Hearing of Old World Monkeys (Cercopithecinae)

WILLIAM C. STEBBINS Kresge Hearing Research Institute and Departments of Otorhinolaryngology and Psychology, The University of Michigan, Ann Arbor, Michigan 48104

ABSTRACT The characteristics of normal hearing were examined in the laboratory for seven species of Old World monkeys. Operant conditioning procedures, coupled with standard audiometric testing methods, were used to assess thresholds of hearing, frequency range of hearing, and differential sensitivity to auditory intensity and frequency. To produce tonal stimulation, an animal was trained to touch and maintain manual contact with a contact-sensitive key and to report hearing the tone by lifting his hand from the key; this response was followed by food reinforcement. When the reporting response occurred without the auditory signal, the animal was punished by a short suspension from the experiment. Additional contingencies were added to ensure stable and reliable responding, and threshold and differential acuity determinations were then made. Threshold was defined as the stimulus value responded to correctly 50% of the time. The frequency range of hearing of all the cercopithecoids tested extended from 60 to 40,000 Hz, an octave above the upper bound of 20,000 Hz for man but well below the 60-70,000 Hz limit for some prosimians. Absolute sensitivity for tonal stimulation in the most sensitive frequency range (1-8 kHz) was about 2 X microbars, comparable to that of other primates tested, including man. Thus, the Old World monkey appears only slightly less sensitive than man to small changes in intensity and frequency of acoustic stimulation. At 1000 Hz at 60 dB above the threshold of audibility, his limit of resolution is about 5 Hz for frequency and 2 dB for intensity.

Hearing has not been one of the distin- the use of the external ear in primates as guishing features traditionally included as a device for temperature regulation and in part of the definition of the order Primates visual communication. About hearing we and yet hearing serves them in a variety of are told only that there is an “auditory functions important, if not always critical, function” (Napier and Napier, ’67: 20); to their daily existence. Intraspecies com- no further details are given. munication and the detection of prey and To a considerable extent, this lack of predator are perhaps two of the most widely attention to the auditory capabilities of primates has reflected man’s ignorance of used of these functions. The emphasis for them. In the past few years, the situation criteria in defining this mammalian order has changed, and we now have reliable has always rested on a contra-distinction audiometric data obtained under laboratory between the dramatic evolutionary develop- conditions from representatives of at least ment of their visual apparatus and the nine primate families (Stebbins, ’71). On equally spectacular reduction in their ol- the basis of their frequency range of hear- factory equipment and sensitivity. If audi- ing and sensitivity, the primates divide tion is ever mentioned in discussions of conveniently into two groups - the pro- primate taxonomy, it is usually briefly dis- simians, resembling the other mammals in patched with a short discussion of external and middle ear morphology. There is a their ability to hear high frequencies (to 70 kHz) and with their somewhat reduced marked lack of concern with how primates use these structures in hearing. Even very 1 The research is supported by grants NS 05077 and NS 05785 from the National Institute of Neurological recently, Napier and Napier (’67) discuss Diseases and Stroke.

AM. J. PHYS. ANTHROP.,38: 357-364. 357 358 WILLIAM C. STEBBINS

sensitivity to low and midrange frequencies the laboratory, the monkeys undergo a six- (below 8 kHz); and the Anthropoidea, lim- week quarantine period which includes a ited to frequencies below 40 kHz with some- series of TB tests and a general health ex- what enhanced sensitivity around 1 kHz. amination. Subsequently they are fitted However, even these generalizations have with collar and leash and trained to leave to be considered tentative since the number their cages to enter special restraint chairs of species tested is less than 20. (fig. 1). They are then placed in a sound- in this paper, the previous review (Steb- treated room where the behavioral condi- bins, ’71) has been updated and our find- tioning and hearing testing procedures are ings on seven species (4 genera) of the carried out. Earphones are carefully fitted Cercopithecinae subfamily of Old World over each ear and a special tube and chute monkeys (Cercopithecidae) are presented : are arranged to deliver small whole-diet Macaca arctoides, M. fascicularis, M. mu- pellets directly to the monkey’s mouth (see latta, M. nemestrina, Cercopithecus aethi- fig. 1). The complete details of the equip- ops, Erythrocebus patas, and Papio papio. ment and of the training and testing pro- The differences between species are no cedures have been described elsewhere greater than the intraspecies differences on (Moody, Stebbins and Miller, ’70; Stebbins, the audiometric measures we have taken, ’70; Stebbins and Coombs, ’72). so that we are able to present our findings The monkey is required to make and for the subfamily Cercopithecinae as a maintain manual contact with a metal disc whole. These findings include auditory located on an extended arm in front of the thresholds for pure tones ranging from 60 chair (see fig. 1) until presented with a Hz to 40 kHz in addition to frequency dif- pure tone. Lifting his hand from the disc ference and intensity difference thresholds. (breaking contact) while the tone is on The difference thresholds, however, are indicates a correct detection and is immedi- limited to two genera: Macaca and Papio. ately followed by a food pellet. Failure to Such information is not closely related to break contact with the disc during the tone the natural acoustic environment of the counts as a failure to hear and has no pro- monkey; rather, it gives us an estimate of grammed consequences. Removal of his the lowest sound Pressure to which his ear hand from the disc when the tone is not is sensitive and of the resolving power of present produces a six-second time-out his ear along the dimensions of wavelength from the experiment which is indicated by and amplitude (intensity or sound pres- the extinction of a red light directly in front sure). The Old World monkey is probably of the animal. During that period, the ex- the most popular choice for field and lab- periment is temporarily interrupted and oratory study and it behooves us to know further experimental events (including something about his hearing acuity. Before delivery of food) are postponed. The time- we can interpret the significance of his out acts as an effective punishment for hearing within the context of his natural guessing, but is without the emotional dis- environment, we must first specify the turbance so often associated with electric characteristics of his ear as an analyzer of shock. sound. The remainder of this paper will be We consider maintained contact with devoted to a description of the methods for the metal disc an observing response which the behavioral measurement of hearing in places the animal in an effective listening nonhuman primates and a discussion of position such that he is able to respond the findings for the Cercopithecinae and quickly to the tone simply by lifting his their relation to other primate families. hand. A two and one-half second “trial” occurs randomly between one and nine sec- METHOD onds after the initiation of a contact re- Most of our animals are juvenile and sponse. On 67% of the trials, a tone is subadult males (probable age range be- actually presented and the animal is rein- tween 3 and 7 years). The majority have forced with food if he removes his hand been obtained from importers, although from the disc before the tone is turned off, two taxa (baboons and vervets) were born The remaining 33% of the trials are con- and raised in this country. Upon arrival in sidered “catch trials”- no tone is pre- HEARING OF OLD WORLD MONKEYS 359

Fig. 1 Subject (Macaca nemestrina) seated in the restraining chair for hearing testing. Earphones, response disc, and feeder tube are shown. The monkey is in listening position with his hand on the response disc prior to tone stimulation. 360 WILLIAM C. STEBBINS sented and lifting his hand from the disc at tensity difference thresholds. For the for- this time is punished by a time-out. These mer, a pure tone is presented repetitively; silent trials are used to monitor the ani- maintained contact with the disc is fol- mal's guess rate. We reject the data if the lowed by a second pure tone of different responses to these catch trials exceed 10% frequency alternating with the first or stan- of the total number presented. A trained dard for two and one-half seconds, The animal rarely responds to these trials and monkey is reinforced with food for cor- almost always reports the tone correctly. rectly discriminating the frequency differ- The picture of a trained animal is serene ence by breaking contact with the disc. and relatively uninteresting. Me sits very Correct responses serve to reduce the fre- quietly in his chair with his hand or fore- quency difference between the two tones; arm draped over the response disc (see fig. failure to break contact with the disc when 1). Upon hearing a tone, he lifts his arm the two tones are alternating serves to in- a few inches from the key and immediately crease the frequency difference between replaces it, at the same time extending his them on the next trial. Difference thresh- tongue to receive the food pellet from the olds are determined for frequency and in delivery chute. a very similar manner for tonal intensity. When the behavior has become stable In figure 2, the tracking procedure is under these conditions, threshold testing illustrated in diagrammatic form. The con- for pure tones is begun. A variant of ditions are those for determination of the Bhkesy audiometry (tracking, up-down, or frequency difference threshold; however, titration) is used. For each frequency, the the basic principle of the tracking proce- tone is presented initially well above the dure - that of adjustment of the stimulus estimated threshold for the animal. Follow- difference by the subject's response - is ing each correct report of the stimulus, its the same regardless of the situation in intensity when next presented is decreased which it is used. The standard stimulus is by 5-10 dB; when the subject fails to re- a I-kHz tone and the second or variable port the stimulus, its intensity is increased stimulus is at an initial value 24 Hz from subsequently by the same amount on the the standard. In a brief warmup, the ani- next trial. In the traditional manner, mal responds correctly to the stimulus dif- threshold is defined as the value of stimu- ference, thus gradually reducing that dif- lation correctly detected in one-half of the ference to about 9 Hz, which is calculated trials in which it was presented. as the average frequency difference thresh- Essentially the same procedure is used old at 1 kHz and is based on the data col- in the determination of frequency and in- lected from trial 10 through trial 29 (see

f 12-

1L 6- 3- IlllllillIIIII1II~~1~~~~~ 5 10 15 20 25 30

Trials Fig. 2 Use of the tracking method in the determination of frequency difference thresholds. The session begins after a brief "warmup" at the point labeled " test" on the figure. Correct reports reduce the difference between the frequencies; following a failure to report, the difference is increased. The calculated threshold is indicated by the dashed horizontal line at 9 Hz. HEARING OF OLD WORLD MONKEYS 36 1

8 I 1 I I I I I I I I

I I 1 I I I I 1 I I I ,062 125 .25 5 I 2 4 8 16 32 64 Frequency (kHz) Fig. 3 Threshold of hearing function from 60 Hz to 45 kHz for Cercopithecinae based on data obtained from four genera.

fig. 2). Thresholds are determined in this range frequencies. The selective pressures manner over a wide range of standard fre- in evolution may have played a role in the quencies. development of high frequency hearing which in turn produced more accurate RESULTS AND DISCUSSION sound localization in some of the mammals A representative hearing function for and in the nocturnal prosimians (Master- pure tones for the cercopithecine monkeys ton, Heffner and Ravizza, ’69). As Master- is shown in figure 3. Although individual ton has pointed out, the anthropoid animals may differ slightly, most of the primates with larger heads and greater essential features of the function are evi- interaural spacing no longer had to rely dent for all animals: on high frequencies for accurate sound 1. The frequency range extends from localization, but instead could make use of below 60 Hz to as high as 40-45 kHz. interaural time difference as a cue for 2. Maximum sensitivity is in the mid- localization. range frequencies between 1 kHz and 8 The absolute threshold function for hear- kHz. ing (fig. 3) provides information on the 3. A “dip” or reduction in sensitivity minimum detectable energy levels of sound characteristically occurs near 4 kHz. to which the monkey is capable of respond- These features distinguish the Old World ing. Two other important characteristics monkeys from the prosimians on the one of the auditory system concern its resolving hand, and from the apes and man on the power for acoustic frequency and ampli- other. The former have a frequency range tude (sound pressure). The experimental of hearing extending to 70 kHz and are question refers to the minimal difference relatively less sensitive to frequencies below in frequency and intensity (sound pressure) 8 kHz; man’s upper frequency limit is about to which the monkey can respond. In fig- 20 kHz and the chimpanzee’s about 30 kHz ure 4, the frequency difference thresholds (Stebbins, ’71 ). Thus the cercopithecine (AF) at 60 dB SL, i.e., 60 dB above the monkey may reflect a stage in the evolution threshold of hearing, for two different of human hearing - a trend toward reduc- species of Macaca are shown as a function tion in high frequency hearing and slightly of the standard frequency. The method of enhanced sensitivity to the low and mid- threshold determination is illustrated in 362 WILLIAM C. STEBBINS

500

- 100 2 v 50 aLL

Frequency (kHz1 Fig. 4 Frequency difference threshold at 60 dB SL as a function of frequency for three subjects representing two species of macaque.

8 I I I I I I I I

. Cercopithecinae

hm T3 Y 4 Ha . 0 . 2 . . i . . .

0 I I t 1 I I I 1 .25 .5 I 2 4 8 16 32 Frequency (kHz) Fig. 5 Intensity difference threshold at 60 dB SL as a function of frequency for four subjects representing two genera of Cercopithecinae. figure 2. For example, for a tone of 1 kHz, for man are about 2 Hz (at 500 Hz) and the monkey’s threshold of discriminability 20 Hz (at 8 kHz), respectively. The increase is slightly above 5 Hz; at 8 kHz, it has in- in AF with frequency for monkey and man creased to about 50 Hz. Comparable values is characteristic; man’s differential acuity HEARING OF OLD WORLD MONKEYS 363

h u) L 0 L2 2 .-v .I

vE Ha

.o I

0 10 20 30 40 50 I (dB SL) Fig. 6 Intensity difference threshold as a function of sensation level, i.e., of intenslty above the threshold of hearing for one typical subject. for frequency differences is slightly sharper Maximum sensitivity is in the midrange (Licklider, '51). at 1 kHz and 8 kHz with a region of re- In figure 5, the intensity difference duced sensitivity between these values. thresholds (AI) at 60 dB SL for two differ- The difference threshold for frequency ent species of Macaca and one of Papio are increases with frequency from a low of shown as a function of frequency. The about 5 Hz at 250 and 500 Hz to a differ- difference threshold is relatively constant ence threshold as high as 500 HZ at 20 at about 2 dB across frequency with a slight kHz. increase suggested at the higher frequen- The difference threshold for sound pres- cies. The shape of the function for man is sure is constant at about 2 dB for all fre- very similar with a lower difference thresh- quencies. old of less than 1 dB. The intensity differ- The difference threshold for sound pres- ence threshold increases as the intensity sure is proportional to the intensity above level at which it is determined is increased. the threshold of hearing at which it is de- These findings are shown for one macaque termined. at several levels above the threshold of These basic functions obtain also for hearing in figure 6. For example, at 5 dB man with certain important differences: SL, A1 = 0.01 pbar; at 50 dB SL, A1 = 1 man's upper frequency limit is 20 kHz; his pbar. The proportional relation obtains for differential acuity for both frequency and man also (Licklider, '51). intensity is slightly superior to that of the CONCLUSIONS Old World monkey. The range of hearing for the cercopithe- LITERATURE CITED cine monkey extends from below 60 Hz to Licklider, J. C. R. 1951 Basic correlates of the about 40 kHz. auditory stimulus. In: Handbook of Experi- 364 WILLIAM C. STEBBINS

mental Psychology. S. S. Stevens, ed. John physics: The design and conduct of sensory Wiley & Sons, Inc., New York, pp. 985-1039. experiments. W. C. Stebbins, ed. Appleton- Masterton, B., H. Heffner and R. Ravizza 1969 Century-Crofts, New York, pp. 41-66. The evolution of human hearing.-- .T. Acoust. 1971 Hearing. In: Behavior of Non- SO~.Amer., 45: 966-985. human Primates. A. M. Schrier and F. Stollnitz, Moody, D. B., W. C. Stebbins and J. M. Miller eds. Academic Press, New York. Vol. 3, pp. 1970 A primate restraint and handling system 159- for auditory research. Behav. Res. Meth. & 192. Instru., 2: 180-182. Stebbins, W. C., and S. Coombs 1973 Chapt. 4. Napier, J. R., and P. H. Napier 1967 A Hand- Behavioral assessment of ototoxicity in non- book of Living Primates. Academic Press, Lon- human primates. In: Proceedings of Conference don. on Behavioral Toxicology. B. Weiss and V. G. Stebbins, W. C. 1970 Studies of hearing and Laties, eds. Appleton-Century-Crofts, New York, hearing loss in the monkey. In: Animal Psycho- in press.