Brain Research, 150 (1978) 29--44

29

Elsevier/North-Holland Biomedical Press

THE AUDITORY

MIDBRAIN

OF A MARSUPIAL:

THE BRUSH-TAILED

POSSUM (TRICHOSURUS VULPECULA)

LINDSAY M. AITKIN, BRIAN M. H. BUSH* and G. RICHARD GATES**

Department of Physiology, Monash University, Clayton, Victoria (Australia) (Accepted November 10th, 1977)

SUMMARY A microelectrode survey was made of the midbrain auditory nuclei of the brushtailed possum (Trichosurus vulpecula), a common Australian marsupial. Information was sought on the tuning characteristics of individual neurones, tonotopic organization and mechanisms of sound localization. It was felt that such information would be of use in future studies of the development and evolution of mammalian hearing. Twelve possums were anaesthetized with ketamine and chloralose-urethane, and recordings were made of extracellular unit discharges in the inferior colliculus during monaural and binaural tonal stimulation. The inferior colliculus of the possum consists of a central nucleus - - a darkly stained, densely packed group of cells - - flanked laterally by an external nucleus with a lower density of paler cells. Tonotopic organization was demonstrated by discretelytuned elements in the central nucleus, but was not observed in the external nucleus. In the latter region broad and irregular tuning was commonly seen. Most units in both divisions were influenced by binaural stimuli, with patterns of binaural interaction similar to those observed in the cat inferior colliculus. Cells influenced by changes in the interaural time and intensity difference were commonly observed, but only a subclass of these were suited in sensitivity for sound localization. In general, the midbrain auditory system of the possum was similar in unit discharge characteristics and organization to those of the eutherian mammals commonly studied.

INTRODUCTION The brush-tailed possum (Trichosurus vulpecula) is a marsupial of widespread * On leave from The Department of Physiology, University of Bristol, Bristol, Great Britain. ** Current address: Department of Psychology, University of Melbourne, Parkville, Victoria, Australia.

30

Fig. 1. a: brush-tailed possum (Trichosurus vulpecula) (from ref. 34, with permission), b: possum brain. RB, restiform body; CN, cochlear nucleus; VIII, auditory nerve; BP, brachium pontis: BC, brachium conjunctivum;IC, inferior colliculus; SC, superior colliculus; MGB, medial geniculate body; LGB, lateral geniculate body. Calibrations: for A, 20 cm (approximate); for B, 1 cm.

31 distribution in Australia a4. It is also a denizen of Australian suburban environments where it coexists with an introduced mammal of similar size - - the domestic cat. Like the latter, it is active nocturnally and has a substantial repertoire of vocalizations which it uses in relation to territorial, parental and mating behaviouraT. This vocal ability is coupled with large and mobile pinnae (Fig. la) and prominent midbrain auditory structures (Fig. lb). No behavioural or physiological tests of auditory capacity have been published for Trichosurus, but studies of the North American marsupial opossum (Didelphis virginiana) indicate that this animal possesses a broad hearing range with rather insensitive thresholds 25. Apart from the latter observations, little is known of auditory function in noneutherian mammals. Brief studies of peripheral auditory function have been published for the monotreme platypus ( Ornithorynchus anatinus) 14, echidna (Tachyglossus aculeatus) 4 and the opossum2L A short account of the auditory region of the cerebral cortex of Didelphis 21 and a larger study of the effects of neocortical removal on sound localization in this species 26 are available. However, most sensory neurophysiological studies of metatherians have been concerned with sensory systems other than that for hearing (e.g. refs. ll, 17, 18, 24, 31 and 35). Studies of the sensory pathways of these mammals are important for two major reasons. First, information may be obtained concerning brain evolution in an alternative stream to that of eutherian mammals. The principles underlying auditory processing in the mammalian brain, which have been elaborated in numerous studies of the cat, dog and monkey, require examination in mammalian species other than placentals. Secondly, basic questions about central neuronal development may be highly amenable to study with marsupials, since much growth and differentiation of the young marsupial occurs outside the uterus, in the pouch a4. The present study was directed towards the former consideration, and examines the physiological organization of the possum midbrain auditory region in relation to sound frequency and localization. METHODS Twelve possums (5 male and 7 female) with weight range from 1.5 to 2.8 kg, were used in this study. Anaesthesia was induced with an injection of ketamine hydrochloride (30 mg/kg i.m.) followed after 30 min. by 1 ~ chloralose - 10~ urethane solution (5 ml/kg i.p.). Atropine sulphate (0.2 mg i.m.) was also administered to reduce the ketamine-induced salivation which was occasionally very marked in this species. Rectal temperature was measured with a thermistor probe and maintained with a DC feedback blanket unit at 36 °C. Artificial respiration via a tracheal cannula was occasionally employed to provide a ventilation level approximately equal to a normal resting volume. Experiments were carried out in a sound-attenuated room. The inferior colliculus of the possum was exposed dorsally by appropriate craniotomy, without the need of brain tissue aspiration. A substantial gap existed between cerebrum and cerebellum and no bony tentorium was present. Instead, a large venous sinus lay immediately dorsal to the inferior colliculus, and microelec-

32 trodes were inserted through this with reference to landmarks offered by the cerebellum and cerebrum. All successful penetrations were histologically confirmed. The prominent pinnae (Fig. la) of Trichosurus narrow to long bony external auditory meati. Because of the oblique angle at which the meati sit in the skull, the animals' heads could not be supported with a conventional stereotaxic apparatus. Instead, the head was held firmly in place by a steel plate screwed to the bone overlying the long nasal sinus of the upper jaw. Calibrated sound channel assemblies. with tapered metal tubes terminating in conical plastic pieces, were positioned close to the eardrums. The structure of the outer ear canal made the incorporation of probe tubes and microphones difficult - - instead, closed field calibrations were used. These were derived from a B and K 0.25 in. microphone (type 4135) used in conjunction with a B and K Heterodyne Analyzer (Type 2010), and were expressed re 0.0002 dyne/sq. cm from 100 Hz to 30 kHz (SPL). Techniques for tonal stimulation, extracellular microelectrode recording and histological reconstruction were identical to those previously reported from this laboratory z,a. Threshold tuning curves were constructed by determining aurally and with the aid of an oscilloscope, the SPL needed to cause a just-detectable, but regular, discharge for each frequency. Taped data were analyzed off-line, after timing with a real-time clock at a resolution of 100 #sec, by a Nova computer programmed to compile response histograms, interspike-interval histograms and selected spike counts. RESULTS

Cytoarchitecture of the inferior colliculus The inferior colliculus of the possum Trichosurus in Nissl sections is similar in appearance to that of the opossum DidelphislL Two major areas may be distinguished cytoarchitectonically: a central and medial aggregation of darkly stained, densely packed cells, and a lateral and rostral population of pale, sparsely distributed neurones. These two regions are similar in disposition and cytoarchitecture to the central and external nuclei of the cat inferior colliculus, respectively7 and the terminologies of ICC and ICX will be used in this study. Similarly, the dorsal-most part of the inferior colliculus will be described as the pericentral nucleus (1CP) because of the small size and dense packing of some of the cells in this region. However, ICP here is not as clearly demarcated from ICC as it is in the cat 7. The nuclei of the inferior colliculi are clearly separate from the superior colliculus at medial sagittal planes owing to the presence of a cell-sparse area lying between these structures (Fig. 2). At more lateral levels a gradual transition occurs between ICX and the tegmental areas lying deep to the superior colliculus, and sharp borders are difficult to define here (Fig. 3). This lateral part of the inferior colliculus and its rostral extension deep to the superior colliculus presumably correspond to the intercollicular area 3° or intercollicular terminal zone 2s described for other species.

Tonotopic organization Long penetrations were made through ICC in four experiments and through

33

POSSUM 1 SWISH SWISH SWISH

5 0* 2" 7 0

I

:)• -5"0 O• 3• SWISH

18"5-

19-0v

E E

T I-ra I.t.l

200-

21.0 ~

0"2

I

0-5

~

~

~

r

I

1"0

1

r

2"0

FREQUENCY

'

I

5"0

~

~

~

'

I

10"0

I

20"0

(kHz)

Fig. 2. Upper: sagittal section (2.3 mm from sagittal plane) at level of recording track, possum 1. Numbers to right of section: best frequencies of units (*) or unit clusters. L. F. SWISH, background discharges to low frequency tones; SC, superior colliculus; ICP, pericentral nucleus of inferior colliculus; ICC, central nucleus of inferior colliculus. 50/~m thickness, thionine stain. Lower: best frequencies of units and unit clusters plotted against micrometer electrode drive reading in mm.

34 ICX in three others. Five further penetrations sampled unit discharges in both subdivisions. The recording sites of a total of 68 units were identified ; 39 were derived from 1CC, 15 from ICX, 5 from ICP, 3 from the adjacent tegmentum deep to superior colliculus, and 6 from border regions between these subnuclei. Central nucleus A penetration through ICC is illustrated for possum 1 (Fig. 2, upper). No recordings were made in ICP, but background discharges to low frequency tones (swish) were recorded in border regions. Following these, single unit discharges and clusters of units were isolated which exhibited progressively higher best frequencies (BF) as a function of depth. At the ventrocaudal border of ICC a return to low frequency swish was observed and a lesion was made. Thus, after an initial fluctuation about a low best frequency value, a basically linear relationship between best frequency (plotted on a log scale) and depth could be demonstrated (Fig. 2, lower). The direction of this best frequency progression is similar to those observed in the cat central nucleus of the inferior colliculus23. Two of the remaining penetrations through ICC exhibited similar trends and sampled from similar regions of the nucleus. The third track passed through ICC at a medial sagittal plane and revealed a restricted (BF 13-21 kHz) but orderly sequence. Although it was always possible to accurately estimate best frequency for units in ICC, it was apparent that extremely sharply tuned units similar to those reported for the cat inferior colliculus5 were infrequently encountered. External nucleus No similar progressions could be observed in ICX, and unequivocal best frequencies could only rarely be determined by aural and oscilloscope monitoring. With cells in this region it was often necessary to plot a tuning curve in order to define a best frequency. Tuning curves so obtained were usually broad, occasionally spanning 6 octaves or more. An example of a penetration through ICX is shown in Fig. 3A. Nine units and two unit clusters were studied and tuning curves were determined for four units (Fig. 3B). These curves are all broad and best frequencies refer to the most sensitive dips in generally fiat curves. For this track the majority of units was more effectively driven by ipsilateral than contralateral stimuli (Fig. 3A), and both onset patterns (Fig. 3C) and sustained patterns (Fig. 3D) could be demonstrated with different units. The 'best' frequencies revealed in this penetration are not arranged in any obviously systematic pattern. In view of the broadness of tuning in those units so studied, it seems unlikely that penetrations at other, possibly more optimal, angles could show a strict tonotopic organization of ICX. Discharges to monaural and binaural stimuli

Binaural interactions were studied with 59 units, 39 of which were located in ICC and the remainder in ICX, ICP and border areas. For both groups a majority

A

POSSUM 7

~ ~ ' - ~

~

~

~ ~ : ~ ~ ~ t :~i ~;:~,~,~.,~"t."~

~J,~:~

//0"3-1kHz

//O-SkHz

\

BROAD

:..

U60E lkHz

1ram

J•il

[ -

7-5 (IPSI)

cx.._

~

~(IP511 _ ~

- - ' ~

(31)

(47)

xJ x"~"x - - - - - - - - - x ~ - - / ~ x

z

\

--,\x/X

(~x

._J LLI

11"oi° x

(12)

o'.2 o'.3o.'4o:0i

o'.1

46

lb

io

FREQUENCY (kHz)

C

UNIT 7-5 lOkHz N= 321

8

0j

I

0

D

UNIT %7 2kHz

BIN 50dB

. . . . . . I. . .

N=321 IPSI 50dB

t

0.5 1 TIME (sec)

I

0 4

015 1 TIME (sec)

Fig. 3. A: sagittal section, 3.5 mm from sagittal plane, at level of recording track, possum 7. U1, U2 etc., unit numbers; EE, both contralateral and ipsilateral stimuli excite; OO/F, no response to contralateral or ipsilateral, but excitation to binaural; EO and OE, monaural only, excitation to contralateral alone or ipsilateral alone, respectively; IE, contralateral inhibits, ipsilateral excites; BROAD, broad cluster response, no clear best frequency; ICX, external nucleus of inferior colliculus; SCS, superficial layers of superior colliculus. B: threshold tuning curves for 4 units from possum 7, plotted on a relative intensity scale for clarity only. Actual SPL indicated for best frequency for each unit in brackets. 7-1, etc, unit numbers. C and D: response histograms for units 7-5 and 7-7. In these and all subsequent histograms, N = total number of spikes in histogram for 20 stimulus trials; ordinate, spike count per bin; abscissa, time in seconds. Tone duration indicated by time bar beneath each histogram.

36 of units was influenced by binaural stimuli (33/39 ICC; 14/20 remainder of units) as in the cat inferior colliculus 5. Three principal forms of binaural interactions were recognizable and were exhibited by units in both ICC and ICX. For 9 units (BF 0.2-2 kHz) the magnitude of binaural discharge depended upon the time delay between the binaural tones. With a further 24 units stimuli to both ears were excitatory but their interaction was not obviously influenced by the interaural time delay. For these units the magnitude of the binaural response varied from occlusion (where it was intermediate between the magnitudes of the monaural responses) to facilitation (greater than the sum of the monaural responses). Inhibitory interactions were predominant with a third group of 14 units, and in these, contralateral stimuli were excitatory and ipsilateral inhibitory in all but two cases, where the reverse situation prevailed. Detailed analyses were made of the discharge characteristics of 16 units, 11 from ICC and 5 from ICX and border regions. Examples of these interactions are instructive since they reveal that units in the possum auditory midbrain may be as sensitive to binaural cues as those in the cat 1~,29.

Influence of interaural time delay Unit 10-2 (Fig. 4) was isolated in the central nucleus and responded weakly to ipsilateral and contralateral stimuli with an onset discharge. Simultaneous binaural stimulation produced a strong facilitation (Fig. 4A) with the emergence of a sustained discharge following the onset response. The binaural tuning curve (Fig. 4B) indicated that discharges were evoked at threshold over a relatively discrete and very low frequency range. Variations in the interaural time difference produced profound alterations in discharge rate and pattern (Fig. 4A). These effects were most clearly manifest at

LU Z

400 -

A UNIT I 0 - 2

B

I I

90-

80-

o 300-

70-

z n,"

_.1

~m60-

°200-

5o-

LI.I

40"

~,i00.

30o

2O

~ o sOO

200 o 260

IPSI LAGS

660

1o'oo

o.~

o!z o!3 o!4 'o76 'o'.8 FREQUENCY (kHz)

CONTRA LAGS /~sec

Fig. 4. Unit 10-2. A: relationship between total spikes during tone for 20 trials (ordinate) and time delay in/~sec between binaural stimuli at 60 dB SPL and 150 or 200 Hz (abscissa). C (O), I (×) etc., spike count with contralateral or ipsilateral stimuli alone. B: threshold tuning curve for simultaneous binaural stimuli.

37 400 - A

60- B D

350-

UNIT 1-2

/ ~

~. 50co

ALON/

,,, 3 0 0 Z

~l

0

O0

•o 40-

30-

, ,,..,

0.50.7

250n¢ rl

200-

~

~:s ~ ' . 5

FREQUENCY (kHz)

CON42dB ALONE o .........

(-7

~:~(+3}

C

h

CONTRA42dB N=200

JIIIIB.I~L

hl

D j ~

n

m 150-

,J

/

0

100-

.,(

CONTRAN=362dB 70

.

/

"o(+33) E

E

~&alJ- m,

~a..___~_dJ_aa

CONTRA42+PSI75dB N=92 , nBh,ld[l~&a J ~ . L z ~

50-

eSPON

F , ......

lb

2b

3'0 ,'0

s'0's

dB SPL

7b

8'0

~=~=

IPSI 75dB N=8

6

I

~.~a~_~

0%

sec

i

Fig. 5. Unit 1-2. A: relationship between total spike count during tone (ordinate) and intensity (abscissa) for 20 trials of contralateral stimuli alone, or with the contralateral stimulus held at 42 dB SPL and combined with ipsilateral stimuli at the stated intensities. Numbers in brackets are interaural intensity differences in dB (--, ipsi less than contra; --, ipsi greater than contra). C, D and E on curves refer to the histograms on the right hand side; SPON, total spontaneous spike count for equivalent time period to tone-on, summed for 20 trials. B : threshold tuning curve for contralateral stimuli. C-F: response histograms at stated intensities and 1500 Hz. Nomenclature as for Fig. 4. frequencies below 300 Hz. At both 150 Hz and 200 Hz maximal firing occurred when the contralateral stimulus was delayed by 0-200 #sec. With long ipsilateral delays firing dropped to below monaural levels and only a small onset response was elicited. The shapes of these curves and the ranges of interaural time delays for which maximal firing occurred suggest a function for units such as this in low frequency sound localization.

Excitatory-inhibitory binaural units Unit 1-2 (Fig. 5), also from the central nucleus, was excited only by contralateral stimuli which elicited sustained discharges (Fig. 5C, D) over a discrete frequency range (best frequency 1.5 kHz, Fig. 5B). Discharges to increasingly intense contralateral stimuli grew in number until, above 62 dB SPL, a diminution in spike count occurred (Fig. 5A). Simultaneous presentation of ipsilateral stimuli, when the intensity of the latter exceeded the contralateral stimulus by 5-10 dB or greater, led to a reduction of the contralateral response (Fig. 5A, E). This ipsilateral inhibitory input could also be observed as an inhibition of spontaneous activity (Fig. 5F). Reductions in contra-

38

A

100-

UNIT ll-10 IPSI

X 80-

"~X---

"o

40-

20

03 0!a 0!s '0'7' '~.

FREQUENCY(kHz)

I00

q~O)

t

C'

.~80

(

to (-60} ~ 60~(_70)~.,43~

~xkk~

'"

~

\

" ~

e2

"

0-

0 -Io

~PS,6o

IPSI 40

~

;

o

N,.

\ "~

+ CONTRA/'\

xCt(-50)

+co.TRAI\\ 30kHz

,Ps, ~o

+ CONTRA

\

IPSl 85 lkHz J + CONTRA \

\

\

(-2o'~'o I

~-'~(-2o) I

\

X \

~-~oI \

b(+1o) I

I

3'o ,o so dB SPL Fig. 6. Unit 11-10. A: threshold tuning curve for ipsilateral stimuli. B: total onset spike count for 20 trials (ordinate) plotted against contralateral intensity in dB SPL (abscissa). All curves are binaural stimulation, with ipsilateral stimuli held at the stated intensity and frequency for each curve, and contralateral varied according to the abscissa and frequency the same as ipsilateral. Numbers in brackets are interaural intensity differences in dB (--, contra less than ipsi; ~, contra greater than

ipsi). lateral response of unit 1-2 occurred, in part, over a physiological range of interaural intensity differences, suggesting a possible role for units of this type in sound localization involving this cue. More detailed observations on the inhibitory influence of one ear stimulus upon the excitatory responses evoked by the other were made for unit 11-10, located at the border of I C C and I C X (Fig. 6). This unit was one of the few excited by ipsilateral stimuli and inhibited by contralateral tones. It fired only at onset and was very broadly tuned (in c o m m o n with most units of ICX), although it exhibited a clear region of frequency sensitivity (in c o m m o n with most units of ICC), in this case, around 30 kHz. Analyses of inhibitory interactions were made at 1, 20 and 30 kHz. At these three frequencies contralateral stimuli inhibited ipsilateral responses but the ranges of interaural intensity differences over which reductions occurred, and the steepness of the slopes of the functions obtained, varied considerably. Thus with two different ipsilateral levels (40 and 60 dB) at 30 kHz, the inhibitory influence of contralateral stimuli was profound at contralateral levels between 20 and as much as 50 dB less

39

than the ipsilateral intensities (Fig. 6B, left). Such intensity differences would not occur in nature for a cat a6, whose head size is similar to that of Trichosurus. On the other hand, at 20 kHz, spike counts were reduced dramatically by an increase in interaural intensity difference of from --10 to + 10 dB (Fig. 6B, right). The relationship for 1 kHz, far removed from the best frequency regions, lay intermediate in sensitivity between these curves (Fig. 6B, right). An examination of these intensity/spike count functions shows that the sensitivity to a particular binaural cue may vary considerably across the range of frequencies which activates a single unit. The binaural characterization of a unit as being excitatory-inhibitory does not specify a role in sound localization at all effective frequencies for that unit.

8o] A

r B°

UNITIt-5

i 6° ~,

/..... 20JI O1

I 02

B

100-

I 0.3

'

I ' ' ' I I 0'5 0-7 1 2 FREQUENCY

[ 3 (kHz)

r

[ ' t ''

5

/t ~X..,. BIN -

2

x

20

60 401 ;

I/I

60 7b dB SPL

O

80, ~_ 70

s'o

I

I L20

s

30

t x/ BIN SUM/.o '~ RA x

IPSt

0

4'0

I

15 20

80-

//

~60z o ~ 40-

I

10

C 3OkHz

100

0 5 kHz

~ ao

30

7

8'o

3b

9(3

.b

s'0

dB SPL

10

i I

\

500Hz

65dB

BIN

//P~ICONTRAN=26 ,Ps,

N:,,

60 o -J 50

I

o 40

I I

30

2oo0 ~6'oo ~oo 8oo ioo 6 IPSI L A G S

4oo 8~o doo ~o ~oo CONTRA

LAGS

psec Fig. 7. Unit ] 1 - ] 5. A : threshold tuning curve for simultaneous binaural stimuli. B mad C: total onset spike count for 20 trials plotted against intensity at 0.5 k H z (B) a n d 30 k H z (C), for conditions of contralateral alone ( C O N T R A ) , ipsilateral alone (IPSI), simultaneous binaural (BIN) and the algebraic sum of responses to contra and ipsi ( S U M ) . D : total onset spike count versus interaural time delay for 20 trials at 0.5 kHz, 65 dB S P L binaural. Monaural response counts ( C O N T R A , IPSI) s h o w n below tone parameters.

40

Binaurally excitatory units An example of a unit with excitatory input from each ear is illustrated in Fig. 7. This unit was located in ICX and is typical, in broad tuning and binaural input, ~f cells of this region. Of 15 ICX units studied, 13 were activated by both ear stimuli. In contrast, only 16 units (including some units sensitive to interaural time delay) of 39 in ICC and none of the 5 units in ICP were excited in this fashion. All of the latter were influenced only by contralateral stimulation; ipsilateral stimuli had no effects. A detailed analysis of excitatory-excitatory interactions was made for unit I 1 5 (Fig. 7). Simultaneous binaural stimuli elicited spikes over a huge frequency range, in excess of 8 octaves (Fig. 7A). Binaural stimuli were generally more effective than either contralateral or ipsilateral stimuli. Thus, at 0.5 kHz a marked facilitation occurred at low intensities and occlusion at the higher intensities (Fig. 7B), although at 30 kHz binaural stimulation was similar in effectiveness to contralateral stimuli alone (Fig. 7C). As for unit 11 10 (Fig. 6), a similar interaction pattern between binaural stimuli occurred across the effective frequency range for unit 11 5, but the effects of binaural stimulation varied quantitatively. It is interesting to note that the excitatory-excitatory pattern of binaural interaction in conjunction with broad tuning, observed for this and other units of ICX. is also characteristic of auditory units in the pontine nuclei and cerebellum z,9 which are known to receive input from the inferior colliculus in the cat e0. Finally, the relationship between spike count and interaural time delay was examined for unit 11-10 at a low frequency (Fig. 7D). The most obvious feature of the resultant plot is a rapid decrease in spike count with a change of 400 #sec, from ipsilateral lagging by 200 #sec to contralateral lagging by the same value. This would be equivalent, for a single sound source located about the head, to the movement of this source from an oblique angle in the contralateral hemifield through azimuth to a similar position in the ipsilateral hemifield (cf. ref. 8). Neurones such as this would be well suited to signalling sound source location and movement about azimuth. DISCUSSION

Organization of the midbrain auditory regions The possum inferior colliculus has several characteristics which resemble those of the cat. First, in a central region of closely packed and darkly stained cells, unit discharges were recorded which were clearly tuned and for which tonotopic organization was demonstrable in a dorsoventral direction. Secondly, tissue lateral, anterolateral and posterolateral to the central region was composed of loosely arranged, larger, paler cells. Unit discharges isolated in this lateral region were broadly tuned to sound frequency and exhibited poorly defined best frequencies. Units studied in regions bordering the central region were similarly broadly tuned, and an impression was gained that a lack of selectivity towards tone frequency was more common for units in possum inferior colliculus than those in the cat 5. Further study is needed to explore this result, However, it seems appropriate to note here that somatic recep-

41 tive fields of units in the cuneate-gracile complex of the opossum appear to be larger in area than those in the same nuclei of many placental animalsis. The similarities between the morphologies and cell responses of the central and lateral regions of the possum auditory midbrain and the corresponding regions of the inferior colliculus of the cat ~ and squirrel monkey12 suggest a homology, and, accordingly, the terms central (ICC) and external (ICX) have been applied, respectively, to them. The high proportion of binaural units in both regions in the possum strengthens this homology. In general, our results suggest that the midbrain auditory organization of this marsupial is indicative of aural capabilities similar to those of some placental mammals. The external nucleus of the inferior colliculus is considered to be part of the intercollicular area in several species, including Didelphis, and receives a significant spinal somatosensory input 27,2s,~°. In their study of the opossum intercollicular terminal zone, Robards and et al. 27,28 described a purely somatosensory input to this region and no unit responses to auditory stimuli were observed. In contrast, auditory and somatosensory responses are intermingled throughout ICX in the cat, with auditory input more widespread than somatosensory input 3. Unit responses to sound in the lateral region of the inferior colliculus in the present study were very similar to those reported for ICX of the cat; whether they would also be influenced by somatic stimuli remains to be determined. However, it would certainly seem incorrect to consider that the external nucleus of the possum or cat is purely somatosensory in function, as has been argued by Robards et al. 28. Binaural responses and sound localization Trichosurus is a medium-sized (1.50-3 kg) but strong animal, nocturnal and arboreal in habit, with a vegetarian diet. Little published information is available about predators of this species, and they may be restricted to the Powerful owP ~, particularly with young possums, and to more urban and less 'natural' enemies including man. Thus, its employment of sound localization is likely to be in relation to territorial, mating and parental behaviour, rather than for the avoidance of predators or the seeking of prey. This study has provided evidence for midbrain mechanisms of sound localization which are very similar to those utilized by neurones in the cat10,15,29, dog 16, kangaroo rat 32,33 and rabbit 1. Thus, at low sound frequencies, binaural neurones may discharge with a rate and pattern dependent upon the interaural time delay. Similarly, for neurones with best frequencies across the entire frequency range, activity evoked by stimulation of one ear may be progressively inhibited by stimuli of increasing intensity applied to the other. The large head size and pinnae of the possum would lead to interaural time and intensity differences with magnitude ranges at least commensurate with those for the cat. The ranges of interaural time and intensity difference over which spike-count changes were marked were usually, but not always, within the (presumed) physiological range. Thus unit 11-10 (Fig. 6) was much more sensitive to changes in interaural

42 intensity difference at 20 kHz than at the other tested frequencies. Furthermore, mo~t changes in discharge rate at these latter frequencies occurred over interaural intensity difference ranges outside those which might be expected to occur in nature. Consequently, although binaural interaction patterns are specified by the binaural input to a neurone, cells with a particular interaction pattern need not be involved in sound localization (cf. ref. 33). Furthermore, units whose discharge rates do change with changes in interaural parameters within the physiological range may not be signalling information about static sound location, but about sound source movement% Hearing range and vocalizations Winter a7 has described the varieties of vocalizations emitted by possums in different behavioral situations. Sounds described as 'screech', 'hiss', 'click' and 'shookshook' contain frequencies of below 1-12 kHz (the limits of the particular analyzer used). In view of the complex nature and broad frequency range of these vocalizations, it might be expected that auditory sensitivity should be appropriately broad. The present physiological data indicate that best frequency thresholds below 10 dB SPL may be observed for individual units from at least 1 to 30 kHz. The small size of the sample of narrowly tuned units in this study restricts the comparison which may be made; however, as noted previously, sharply tuned units were less commonly encountered in the possum than in the cat inferior colliculus 5. Individual tuning curves (e.g. Figs. 3, 6 and 7) could spread from 1 to 30 kHz, parallelling individual unit best frequency threshold distribution. The frequency range to which Triehosurus may be responsive in its natural environment, as suggested by these physiological measures, closely resembles that described, using behavioural methods, for the North American opossum Didelphis 2~. However, Trichosurus exhibits more sensitive thresholds than those shown by Didelphis. This may be due to the different methods of assessing auditory sensitivity rather than to a real species difference. ACKNOWLEDGEMENTS The authors wish to thank Moyra Farrington, Jill Maplesden and Lynne Hepburn for their help with histology, illustrations and typing. This work was supported by grants from the Australian Research Grants Committee. REFERENCES l Aitkin, L. M., Blake, D. W., Fryman, S. and Bock, G. R., Responses of neurones in the rabbit inferior colliculus. II. Influence of binaural tonal stimulation, Brain Research, 47 (1972) 91-101. 2 Aitkin, L. M. and Boyd, J., Responses of single units in the cerebellar vermis of the cat to monaural and binaural stimuli, J. Neurophysiol., 38 (1975) 418~,29. 3 Aitkin, L. M., Dickhaus, H., Schult, W. and Zimmermann, M., External nucleus of inferior colliculus: auditory and spinal somatosensory afferents and their interactions, J. Neurophysiol., in press.

43 4 Aitkin, L. M. and Johnstone, B. M., Middle ear function in a monotreme: the Echidna (Tachyglossus aculeatus), J. exp. Zool., 180 (1972) 245-250. 5 Aitkin, L. M., Webster, W. R., Veale, J. L. and Crosby, D. C., Inferior colliculus. I. Comparison of response properties of neurons in central, pericentral and external nuclei of adult cat, J. Neurophysiol., 38 (1975) 1196-1207. 6 Altman, J. A., Are there neurons detecting direction of sound source motion, Exp. NeuroL, 22 (1968) 13-25. 7 Berman, A. L., The Brain Stem of the Cat, University of Wisconsin Press, Madison, 1968. 8 Bock, G. R. and Webster, W. R., Coding of spatial location by single units in the inferior colliculus of the alert cat, Exp. Brain Res., 21 (1974) 387-398. 9 Boyd, J. and Aitkin, L. M., Responses of single units in the pontine nuclei of the cat to acoustic stimulation, Neurosci. Lett., 3 (1976) 259-263. 10 Brugge, J. F., Anderson, D. J. and Aitkin, L. M., Responses of neurons in the dorsal nucleus of the lateral lemniscus to binaural tonal stimulation, J. Neurophysiol., 33 (1970) 441-458. 11 Erickson, R. P., Jane, J. A., Waite, R. and Diamond, I. T., Single neuron investigation of sensory thalamus of the opossum, J. Neurophysiol., 27 (1964) 1026-1047. 12 Fitzpatrick, K., Cellular architecture and topographic organization of the inferior colliculus of the squirrel monkey, J. comp. Neurol., 164 (1975) 185-208. 13 Fleay, D., Nightwatchman of Bush and Plain, Churchill, Brisbane, 1968. 14 Gates, G. R., Saunders, J. C., Bock, G. R., Aitkin, L. M. and Elliott, N. A., Peripheral auditory function in the platypus (Ornithorynchus anatinus), J. acoust. Soc. Amer., 56 (1973) 152-156. 15 Geisler, C. D., Rhode, W. S. and Hazelton, D. W., Responses of inferior colliculus neurons in the cat to acoustic stimuli having wide-band spectra, J. Neurophysiol., 32 (1969) 960-974. 16 Goldberg, J. M. and Brown, P. B., Responses of binaural neurons of dog superior olivary complex to dichotic tonal stimuli : some physiological mechanisms of sound localization, J. Neurophysiol., 32 (1969) 613-636. 17 Goodwin, H. E. and Hill, R. M., Receptive fields of a marsupial visual system. 1. The superior colliculus, Amer. J. Optom., 45 (1968) 358-363. 18 Hamilton, T. C. and Johnson, J. I., Somatotopic organization related to nuclear morphology in the cuneate-gracile complex of opossums (Didelphis marsupialis virginiana), Brain Research, 51 (1973) 125-140. 19 Huber, G. C. and Crosby, E. C., A comparison of the mammalian and reptilian tecta, J. comp. NeuroL, 78 (1943) 133-168. 20 Kawamura, K., The pontine projection from the inferior colliculus in the cat. An experimental study, Brain Research, 95 (1975) 309-322. 21 Lende, R. A., Sensory representation in the cerebral cortex of the opossum (Didelphis virginiana), J. comp. Neurol., 121 (1963) 395-403. 22 McCrady, E., Jr., Wever, E. G. and Bray, C. W., The development of hearing in the opossum, J. exp. Zool., 75 (1937) 503-517. 23 Merzenich, M. M. and Reid, M. D., Representation of the cochlea within the inferior colliculus of the cat, Brain Research, 77 (1974) 397-415. 2J, Pubols, B. H., Jr. and Pubols, L. M., Somatic sensory representation in the thalamic ventrobasal complex of the Virginia opossum, J. comp. Neurol., 127 (1966) 19-34. 25 Ravizza, R. J., Heffner, H. E. and Masterton, B., Hearing in primitive mammal I: Opossum (Didelphis virginiana), J. aud. Res., 9 (1969) 1-7. 26 Ravizza, R. J. and Masterton, B., Contribution of neocortex to sound localization in opossum ( Didelphis virginiana), J. Neurophysiol., 35 (1972) 344-356. 27 Robards, M. J., Watkins, D. W. and Masterton, R. B., Nucleus intercollicularis: the somesthetic tectum, Anat. Rec., 178 (1974) 448. 28 Robards, M..I., Watkins, D. W. and Masterton, R. B., An anatomical study of some somesthetic afferents to the intercollicular terminal zone of the midbrain of the opossum, J. comp. Neurol., 170 (1976) 499-524. 29 Rose, J. E., Gross, N. B., Geisler, C. D. and Hind, J. E., Some neural mechanisms in the inferior colliculus of the cat which may be relevant to localization of a sound source in space, J. NeurophysioL, 29 (1966) 288-314. 30 Schroeder, D. M. and Jane, J. A., The intercollicular area of the inferior colliculus, Brain Behav. EvoL, 13 (1976) 125-141. 31 Sousa, A. P. B., Oswaldo-Cruz, E. and Gattass, R., Somatotopic organization and response pro-

44

32 33 34 35 36 37

perties of neurons of the ventrobasal complex of the opossum, J. comp. Neurol., 142 (1971) 231-238. Stillman, R. D., Characteristic delay neurons in the inferior colliculus of the kangaroo rat, /Lvp. Neurol., 32 (1971)404-412. Stillman, R. D., Responses of high-frequency inferior colliculus neurons to interaural intensily differences, Exp. Neurol., 36 (1972) 118-126. Tyndale-Biscoe, H., Life oJ Marsupials, Arnold, London, 1973. Weller, W. L., Barrels in somatic sensory neocortex of the marsupial Trichosurus vulpecula (brushtailed possum), Brain Research, 43 (1972) 11-24. Wiener, F. M., Pfeiffer, R. R. and Backus, A. S. N., On the sound pressure transformation by the head and auditory meatus of the cat, ,4cta oto-laryng. (Stockh.), 61 (1966) 255-269. Winter, J. W., The Behavior and Social Organization of the Brush-Tailed Possum (Trichosurus Vulpecula (Kerr)), Ph.D. Thesis. University of Queensland, 1976.