPL. Data for waves i, ii, and iv are shown. Each data point represents the mean peak latency of six bats. The error bars show the sample standard deviation.

Click 2.7

'•'• 2.3

1.5 50 60 70 80 90

dB pSPL

FIG. 4. Mean i-iv interval is plotted across click pSPL. Each data point represents the mean i-iv interval of six bats. The error bars show the sample standard deviation.

803 J. Acoust. Soc. Am., Vol. 96, No. 2, Pt. 1, August 1994 R. Burkard and C. F. Moss: Bat BAERs to clicks and FM sweeps 803

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 128.220.159.2 On: Wed, 11 May 2016 20:15:42

Page 4: The brain-stem auditory-evoked response in the big brown ... brain-stem... · The brain-stem auditory-evoked response in the big brown bat (Eptesicus fuscus) to clicks and frequency-modulated

25.0 - .

-

20.0

Click

•-'•15.0 -

'-- 10.0

5.0 •

-

.

0.0 5'0 ........ '6'0 .... '"' '7'0 ........ ........ '9'0 dB pSPL

FIG. 5. Mean peak amplitudes are plotted across click pSPL. Data for waves i, ii, and iv are shown. Each data point represents the mean peak amplitude of six bats.

III. EXPERIMENT 2

A. Introduction

The big brown bat emits biosonar signals consisting of brief FM sounds, which sweep from approximately 100 to 20 kHz in the approach phase of insect pursuit (Griffin, 1958; Kick and Simmons, 1984; Fattu and Suthers, 1981). In the present investigation, we used approximations to these bio- acoustic signals as stimuli and investigated the effects of FM-sweep level on the BAER.

B. Methods

-18 4

ß = -36.8

. -55.2

-73.6

20 40 60 80 100

Frequency (kHz)

FIG. 6. The acoustic spectrum for the FM-sweep stimuli used in experiment 2 is shown. The time-domain waveform is shown in the inset. For the inset, the time base is 0.2 ms per division.

C. Results

An FM-sweep intensity series from one representative bat is shown in Fig. 7. These responses differed from the BAERs to click stimuli in several ways. First, wave iii was rarely observed in response to FM sweeps. Second, at lower FM-sweep levels (e.g., see the response to the 50 dB pSPL stimulus), the response appeared similar to the click-evoked BAER, with three prominent waves. However, with higher level FM sweeps an early wave occurred prior to the first

dB

pSPL

BAERS were obtained from six animals. Responses 90 were recorded to FM stimuli sweeping from 100 to 20 kHz in I ms. FM sweeps were presented at a rate of 27 Hz at 80 levels of 20 to 90 dB pSPL in 10-dB steps, and the order of stimulus presentation was randomized. FM sweep band- width, duration and presentation rate were selected to ap- 70 proximate the bat's biosonar pulses during the approach phase of insect pursuit (Griffin, 1958; Kick and Simmons, 60 1984; Simmons, 1989). The 27-Hz rate has also been used in earlier investigations of the BAER in humans and gerbils (Burkard and Hecox, 1983a; Burkard and Voigt, 1989), and 50 this choice facilitates cross-species comparisons. The FM sounds were generated by a Stanford Research Systems 40 model D5245 function generator which was gated with a 30 0.25-ms cosine-squared onset and offset (Wilsonics model BSIT). The electronic switch was controlled by a gating sig- 20 nal produced by a pulse generator (Grass S88). The time- and frequency-domain representations of the FM sweeps used in experiment 2 are shown in Fig. 6. The 10-dB down bandwidth is approximately 20-70 kHz. There is less low- frequency energy in the FM sweep than in the click (Fig. 1) due to the 20-kHz cutoff of the FM sweep.

FM SWEEP

2 4 6 8 10

Time (ms)

FIG. 7. BAERs from one bat for an FM-sweep intensity series are shown. For each click level, two responses are superimposed.

804 J. Acoust. Soc. Am., Vol. 96, No. 2, Pt. 1, August 1994 R. Burkard and C. F. Moss: Bat BAERs to clicks and FM sweeps 804

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 128.220.159.2 On: Wed, 11 May 2016 20:15:42

Page 5: The brain-stem auditory-evoked response in the big brown ... brain-stem... · The brain-stem auditory-evoked response in the big brown bat (Eptesicus fuscus) to clicks and frequency-modulated

FM SWEEP

i '- 1,8

i !! • •'• 1,7 0.8 1.5 -- i I i i i i i i i i i i i i i i i i i i i i i i i i i i i I i i I i i •5 • • •7• •1• ;:Jb 4.•) 50 60 70 80 90

dB pSPL dB pSPL

FIG. 8. Mean BAER peak latencies are plotted across FM-sweep pSPL. Data for waves ia, ib, ii, and iv are shown. Each data point represents the mean latency of six bats for waves ib, ii, and iv. For wave ia, from 3-6 latency values contribute to each mean value. Error bars show the sample standard deviation.

peak observed with lower-level FM sweeps. This early peak to high-level FM sweeps was labeled wave ia, while the later peak was labeled wave ib. Most bats showed an identifiable response to 30 dB pSPL FM sweeps, and all bats showed a BAER at levels of 40 dB pSPL and above. The threshold of wave ia varied widely across animals, ranging from 30 to 80 dB pSPL. Assigning a threshold for the appearance of this peak was difficult in some animals. For example, the shoul- der occurring prior to wave ib for the 50 dB pSPL condition in Fig. 7 may represent activity contributing to wave ia.

Mean peak latencies, and standard deviations, across FM-sweep level are shown in Fig. 8. With increasing stimu- lus level, there is a decrease in mean peak latency of waves ia, ib, ii, and iv. Waves ib, ii, and iv were observed in all bats for FM-sweep levels of 40 dB pSPL and above. In contrast, wave ia first appeared in three of the six bats at the 60 dB pSPL FM-sweep level. All six bats showed wave ia at 80 dB pSPL and above. With increasing stimulus level, there is a decrease in BAER peak latency. Mean wave iv latency pla- teaus at the 80 to 90 dB pSPL FM-sweep levels.

Wave ib, ii, and iv peak latency data were fit with linear regression lines, and the slopes of these latency/intensity functions are shown in Table II. The first row shows the

latency/intensity function slope for 40-90 dB pSPL FM

TABLE II. Latency/intensity function slopes (/xs/dB) for FM sweeps. These linear-regression slopes are based on the peak latencies of six bats.

Sweep levels Wave ib Wave ii Wave iv

40-90 dB pSPL -5.82 -8.55 - 14.27 40-50 dB pSPL - 11.33 - 12.83 -21.83 50-60 dB pSPL -9.83 - 14.00 -30.33 60-70 dB pSPL -5.17 -6.67 -8.83 70-80 dB pSPL - 1.50 -3.50 - 10.50 80-90 dB pSPL -2.00 -7.00 +3.17

FIG. 9. Mean ib-iv interval is plotted across FM-sweep pSPL. Each data point represents the mean ib-iv interval of six bats. Error bars show the sample standard deviation.

sweeps. As seen for clicks, there is an increase in latency/ intensity function slope for the later BAER peaks. Latency/ intensity function slopes are -6, -9, and -14 /.ts/dB for waves ib, ii, and iv, respectively. Latency/intensity function slopes are generally shallower for higher FM sweep levels, although slope does not vary monotonically with FM-sweep level. The small increase in mean wave iv latency for an increase in FM-sweep level from 80 to 90 dB pSPL results in a small positive latency/intensity function slope. A paired t-test was performed to evaluate whether wave iv latency was significantly different for 80 and 90 dB pSPL FM sweeps and the results indicate no difference (p=0.5969). The effects of FM sweep level on wave ib, ii, and iv latencies were evaluated statistically with three one-way repeated- measure ANOVAs, and all were significant (p<0.01).

Mean ib-iv intervals, and standard deviations, are plot- ted across FM-sweep level in Fig. 9. There is a decrease in ib-iv interval with increasing FM-sweep level. This change in ib-iv interval with stimulus level reflects the steeper latency/intensity function for wave iv, as compared to wave ib. A comparison of Figs. 9 and 4 shows that the ib-iv inter- val across FM-sweep level is similar in magnitude to the i-iv interval to click stimuli. This suggests that wave ib to FM sweeps is homologous to wave i to click stimuli. A one-way ANOVA evaluating the effect of FM-sweep level on ib-iv interval reached significance (p<0.01).

Mean peak amplitudes are plotted across FM-sweep level in Fig. 10. For all stimulus levels, wave ia is small in amplitude, compared to waves ib, ii, and iv. All waves tend to increase in amplitude with increasing FM-sweep level. Waves ib, ii, and iv are of similar magnitude for FM-sweep levels of 40 to 60 or 70 dB pSPL. For FM-sweep levels above 70 dB pSPL, the amplitude of wave ib continues to increase with increasing stimulus level, while waves ii and iv do not change in amplitude with increasing stimulus level above 70 dB pSPL. The effect of FM-sweep level on wave

805 J. Acoust. Soc. Am., Vol. 96, No. 2, Pt. 1, August 1994 R. Burkard and C. F. Moss: Bat BAERs to clicks and FM sweeps 805

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 128.220.159.2 On: Wed, 11 May 2016 20:15:42

Page 6: The brain-stem auditory-evoked response in the big brown ... brain-stem... · The brain-stem auditory-evoked response in the big brown bat (Eptesicus fuscus) to clicks and frequency-modulated

10.0

8.0

6.0

4.0

2.0

0.0

FM SWEEP

,

ß , ii /

,.., 0

I I I I I I I I I I I I I I I I I I I I I I I I I I I Jl I I III Ill II I Ill llll I I I i I 4-0 50 60 70 80 90

dB pSPL

FIG. 10. Mean peak amplitudes are plotted across FM-sweep pSPL. Data for waves ia, ib, ii, and iv are shown. Each data point represents the mean amplitude of six bats for waves ib, ii, and iv. For wave ia, from 3-6 ampli- tude values contribute to each mean value.

ib, ii, and iv amplitude was evaluated statistically with three one-way repeated-measures ANOVAs, and all were signifi- cant (p <0.01).

IV. EXPERIMENT 3

A. Introduction

The BAER data for click and FM stimuli, for experi- ments 1 and 2, show remarkable similarities. However, one noteworthy difference is the presence of wave ia at high FM-sweep levels. For click stimuli, only one wave i peak occurred at high stimulus levels. One possible explanation for the presence of wave ia at higher FM-sweep levels is that wave ia represents activity from the high frequency regions of the cochlea, and wave ib represents activity of more apical regions of the cochlea. At high levels, there is a response to the onset of the stimulus, which contains the highest frequen- cies in the sweep (the first 0.25 ms represents 100-80 kHz). There is a second response in more apical cochlear regions toward the middle or end of the sweep (the last half of the stimulus represents the frequency region from 60-20 kHz). The temporal dispersion of the frequency content of the FM sweep makes it possible to independently activate different cochlear regions, and to observe multiple responses in the BAER. Perhaps double later waves (e.g., a second wave ii) were not observed in the waveform because they are embed- ded in the ongoing response. As multiple wave i peaks were not observed to clicks, it appears that temporal dispersion of the frequency content of the stimulus is necessary to evoke multiple wave i peaks. Moreover, wave ia may not be ob- served at lower FM-sweep levels because of less stimulus energy in the high frequencies (see Fig. 6).

In order to evaluate the possibility that wave ia arises from the temporal dispersion of stimulus frequency, we used 90 dB pSPL 100-20 kHz FM sweeps in combination with high-pass masking noise. The broadband masking noise is

HIGH PASS

unmasked

80 kHz

56.6 kHz

40 kHz

28.3 kHz

20 kHz

2 4 6 8 10

Time (ms)

FIG. 11. BAERs to 90 dB pSPL 100-20 kHz FM sweeps from one bat for all high-pass masker cutoff frequencies used in experiment 3 are shown. For each condition, two responses are superimposed.

raised in level until the response to the FM sweep is elimi- nated. Then the masking noise is steeply high-pass filtered, to eliminate the activity from cochlear regions which are basal to the masker high-pass cutoff frequency. If our hy- pothesis that wave ia represents basal cochlear activation is correct, then this peak should disappear as the cutoff fre- quency of the high-pass masker is reduced.

B. Methods

BAERs were recorded from four animals. FM sounds

were generated, shaped and gated as described above (see methods, experiment 2), but were attenuated by an HP350D attenuator. The FM sounds were presented at 27 Hz at 90 dB pSPL. White noise (Scott model 811-B) was bandpass fil- tered from 100 Hz to 110 kHz (Kemo VBF-8), and high-pass filtered at a nominal 96 dB/oct by ganging two channels of a second Kemo VBF-8 filter. The noise level was controlled by a second HP350D attenuator. The output of the FM sweep and noise attenuators were electrically mixed by a custom- built passive mixer, and routed to the Ultrasound Advice transducer. Broadband noise was raised in level until the

BAER was no longer observed. Keeping masker spectrum level constant, BAERs were recorded when this noise was high-pass filtered at 80, 56.6, 40, 28.3, and 20 kHz (i.e., at half octave intervals from 20 to 80 kHz). A response was also recorded with no masking noise. High-pass responses were obtained in the order of decreasing masker cutoff fre- quency.

806 d. Acoust. Soc. Am., Vol. 96, No. 2, Pt. 1, August 1994 R. Burkard and C. F. Moss: Bat BAERs to clicks and FM sweeps 806

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 128.220.159.2 On: Wed, 11 May 2016 20:15:42

Page 7: The brain-stem auditory-evoked response in the big brown ... brain-stem... · The brain-stem auditory-evoked response in the big brown bat (Eptesicus fuscus) to clicks and frequency-modulated

1.9o -'

: .

• -

[: 1.70 '

>,, : 0 - c- 1.50 (D

• -

_J

1.30 -

(D : Q_ :

1.10 - .

-

: -

.

0.90 -

a Bat22 • Bat23 , Bat24 o Bat25

2•5 ......... 4'5 ......... 6'5 ......... 8'5 .... t•nasked High-Pass Masker Cutoff (kHz)

FIG. 12. Peak latencies for waves ia and ib are plotted across high-pass masker cutoff frequency. Each symbol represents a peak latency for one bat. The solid lines connect mean peak latencies at each masker cutoff fre- quency. Small symbols represent latencies for wave ia, while larger symbols represent latencies for wave ib.

C. Results

In all four bats, there was a weak response to the 90 dB pSPL FM sweep in the presence of 80-dB SPL noise, but there was little or no response observed when a broadband noise level of 85 dB SPL was presented. Responses from one bat to a high-pass masker cutoff series are shown in Fig. 11. The BAERs in this bat are representative of those observed in all four animals, and several points are noteworthy. First, wave ia disappeared for high-pass masker cutoff frequencies of 56.6 kHz and below. Second, the BAER was typically observed for high-pass maskers down to 28.3 kHz. When the high-pass masker cutoff was set to 20 kHz, the response disappeared. Third, the latency of wave ib decreased slightly when the high-pass masker cutoff frequency was reduced from 80 to 56.6 kHz. It is interesting that this occurred at the cutoff frequency where wave ia disappeared. Fourth, two peaks clearly were present in the latency region associated with wave iv for the unmasked condition in this bat, and other bats in this series. Note that as the high-pass masker cutoff frequency decreased to 40 kHz, the wave iv response tended to become more like that seen to click stimuli.

Figure 12 plots waves ia and ib across masker cutoff frequency. Each symbol represents the latency for one bat, with the solid line representing the mean data. In all four bats, wave ia was observed for the unmasked condition, and for the 80-kHz cutoff frequency. Wave ia latency increased in the presence of the 80-kHz high-pass masker. There was no observable wave ia for masker cutoff frequencies below 80 kHz for any animal, with the exception of the ia response for bat 23 with a 28-kHz cutoff frequency. Wave ib latency was similar for the unmasked condition and for the 80-kHz high- pass masker. In three of four bats, there was a decrease in wave ib latency when the high-pass masker cutoff frequency was decreased from 80 to 56.6 kHz. Finally, wave ib latency

increased with decreasing cutoffs frequency below 56.6 kHz.

v. DISCUSSION

In the present investigation, we evaluated the effects of click and FM-sweep level on the latency and amplitude of the BAER peaks recorded from the big brown bat, an animal that uses its auditory system for biosonar. Below, we discuss these findings in the context of far-field BAER data from other mammalian species.

Click-evoked BAERs have been extensively investi- gated, and we conducted experiment 1 to permit comparison of the bat BAER with other animals. In the big brown bat, we found an increase in peak amplitude and a decrease in peak latency with increasing click level. These effects of click level have been reported for a variety of species, in- cluding humans, gerbils, rats, and monkeys (Cox, 1985; Burkard and Voigt, 1989; Burkard and Voigt, 1989; Fria et al., 1982). We found a decrease in the i-iv interval with increasing click level. In other mammalian species, such as gerbils, cats, monkeys and humans (Burkard et al., 1989; Huang and Buchwald, 1978; Fria et al., 1982; Stockard et al., 1979; Cox, 1985), there is little change in the inter- wave intervals with click level. In fact, careful inspection of published tables and figures suggests a slight increase in the human I-V (Cox, 1985; Stockard et al., 1979) and gerbil i-v (Burkard et al., 1989) intervals with increasing click level. In the rat, there is little change in the i-iv interval with click level, while the i-v interval appears to increase slightly with decreasing click level, from 3.16 ms at 90 dB pSPL to 3.24 at 60 dB pSPL (Burkard et al., 1990). Thus, at least one study in the rat suggests a small decrease in an interpeak interval with increasing click level, and parallels the findings of the present investigation in the bat. However, the magnitude of the decrease in the rat i-v interval with increasing click level is substantially smaller than that observed in the present in- vestigation.

Experiment 2 used FM sweeps as stimuli, simulating the biosonar sounds these animals use for echolocation. The data

for this experiment showed stimulus dependencies that were similar to those found for the click stimuli. Specifically, with increasing FM-sweep level, there was a decrease in peak latency, an increase in peak amplitude and a decrease in the i-iv interval. Thus, in the big brown bat, the change in i-iv interval with increasing level appears for both click and simulated biosonar stimuli.

In a related study, Simmons et al. (1990a) studied N1 (eighth nerve) and N4 (lateral lemniscus) potentials recorded with a tungsten electrode placed on the dorsal surface of Eptesicus inferior colliculus. They used 1-ms duration, 110-20 kHz FM sweeps as stimuli. They found slopes of -13 and -18 /xs/dB at 15 dB above threshold for N1 and N4, respectively. These slopes are in good agreement with our overall slopes for clicks and FM sweeps in the present investigation. Furthermore, the steeper slope of N4 as com- pared to N1 would result in a larger N1-N4 interval at lower click levels, and this is consistent with the results of the present study.

Frequency modulation is present in human speech, and previous studies have evaluated the human BAER to FM

807 J. Acoust. Soc. Am., Vol. 96, No. 2, Pt. 1, August 1994 R. Burkard and C. F. Moss: Bat BAERs to clicks and FM sweeps 807

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 128.220.159.2 On: Wed, 11 May 2016 20:15:42

Page 8: The brain-stem auditory-evoked response in the big brown ... brain-stem... · The brain-stem auditory-evoked response in the big brown bat (Eptesicus fuscus) to clicks and frequency-modulated

stimuli. Trenque and Gazeaud (1978) presented FM sweeps at a rate of 30 Hz, and found that wave V latency was greater for rising FM sweeps. Lenhardt (1982) studied responses to FM sweeps, evaluating wave V latency across sweep fre- quency, direction of sweep and rate of FM-sweep presenta- tion. In general wave V latency increased with increasing rate of presentation. For slow rates of presentation, the la- tency of wave V to rising FM sweeps was greater than the latency observed to falling FM sweeps. For a stimulation rate of 100 Hz, the latency of wave V to falling FM sweeps was greater than the latency observed to rising FM sweeps. The shorter latency to falling FM sweeps at lower presentation rates is consistent with the view of the BAER as on onset

response, and a rising sweep has a lower frequency energy distribution at burst onset than a falling sweep. The shorter latency for the falling sweep is consistent with the shorter traveling-wave delay to basal cochlear regions. The reversal of this trend at higher rates of stimulation is perplexing, and suggests that there may be a frequency dependence of rate- induced wave V latency shift. Differences between the sweep rate, bandwidth and presentation rate of stimuli used in these studies preclude direct comparisons with our data in the bat. However, it is noteworthy that no previous study using FM sweeps has reported a change in interpeak intensity with stimulus level.

The dependence of interaural stimulus level on the re- sponse of single (binaural) cells of the inferior colliculus of the Mexican free-tailed bat has been described by Pollak (1988). In this study, interaural level differences evoking maximum neural discharge could be offset by advancing or delaying the arrival time of the ipsilateral (inhibitory) stimu- lus. Time-intensity trading values for single cells in the bat inferior colliculus ranged from -10 to -180/xs/dB, with an average value of -47 /xs/dB. Differences in time-intensity trading from the single-unit work of Pollak (1988; mean val- ues of -47 /xs/dB) and the present investigation (wave iv slopes of -18 and -13 /xs/dB for clicks and FM sweeps, respectively) could be due to differences in bat species, dif- ferences in the acoustic stimuli and their delivery (free field versus dichotic), or differences in time-intensity trading for single-unit responses and gross potentials. A comparison be- tween Pollak's data and our own is constrained further by our limited understanding of the generators of the individual waves of the bat BAER. However, data in other mammals (e.g., in the cat: Buchwald and Huang, 1975; Huang and Buchwald, 1977; Achor and Starr, 1980a,b) suggest that wave iv reflects activity at or below the level of the rostral pons and midbrain.

The click-evoked BAERs observed in experiment 1 and the FM-sweep BAERs observed in experiment 2 were very similar in waveform morphology. One noteworthy difference in these responses is the appearance of wave ia at higher FM-sweep levels, but not at higher click levels. Wave ia appears to represent a response to the high frequency energy present at FM-sweep onset. This interpretation is supported by the results of experiment 3, which showed that wave ia was eliminated by the presence of masking noise high passed at 56.6 kHz or below, while wave ib was present for masker frequencies as low as 28.3 kHz. At lower click levels, the

energy at these high frequencies appears to be inadequate to elicit a BAER, and wave ib appears to be the result of syn- chronous activity in more apical cochlear regions. Thus to low level FM sweeps, we obtained an onset response in more apical cochlear regions, perhaps to energy contained in the middle of the FM sweep. At higher levels, there were two distinct onset responses: one corresponding to basal regions excited by the high-frequency energy at FM sweep onset (wave ia), and a second in more apical cochlear regions in response to lower-frequency energy (and occurring later in the FM sweep: wave ib). It seems that the appearance of wave ia requires that there be adequate high-frequency en- ergy in the stimulus, and that this spectral energy be tempo- rally distributed, as we found no evidence of wave ia to click stimuli. For 90 dB pSPL stimuli, mean wave i latency to clicks was 0.98 ms, while mean wave ia latency to FM sweeps was 1.16 ms. The longer latency for wave ia presum- ably reflects the effects of FM-sweep rise time (0.25 ms) and effective stimulus level on BAER peak latency.

We observed one consistent difference between the FM-

sweep responses obtained in experiments 2 and 3. Specifi- cally, there was a prominent peak following wave iv in BAERs recorded in experiment 3, which was not commonly observed in the BAERs to FM sweeps obtained in experi- ment 2. We used identical FM stimuli and recording param- eters in the two experiments. However, we used different attenuators in experiment 3, in order to maximize high- frequency energy. It is possible that the additional high- frequency energy in the acoustic spectrum of the FM sweep in experiment 3 can account for the difference in waveform morphology observed in experiments 2 and 3. Support for this notion comes from close examination of the high-pass masking data. A decrease in the high-pass masker cutoff fre- quency attenuated this later peak in the wave iv complex (see Fig. 11). Had the two-peaked wave iv represented a parallel of the wave ia-ib complex, the earliest peak in wave iv would have disappeared with decreasing high-pass masker cutoff frequency. However, the later peak disappeared with decreasing masker cutoff frequency, and hence it is unlikely that this complex represents a wave iv pair, i.e., an early wave iv responding to high-frequency energy at sweep onset, and a later wave to lower-frequency energy occurring later in the sweep waveform.

The pattern of BAER peak amplitude growth with in- creasing stimulus level differed for clicks and FM sweeps (see Figs. 5 and 10). The amplitude of waves i, ii, and iv increased monotonically with increasing click level. In con- trast, using FM-sweep stimuli, the amplitude of waves ii and iv plateaued at and above 70 dB pSPL. Interestingly, waves ia and ib showed a monotonic increase in mean peak ampli- tude with increasing FM-sweep level. Thus the nonmono- tonic behavior of waves ii and iv does not appear to be the result of a saturating eighth nerve response, as wave ia and ib amplitudes did not show amplitude saturation for the levels used in this study. It is interesting to note that the wave ii and wave iv saturation occurred at an FM-sweep level just above that where a wave ia response was observed in most of the animals. It is possible that the activity producing peak ia leads to activity in the generators of waves ii and iv that is

808 J. Acoust. Soc. Am., Vol. 96, No. 2, Pt. 1, August 1994 R. Burkard and C. F. Moss: Bat BAERs to clicks and FM sweeps 808

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 128.220.159.2 On: Wed, 11 May 2016 20:15:42

Page 9: The brain-stem auditory-evoked response in the big brown ... brain-stem... · The brain-stem auditory-evoked response in the big brown bat (Eptesicus fuscus) to clicks and frequency-modulated

out of phase with the activity produced in these generators by wave ib. The vector sum of these two contributions could then lead to phase cancellation, possibly resulting in the ob- served saturation of waves ii and iv. This idea could be tested

by using high-pass masking procedures, or by varying the frequency range of the FM sweep. The prediction is that the saturation of wave ii and iv amplitudes would disappear when one eliminated activity from the basal regions of the cochlear partition.

In the echolocating bat, clicks and FM stimuli give rise to BAER latency/intensity functions with similar slopes. The similarity between latency/intensity function data for the clicks and FM sweeps is noteworthy, because the spectral characteristics of these two stimuli are similar, while the temporal characteristics are distinct (see Figs. 1 and 6). In addition, the latency-intensity function slopes of the wave i and wave iv responses to FM sweeps and clicks compare well with the behavioral data on time-intensity trading for the perception of target range in Eptesicus fuscus (Simmons et al., 1990a,b). Unlike other mammals, (e.g., humans and gerbils), the i-iv interval in the bat decreases with increasing stimulus amplitude, a finding which may reflect a specializa- tion in the brain stem of this animal for the processing of acoustic information contained in sonar cries and echoes.

One might ask whether the different latency/intensity slopes for waves i and iv (and resulting level dependence of the i-iv interval) reflects a species specialization. If so, then this spe- cialization may be related to echolocation. It is interesting that the i-iv interval appears to remain relatively constant at high stimulus levels, and to increase with decreasing stimu- lus level for click levels below 70 dB pSPL and for FM- sweep levels below 60 dB pSPL. The relatively constant i-iv interval occurs at stimulus levels corresponding to those ob- served for emitted biosonar signals (Kick, 1982; Griffin, 1958). in contrast, for the lower stimulus levels, which would correspond to the sound level of returning echoes, there appears to be level-dependent i-iv interval increases. The changes in interwave interval across stimulus level were observed for both clicks and FM sweeps, suggesting that if this is indeed a specialization, it is not dependent on the temporal characteristics of the stimulus.

We have several questions left unanswered that can be evaluated empirically in future studies. First, are BAERs to FM sweeps dependent on the bandwidth and direction of the sweeps, and do changes in these parameters influence the ia-ib complex? Second, does the double-peaked wave iv to FM-sweep stimuli arise from the high-frequency energy of the FM sweep? Third, does the decrease of the i-iv interval with increasing level observed in the bat result from a spe- cies specialization or from the stimulation and/or recording parameters? To our knowledge, the only other far-field BAER bat study reported in the literature was conducted by Belknap and Suthers (1982), who studied BAERs in the megachiroptera bat species Rousettus aegyptiacus. Due to differences in bat species, their use of toneburst stimuli and the response variables investigated, the data of Belknap and Suthers does not shed light on the issue of species special- ization.

In this investigation, we used stimuli with high-

frequency energy extending above 80 kHz. Previous work in humans and gerbils on the level dependence of human and gerbil BAERs used stimuli with spectral content which was considerably lower in frequency than that used in the present studies. The spectral content of the signal may have contrib- uted to the change in i-iv interval with stimulus level shown here. Alternatively, the free-field and thus binaural stimula- tion may have contributed to the level dependence of the i-iv interval. Gerbils are able to hear tonal stimuli up to or beyond 60 kHz (Ryan, 1976). A comparison of the bat and gerbil BAER to FM sweeps with similar bandwidths, under identical stimulation and recording parameters would allow us to determine whether the observations reported in the present investigation are the result of stimulus bandwidth, recording parameters, or a species specialization in the bat.

Vl. SUMMARY

(i) The latency/intensity function slopes of the bat BAER are steeper for wave iv than wave i for clicks and FM sweeps. This results in a i-iv interval decrease with increas- ing stimulus level.

(ii) The latency/intensity function slopes are similar for clicks and for 100-20 kHz FM sweeps.

(iii) For 100-20 kHz FM sweeps, two wave i peaks (ia and ib) are observed at high stimulus levels. The first peak is thought to arise from basal cochlear regions in response to the high-frequency energy at the onset of the FM burst, while the second is thought to arise from more apical cochlear regions in response to the lower-frequency energy occurring in the later portion of the FM burst. This is supported by the elimination of wave ia when high-pass masking noise cutoff is reduced below 80 kHz, while wave ib is observed for high-pass maskers as low as 28.3 kHz.

(iv) This is the first investigation using simulated species-specific vocalizations as stimuli to elicit BAERs. The effect of stimulus level on the BAER to FM sweeps parallels that observed to clicks, with the exception of the appearance of a peak occurring prior to wave ib at the higher FM sweep levels. The change in i-iv (and ib-iv) interval with increas- ing click and FM-sweep level could reflect a specialization specific to echolocation behavior. However, substantiation of such a thesis requires further investigation.

ACKNOWLEDGMENTS

Portions of these data were presented at the 1993 mid- winter meeting of the Association for Research in Otolaryn- gology. We acknowledge LeCroy Instruments for donating the FFT software used in our acoustic calibrations. We thank

Doreen Valentine and Jennifer Melcher for commenting on an earlier draft of this manuscript, and acknowledge the two anonymous reviewers of this manuscript. This work was par- tially supported by NIH NINDCD DC00399 (RB) and an NSF Young Investigator Award IBN-9258255 (CFM).

Achor, L., and Starr, A. (1980a). "Auditory brainstem responses in the cat. I. Intracranial and extracranial recordings," Electroencephalogr. Clin. Neurophysiol. 48, 154-173.

Achor, L., and Starr, A. (1980b). "Auditory brainstem responses in the cat. II. Effects of lesions," Electroencephalogr. Clin. Neurophysiol. 48, 174- 190.

809 J. Acoust. Soc. Am., Vol. 96, No. 2, Pt. 1, August 1994 R. Burkard and C. F. Moss: Bat BAERs to clicks and FM sweeps 809

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 128.220.159.2 On: Wed, 11 May 2016 20:15:42

Page 10: The brain-stem auditory-evoked response in the big brown ... brain-stem... · The brain-stem auditory-evoked response in the big brown bat (Eptesicus fuscus) to clicks and frequency-modulated

Belknap, D., and Suthers, R. (1982). "Brainstem auditory evoked responses to tone bursts in the echolocating bat, Rousettus," J. Comp. Physiol. 146, 283-289.

Buchwald, J., and Huang, C. (1975). "Far-field acoustic response: Origins in the cat," Science 189, 382-384.

Burkard, R., Feldman, M., and Voigt, H. (1990). "Brainstem Auditory- evoked response in the rat. Normative studies, with observations concern- ing the effects of ossicular disruption," Audiology 29, 146-162.

Burkard, R., and Hecox, K. (1983). "The effect of broadband noise on the human brainstem auditory evoked response. I. Rate and intensity effects," J. Acoust. Soc. Am. 74, 1204-1213.

Burkard, R., and Voigt, H. (1989). "Stimulus dependencies of the gerbil brain-stem auditory evoked response (BAER). I: Effects of click level, rate and polarity," J. Acoust. Soc. Am. 85, 2514-2525.

Cox, L. (1985). "Infant assessment: Developmental and age-related consid- erations," in The Auditory Brainstem Response, edited by J. Jacobson (College-Hill, San Diego), pp. 297-316.

Fattu J., and Suthers, R. (1981). "Subglottic pressure and the control of phonation in the echolocating bat, Eptesicus," J. Comp. Physiol. 143, 465-475.

Fria, T., Saad, M., Doyle, W., and Cantekin, E. (1982). "Auditory brain stem responses in rhesus monkey with otitis media with effusion," Otolaryngol. Head Neck Surg. 90, 824-830.

Griffin, D. (1958). Listening in the Dark (Yale U. P., New Haven). Henry, K. (1979). "Auditory nerve and brain stem volume-conducted po-

tentials evoked by pure-tone pips in the CBA/J laboratory mouse," Audi- ology 18, 93-108.

Huang, C., and Buchwald, J. (1977). "Interpretation of the vertex short latency acoustic response: A study of single neurons in the brainstem," Br. Res. 137, 291-303.

Huang, C., and Buchwald, J. (1978). "Factors that affect the amplitudes and latencies of the vertex short latency acoustic responses in the cat," Elec- troencephalogr. Clin. Neurophysiol. 44, 179-186.

Kick, S. (1982). "Target-detection by the echolocating bat, Eptesicus fus- cus," J. Comp. Physiol. 145, 431-435.

Kick, S., and Simmons, J. (1984). "Automatic gain control in the bat's sonar receiver and the neuroethology of echolocation," J. Neurosci. 4, 2705- 2737.

Kraus, N., Smith, D., Reed, N., Willott, J., and Erwin, J. (1985). "Auditory brainstem responses in non-human primates," Hear. Res. 17, 219-226.

Lenhardt, M. (1982). "Wave V latency and chirp (linear frequency ramp): Repetition rate," Audiology 21, 425-432.

Pollak, G. (1988). "Time is.traded for intensity in the bat's auditory sys- tem," Hear. Res. 36, 107-124.

Ryan, A. (1976). "Hearing sensitivity of the Mongolian gerbil, Meriones unguiculatus," J. Acoust. Soc. Am. 59, 1222-1226.

Simmons, J. (1973). "The resolution of target range by echolocating bats," J. Acoust. Soc. Am. 54, 157-173.

Simmons, J. (1989). "A view of the world through the bat's ear: The for- mation of acoustic images in echolocation," Cognition 33, 155-199.

Simmons, J., Moss, C., and Ferragamo, M. (1990a). "Convergence of tem- poral and spectral information into acoustic images of complex sonar tar- gets perceived by the echolocating bat, Eptesicus fuscus," J. Comp. Physiol. 166, 449-470.

Simmons, J., Ferragamo, M., Moss, C., Stevenson, S., and Altes, R. (1990b). "Discrimination of jittered sonar echoes by the echolocating bat, Eptesicus fuscus: The shape of target images in echolocation," J. Comp. Physiol. 167, 589-616.

Stockard, J., Stockard, J., Westmoreland, B., and Corfits, J. (1979). "Brain- stem auditory-evoked responses. Normal variation as a function of stimu- lus and subject characteristics," Arch. Neurol. 36, 823-831.

Trenque, P., and Gazeaud, F. (1978). "Latency of brain stem responses to chirps (linear frequency-ramp bursts)," Audiology 17, 213-231.

810 J. Acoust. Soc. Am., Vol. 96, No. 2, Pt. 1, August 1994 R. Burkard and C. F. Moss: Bat BAERs to clicks and FM sweeps 810

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 128.220.159.2 On: Wed, 11 May 2016 20:15:42