Electromyography of pongid shoulder muscles. III. Quadrupedal positional behaviorкод для вставкиСкачать
Electrotnyography of Pongid Shoulder Muscles Ill. QUADRUPEDAL POSITIONAL BEHAVIOR ' RUSSELL H. TUTTLE A N D JOHN V. BASMAJIAN Department of Anthropology and Committee on Evolutionary Biology, The University of Chicago, Chicago, Illinois fiO6.37 and McMaster University and Rehabilitation Centre, Chedoke Hospitals, Hamilton, Ontario, Canada L8N 3Lfi - K E Y WORDS Apes Electromyography . Shoulder muscles Positional behavior . Knuckle-walking . Quadrupedalism . ABSTRACT Electromyographic (EMG) recordings were taken from 14 shoulder muscles (or major parts of them) in a gorilla, a chimpanzee and an orangutan as they stood quadrupedally and tripedally, descended from elevated substrates, crutch-walked, and progressed quadrupedally on inclined and level substrates. In the African apes, low potentials commonly (but not always) occurred in the sternocostal pectoralis major, anterior deltoid, supraspinatus and subscapularis muscles during quadrupedal stance. The quadrupedal orangutan always exhibited low potentials in the pectoralis major muscle and EMG activity commonly occurred in her supraspinatus and subscapularis muscles. Quiescent tripedal stances were not accompanied by striking changes in EMG patterns from those which characterized quadrupedal stances. Per contra, eccentric loadings of the forelimb during descents from elevated substrates generally recruited notable EMG activity in the deltoid, supraspinatus and, to a lesser extent, infraspinatus muscles of the three pongid apes. The pectoralis major and caudal serratus anterior muscles were much more active in Pongo and Pun during these descents. Supportive segments of quadrupedal locomotive cycles were generally accompanied by EMG activity in the pectoralis major, intermediate and posterior deltoid and supraspinatus muscles. The intermediate and posterior deltoid muscles were characteristically active during pre-release of the hand and early swing phase. The cranial trapezius and supraspinatus muscles also may act during early swing phase. We conclude that the pectoralis major and perhaps the supraspinatus and subscapularis might serve regularly as postural muscles during static terrestrial quadrupedalism in pongid apes. The lack of dramatic differences between the EMG patterns exhibited during fist-walking versus knuckle-walking indicates that an evolutionary transformation from a shoulder complex like that of Pongo to ones like Pun or vice versa need not entail major changes in myological features. The main purpose of this report is to present the first description of electromyographic (EMG) activities of the shoulder muscles in pongid apes during spontaneous quadrupedal positional behavior. In other publications (Tuttle and Basmajian7 '77, '78) we presented EMG information about the same muscles as AM. J. PHYS. ANTHROP. (1978) 49: 57-70. the subjects raised their arms and engaged in suspensory behavior and suggested how this might relate to several current problems on scapular shape and hominoid phylogeny. Herein we will focus on mechanisms of pongid Parts of this paper were read at the AAPA meeting in Toronto, Canada, April. 1978. 57 Caudal serratus anterior Cranial trapezius 1 Teres major c 0 G c 0 G c 0 1 1 1 1 2 1 1 2 G 0 1 0 0 0 0 0 0 0 0 0 0 0 0 c G 1 2 0; + O;l+l ;+t+;i~~ 0 0 0 0 0 0 0 O;i+J - 0;+; 0; + - 0 ++ 0 0. + [+ +I O;i+l 0 ++ + 0; 0 0 0 - - 0 0 0 0 0 0;[+1 O;[+I 0;l-I 0 0 o;+ 0 - 0 O;l+l 0 - 0 O;l+l - 0- 0 o;++ 0- o;+ ++;+++ I++I +;++ [++I -;++ t ;I01; + M - 0 0 + 0; 0; + 0 0; 0 + 0 0 0 0 0 0 0 0; + 0 0 0 0 S O;I+l - 0 - 0 0 @I+; + +I 0;I-1 - 0;itl 0 O;+;[++l 0;i-1 r++;++-i +;++; I- + + I +;101; +;++; l i i i l R 0 + + 0; * 0 0; it+l 0;i ; 0; - 0 0 0 0 0 0 0 0 ;+ 0 0 0 M Swing phase Quadrupedal walking and lrunningl Stance phase t ; + S +;++ ++ - walking Crutch +, 0 0 - 0 0 0 0 - 0 - 0 0 ;7 0 [++I - t;t+ 0 0; + ; I + + I Descent onto hand - 0; t 0 +;ro;++i + 0 0 0 + 0; G 0 3 0; + stance + c 0 c 2 Quiet tripedal 0; 1 G 2 Quiet quadrupedal ntance 2 2 Subject N Vertebrocostal latissimus dorsi Iliac latissirnus dorsi Sterno,costal pectoralis major Muscle TABLE 1 Activities of pongid shoulder musrles during quadrupedal positional behavior (symbols: N , number of 60-120 rninute experiments; G, gorilla; C, chimpanzee; 0, orangutan; -, no d a t a ; 0,silent; +, low EMG; + moderate EMG; + + +, high EMG; S, sluw pace; M, moderate pace; R, rapid pace). I1 indicate relatively rare occurrences or mainly while the subject w a s groggy. - 0 I+ + I 0:+; - 0 ;+ - 0 0 0 0 - 0 0 0 0 ;t 0 0 0 R sz z Ez m z c 2z tr z b m 2 5 H s: r r M m M c EMG OF PONGID SHOULDER MIJSCLES AND QUADRUPEDALISM f 0 t. +- + o .~ + ._ + 0-k -_ ++ t"; +, + + -+ tt i + + . + . ++ + I .0 t + + + +. t + - ...t + + + + O + f + +; + -+ ." . - + t. + I + + + 2 + y + . -+ I I t. _ c + + + ++ tt + + 0+ to + + + . 0 ++ t ?"Z + ++ ,++T a+o+ + t ++ 2 ++ < + t q+- 4-+ ;" + 0-+ 0-+ +-+ ._ + + 3 + % 3 + ;+ a + a+a" 0 3 + + +. ^ ' ^ t t I 0 + ." c $" - t + + + f +- !- 0 0 t 3 0 0 f. 0 + .^ +. _ t o 0 30 0 u 0 4 i N N r n 1 z z c $2 :2 -oi 8Q L .+ B E 3 cn 0 59 60 RUSSELL H. TUTTLE AND JOHN V. BASMAJIAN quhdrupedalism and evolutionary problems of knuckle-walking. The muscles or major parts thereof are listed in table 1. SUBJECTS AND METHODS The subjects in this study were a 6%-to 7%year-old female Sumatran orangutan (Pongo pygmaeus); a 5?h- to 6lh-year-old female lowland gorilla (Pan gorilla); and a 4%- to 5%year-old male chimpanzee (Pun troglodytes). Indwelling fine-wire bipolar electrodes (Basmajian, '74; pp. 35-39) were used according to procedures which have been described in previous papers (Tuttle and Basmajian, '74a,b). The subjects could move freely in the testing area. They had opportunity to stand tripedally and quadrupedally on level and inclined surfaces (figs. l a , 2a,b,d), to eccentrically load the forelimb during descents from a ramp (fig. lb) and platform, and to walk quadrupedally a t various speeds on the ramp and floor (figs. lc,d, 2c). We induced bouts of locomotion up and down the ramp by placing M & M candies alternately a t its apex and near its base. Quadrupedal behavior generally commenced during the initial 20 minutes after arrival in the testing area and was exhibited intermittently throughout the recording period. The 17.8" inclined ramp was 244 cm long, 76 cm wide and 51 cm high a t the apex. The platform was 91 cm x 76 cm x 51 cm. RESULTS The number of recording sessions which provided useful data on each muscle in the three pongid apes and general levels of activity during quiescent quadrupedal and tripedal stances, descents from elevated substrates with the tested forelimb eccentrically loaded, crutch-walking, and quadrupedal locomotion are listed in table 1. Hereinbelow we will discuss the activities of certain muscles and compare results obtained from the three subjects especially insofar as they might reveal basic similarities and differences between the habitual knuckle-walkers (gorilla and chimpanzee) and the orangutan. Unless otherwise mentioned all descriptions are based on data from fully alert subjects. Quadrupedal stance When the gorilla stood quadrupedally, she habitually extended her elbow so that the arm and forearm were virtually aligned. Although the dorsum of the knuckled hand faced directly anteriorly, t h e ulnar olecranon process was oriented more laterad than posteriorly (as in figs. lb,c). The shoulders appeared to be hunched forward somewhat. The chief contact points with the substrate were the dorsal surfaces of the middle segments of digits 11-IV. The wrist was slightly volarflexed and adducted, probably loading digits I1 and I11 more than digit IV. She very rarely stood with a supportive hand fisted. During quadrupedal stance, the chimpanzee exhibited a greater variety of forelimb postures than the gorilla did. The dorsal surfaces of his knuckled hands faced laterad, anterolaterad, or anteriorly. The olecranon process was posterior when the dorsum of the hand faced laterad. But, as in gorilla, it was laterad when the dorsum of the hand faced anteriorly. Often weight was borne primarily on digits 111 and IV while digit I1 was rather widely abducted and flexed so that i t only lightly touched the substrate. The wrist was slightly volarflexed and adducted. The elbow joint commonly exhibited slight flexure. The chimpanzee's shoulders appeared to jut anteriorly somewhat less than the gorilla's did. The chimpanzee commonly stood with one or both hands fisted, especially when he was groggy or immediately following a bout of rapid progression in which a fisted hand was used during the stance phase of the final locomotive cycle. The orangutan exhibited the greatest variety of forelimb postures during quadrupedal stance. Fisted hands, with the dorsum facing anteriorly, obliquely or laterad and the ulnar olecranon process concordantly oriented as described hereinabove for the chimpanzee, were most commonly employed for static support. But she also occasionally placed her hands palmigrade (Tuttle, '671, in a knuckled posture (fig. 2b), or with only the middle fingertips touching the floor. The latter two postures were employed as she stood beneath a desired object and very slowly rose to a bipedal posture (fig. 2d) and as she returned from bipedal foraging to a more fully quadrupedal stance. Sometimes the arm and forearm were well aligned. At other times the elbow exhibited flexure during various fisted, palmigrade and knuckled stances. In the gorilla and chimpanzee, none of the shoulder muscles always produced EMG potentials during quadrupedal stance. However, low activity commonly appeared in their sternocostal pectoralis major, anterior deltoid, supraspinatus, and subscapularis muscles during quadrupedal stance. Several muscles, in- EMG OF PONGID SHOULDER MUSCLES AND QUADRUPEDALISM 61 Fig. 1 Gorilla female a, standing tripedally on ramp while scratching t h e dorsum of her neck with the left hand; b, eccentrically loading t h e extended right forelimb during a descent from the ramp; and c,d, knuckle-walking, Note (a-c) that the olecranon process of the ulna faces laterad when the dorsum of the knuckled hand is oriented anteriorly. cluding the latissimus dorsi, teres major, cranial trapezius, caudal serratus anterior and posterior deltoid, were always virtually silent as the African apes quiescently stood quad- rupedally. The two apes differed from one another chiefly in the relatively greater incidences of low potentials in the gorilla’s infraspinatus muscle and of moderate poten- 62 RUSSELL H. TUTTLE AND JOHN V. BASMAJIAN Fig. 2 Orangutan female a, standing tripedally with the right hand fisted; b, standing quadrupedally with both hands knuckled; c , fist-walking (note marked dorsiflexion of left wrist); and d, reaching overhead from a quasi-bipedal posture with t h e right hand knuckled. EMG OF PONGID SHOIJLDER MUSCLES AND QUADRUPEDALISM 63 a Fig. 3 EMG recording of right a, sternocostal pectoralis major; b, teres major; and c , supraspinatus muscles in a n orangutan quasi-crouched alternately tripedally and quadrupedally on the floor while eating food from t h e apex of a ramp. EMG activity increased from low to moderate levels in the supraspinatus (secs. 7121 as she eccentrically loaded t h e right forelimb while leaning forwards to lick t h e top of the ramp. t ; time in seconds. tials in t h e chimpanzee's supraspinatus muscle. In the quadrupedal orangutan, low potentials always occurred in the sternocostal pectoralis major muscle and they commonly also appeared in the supraspinatus and subscapularis muscles (fig. 3). Unlike the African apes, the orangutan's anterior deltoid muscle remained silent during quadrupedal stance. Many other muscles, including the latissimus dorsi, teres major (fig. 3), cranial trapezius, caudal serratus anterior, rhomboid, deltoid, infraspinatus and teres minor, were silent in the quadrupedal orangutan. Tripedal stance During tripedal stance, positions of the forelimb were not noticeably different from those employed in quadrupedal stance. Some tripedal stances were only momentary, as when grabbing or striking at objects and investigators, while those used during feeding and reaching unimanually overhead might last the better part of a minute before quadrupedal, bipedal or other postures were adopted. Generally the level of EMG activity did not change during long tripedal stances unless there was bodily movement (fig. 3 ) . Overall the EMG patterns observed during tripedal stance in the three subjects are remarkablv similar to those exhibited during " quadrupedal stance. In addition, low potentials occurred in the gorilla's posterior deltoid, in the chimpanzee's infraspinatus, and in the orangutan's iliac latissimus dorsi muscles and moderate potentials were sometimes exhibited by the orangutan's supraspinatus muscle during tripedal stance (fig. 3). These are the major instances in which tripedal stance appears to have induced higher levels of EMG activity than quadrupedal stance did. Descent onto the hand Eccentric loadings of the forelimb during descents from a ramp (fig. lb) and a level platform onto a knuckled (Pan) or fisted (Pongo) hand commonly were accompanied by greater EMG activity than was exhibited in tripedal stance (table 1).This was particularly striking in t h e orangutan. In all subjects, the three heads of the deltoid, the supraspinatus, and, to a lesser extent, the infraspinatus muscles were quite active during the supportive phase of descents onto the hand. In the African apes there was little difference in the EMG pattern of the sternocostal pectoralis major muscle during stance and the supportive phase of descents. 64 RUSSELL H. TUTTLE AND JOHN V. BASMAJIAN facing anteriorly as i t did during the stance phase. Position of the wrist generally changed little between stance and swing phase. The elbow was flexed slightly and the shoulder was raised subtly so that the anteriorly swinging hand cleared the substrate. Like stance, locomotor positions of the hand in the knuckle-walking, or more rarely, fistwalking chimpanzee varied more than in the gorilla. For instance, the dorsum of his hand sometimes faced laterad in stance phase, particularly during fist-walking steps. The orangutan’s quadrupedal locomotion Crutch-walking was of shorter duration, more variable in Crutch-walking was executed by swinging speed, and often appeared to be less well cooror sliding the flexed hindlimbs and torso be- dinated than that of the African apes. She tween the abducted forelimbs. Knuckled (Pan usually walked with her hands tightly fisted and with the dorsal aspects of proximal phagorilla, Pan troglodytes) or fisted fPongo pygmaeus) hands served as chief contacts with langes 11-V on the substrate (fig. 2c). She very the substrate. Crutch-walking was infrequent rarely moved with a hand palmigrade. She and consisted of only one or two steps per never knuckle-walked. The dorsum of the locomotor bout on the floor or off the base of fisted hand faced anteriorly, laterad or interthe ramp. I t was accompanied by considerable mediately during stance phases. The extended EMG activity while the forelimbs were sup- fingers were directed laterad during the portive, expecially in the sternocostal pec- stance phase of palmigrade steps. The orangtoralis major, iliac latissimus dorsi, caudal utan often shuffled forward without fully serratus anterior P a n gorilla only) and rhom- elevating her hand from the substrate during boid (especially P. gorilla) muscles (table 1). “swing phase.” Unlike the African apes, in We cannot attempt close comparisons among Pongo the shoulder did not pass over the supthe three subjects because our data is frag- portive hand during midstance phase. During the stance phase of locomotor cycles, mentary. the sternocostal pectoralis major, intermediQuadrupedal locomotion ate and posterior deltoid, and supraspinatus For purposes of analysis and description we muscles were regularly active in all subjects. arbitrarily divided each forelimb locomotor In the African apes, the pectoralis major gencycle into a stance phase, encompassing the erally acted a t low EMG levels throughout the period between initial contact and release of stance phases of slow and moderately paced the hand from the floor or ramp, and a swing locomotor cycles. Higher potentials sometimes phase, embracing the period during which the accompanied rapid knuckle-walking though hand was free of the substrate en route to lower potentials were still more common (fig. final repositioning. Because release of the 4). Similarly, in Pongo both low and moderate hand is initiated proximally by subtle eleva- potentials were exhibited by the sternocostal tion of the shoulder and flexion of the elbow pectoralis major muscle at all locomotor paces joint support does not occur throughout the and high potentials occurred rarely during stance phase. Instead support is concentrated rapid fist-walking. in t h e early and especially the middle segLow or moderate potentials characterisments of stance phase. tically occurred in the intermediate and posteDuring the stance phase of locomotor cy- rior deltoid muscles of the three subjects durcles, the gorilla quite consistently placed her ing the stance phases of slow, moderate and knuckled hand with the dorsum facing anteri- rapid locomotor cycles (fig. 4). In Pan and orly (fig. lc). Concordantly, the ulnar olecra- Pongo the intermediate deltoid muscle gennon process was oriented laterad. During early erally commenced activity toward the end of swing phase, the hand might be pronated so stance phase, before the elbow flexed and (in t h a t the palm faced somewhat laterad (fig. Id) Pan only) before adduction of the wrist was or it might be carried forward with t h e dorsum decreased. EMG activity continued as weight Per contra, in the orangutan i t exhibited potentials ranging from low to high. Similarly, the caudal serratus anterior muscle was much more active in the orangutan than in the African apes. High EMG potentials occurred more commonly in the anterior deltoid and supraspinatus muscles of Pongo than in those of Pan. The major notable difference between the shoulder muscles of the two African apes during the supportive phase of descent is the higher EMG activity of the subscapularis muscle in the chimpanzee. EMG OF PONGID SHOULDER M1JSCLES AND QUADRUPEDALISM a t locomotor bouts, low potentials commonly occurred. In Pan troglodytes, t h e posterior deltoid muscle generally was active during pre-release and early swing phase of the hand, especially during rapid knuckle-walking (fig. 4). Occasionally, it was silent throughout a locomotor cycle or was only active during prerelease of the hand. In fist-walking Pongo pygmaeus, the posterior deltoid muscle generally evinced either low or moderate and rarely marked potentials during pre-release of the hand and lower potentials during early swing phase. In Pun gorilla, the supraspinatus muscle acted quite variably during quadrupedal locomotion. Some locomotor cycles were accompanied by virtually continuous low or widely spaced single potentials while more commonly consequential potentials were concentrated during early swing phase. On a few occasions, the supraspinatus muscle was silent during swing phase, especially when knuckle-walking up and down the ramp. Hence it appears that forward swing of the forelimb can occur without action of the supraspinatus muscle in the gorilla. EMG potentials, including some high ones, often occurred when there was noticeable lateral rotation of the humerus a t the glenohumeral joint during initial and early swing phase, Sometimes the potentials which occurred during stance phase were highest a t pre-release of the hand, i.e., a t the outset of repositioning the forelimb. In Pan troglodytes, the supraspinatus muscle was virtually always active a t low or moderate levels during much of the locomotive cycle. Most commonly, it was active from the latter half of swing phase through the supportive segments of stance phase. But on otter occasions, EMG activity occurred predominantly during stance phase and only briefly or not at all during swing phase. Characteristically, it was silent a t pre-release of the hand. In Pongo pygmaeus, the supraspinatus muscle was active a t various EMG levels ranging from low to moderate during support phase, especially from midstance through release of the hand, and early swing phase. Sometimes it was active only during pre-release and early swing. I t was silent during some very slow fist-walking bouts. Thus, in Pongo it seems to he particularly related to the initiation of manual repositioning during fist-walking a t moderate and brisk paces. In the African apes the infraspinatus mus- * 1 2 3 4 5 6 sec. Fig. 4 EMG recording of right a, sternocostal pectoralis major; and b, anterior; c, intermediate; and d, posterior deltoid muscles in a briskly knuckle-walking chimpanzee. a is active throught supportive aspects of stance phase; b,c are chiefly active (low potentials) during mid and late stance phase; d, acting to lift and perhaps also laterally rotate the arm, exhibits moderate potentials during pre-release of the hand and swing phase. t, time is seconds. was removed from the hand. It ceased abruptly during the early segment of swing phase. Like the pectoralis major muscle, the intermediate deltoid did not exhibit higher potentials during rapid locomotion unless the supportive forelimb was subjected to increased loads. Indeed EMG activity was often quite low during rapid locomotion on level surfaces unless quick turns were being executed (fig. 4). In Pan gorilla, the posterior deltoid muscle exhibited EMG activity during pre-release of the hand at the end of stance phase. During rapid locomotion about the room with the forelimb acting as a pivot for sudden turns, the posterior deltoid probably contributed notable propulsive force. But during other 65 66 RUSSELL 11 TUTTLE AND JOHN V BASMAJIAN cle was commonly active briefly a t the outset of manual repositioning, i,e., during pre-release and earliest swing phase. This is when lateral rotation of the humerus at the glenohumeral joint. occurs. In the gorilla, moderate potentials accompanied brisk knuckle-walking on the floor and ramp. EMG potentials were generally low in the chimpanzee’s infraspinatus muscle even during brisk locomotion up and down the ramp. In the chimpanzee and groggy gorilla, the subscapularis muscle was sometimes active at low levels during late swing and/or most of the succeeding support phase. Thus it may be related to medial rotation of the humerus a t the glenohumeral joint which occurs a t that point in the locomotive cycle, In the fist-walking orangutan, when low potentials occurred in the subscapularis muscle they were confined to the support phase as the humerus was rotating medially a t the glenohumeral joint. In the knuckle-walking apes, the cranial trapezius muscle generally exhibited brief bursts of quite low potentials at pre-release of the hand. In the orangutan, very low potentials commonly accompanied pre-release of the hand and continued well into swing phase. However, in one locomotor bout, moderate potentials occurred during the early swing phases of several fist-walking steps. In the gorilla, the caudal serratus anterior muscle sometimes exhibited low potentials during swing phases of moderately and briskly paced knuckle-walking, On other occasions only inconsequential single potentials or silence accompanied knuckle-walking. In the chimpanzee, knuckle-walking was usually accompanied by nil activity in the caudal serratus anterior muscle though occasionally very low potentials appeared during the propulsive segment of stance phase. In Pongo very low potentials occasionally accompanied the early and midstance phase or pre-release and early swing phase. But generally the caudal serratus anterior muscle was virtually silent during fist-walking. In the gorilla, the rhomboid muscle commonly exhibited low potentials during stance phase, particularly when she used full versus short forelimb strides and knuckle-walked briskly on the floor and ramp. The chimpanzee’s rhomboid muscle acted at low levels during release of the hand and early swing phase of‘ brisk knuckle-walking and fist-walking on the ramp and floor. In Pongo, the rhom- boid muscle exhibited low and moderate potentials during late stance phases of fistwalking. The anterior deltoid muscle of the gorilla generally acted a t low and occasionally mode r a t e levels during t h e stance phase of knuckle-walking. It was silent during swing phase. In the knuckle-walking chimpanzee, very low potentials characterized the anterior deltoid muscle during stance phase (fig. 4)and occasionally it was silent. In the fist-walking orangutan’s anterior deltoid muscle only a brief burst of very low potentials occurred at the end of swing phase. In summary, the following general features emerge from this study re the role of shoulder muscles during quadrupedal locomotion: 1. The sternocostal pectoralis major muscle is a regular supporter of the upper body weight and probably also a propulsive element during the stance phases of knucklewalking and fist-walking steps in all three pongid apes. 2. The intermediate and posterior deltoid muscles provide propulsive force during late stance phase and concurrently they may also augment st,ability in the shoulder complex. 3. The subscapularis muscle often contributes to support or propulsion or both functions during stance phase. It probably also contributes to stability of the glenohumeral joint by acting as a medial rotator of the humerus. 4. Propulsion, support and shoulder stability may be augmented severally by the rhomboid (P. gorilla, Pongo), supraspinatus (P. troglodytes, Pongo), the anterior deltoid (Pardt and the iliac segment of the latissimus dorsi (P. gorilla) muscles. 5. Release of the hand prior to repositioning regularly involves the intermediate and posterior deltoid and cranial trapezius muscles in all three great apes and the infraspinatus muscle (probably acting as a lateral rotator of the humerus) in the African apes. 6. The supraspinatus was the only shoulder muscle which was commonly active during early swing in the three subjects. The infraspinatus muscle also acted during early swing phase in the African apes. 7. During late swing phase, as the hand descended toward the substrate, the subscapularis muscle (probably acting as a medial EMG OF PONGID SHOULDER MUSCIXS ANT) QUADRUPEDALISM rotator of the humerus) was active in the African apes whereas the anterior deltoid muscle was active in the orangutan. Irregularly the iliac segment of the latissimus dorsi muscle was active in the gorilla just prior to hand contact. DISCUSSION Pongid versus other mammalian quadrupedalism Textbooks on mammalian functional morphology (e.g., Young, '57; p. 161; Leach, '61: p. 159) commonly depict serratus anterior in pronograde mammals as a postural muscle which supports the upper body weight during quadrupedal positional behavior. Our EMG studies indicate that the thick caudal segments of pongid serratus anterior muscles do not act consistently in accordance with this model. In all subjects, the caudal serratus anterior muscles were silent during tripedal and quadrupedal stances. And they were silent or minimally active during the stance phases of most locomotor cycles. It is possible that cranial digitations of pongid serratus anterior muscles are more active during quadrupedalism. As predicted by Young ('57: p. 161),the cranial trapezius did not serve as a postural muscle in our quadrupedal subjects. Unimpressive EMG levels during the swing phases of locomotor cycles highlight t h a t its principal activity occurs during arm-raising in apes as it does in orthograde man (Tuttle and Basmajian, '77). Among the shoulder muscles which we tested, only the pectoralis major and perhaps the supraspinatus and subscapularis might be singled out as postural muscles during static terrestrial quadrupedalism in pongid apes. This generalization might apply less to the African apes than t o the orangutan, especially when its forelimb is eccentrically loaded. Stern et al. ('77) have published the only EMG study on non-hominoid primate shoulder muscles to which we can compare broadly our results. They present highly condensed, schematic renderings of EMG records from the latissimus dorsi, caudal serratus magnus (= anterior). middle ( = intermediate) deltoid, and pectoralis major muscles of two Ateles and one Lagothrix during pronograde quadrupedalism along a horizontal branch. In the ateline monkeys, the sternocostal ( = caudal) pectoralis major muscle exhibited 67 uniphasic periods of EMG activity commencing a t mid (Lagothrix) or late (Ateles) swing phase and ending during early (Lagothrix) or mid fAteles) stance ( = support) phase of the quadrupedal locomotor cycle (Stern et al., ' 7 7 ) . This contrasts noticeably with the pongid apes in which the sternocostal pectoralis major muscle acted during most or all of the stance phase and was silent during the swing phase of quadrupedal locomotor cycles. Like the pongid apes, the ateline monkeys irregularly exhibited EMG activity in the latissimus dorsi muscle during quadrupedal locomotion. In Lagothrix, the low potentials were confined to the late swing phase of the locomotor cycle. In the fully alert chimpanzee and orangutan, the latissimus dorsi muscle was virtually silent during quadrupedal progression on the floor and ramp. The iliac segment of the latissimus dorsi muscle in the gorilla sometimes exhibited low potentials during late swing and most of the stance phase as she progressed quadrupedally on the floor. Once when she rapidly ascended the ramp, moderate potentials occurred in the iliac segment of the latissimus dorsi muscle as she reached the apex (table 1). The caudal serratus anterior muscle seems to act quite variably in the ateline monkeys and the pongid apes during quadrupedal locomotion. Although the overall patterns of activity in Ateles and Lagothrix are far from identical (figs. 3, 5 in Stern et al., '771, both species consistently evinced prominent EMG activity during stance and swing phases of the quadrupedal locomotor cycle. In the pongid apes, the caudal serratus anterior muscle was less consistently active than it was in the ateline monkeys and i t exhibited only low potentials. These were largely confined to the swing phase in Pan gorilla; occurred only during the stance phase in Pan troglodytes; and could occur during swing andlor stance phases of moderately paced fist-walking in Pongo pygmae us. The intermediate deltoid muscle was consistently biphasically active during early and mid stance phases and mid swing phases of quadrupedal locomotor cycles in Ateles. This pattern was only partly exhibited by Lagothrix during rapid progression. At slower paces, Lagothrix evinced only a short period of EMG activity toward the end of the stance phase (Stern et al., ' 7 7 ) . Unlike Ateles, in the pongid apes, the intermediate deltoid muscle 68 RUSSELL H. TUTTLE AND JOHN V. BASMAJIAN acted chiefly during mid and especially late stance phase and its activity generally continued into the beginning of swing phase. We would need studies of branch-walking pongid apes and ground-walking Ateles and Lagothrix before we could attempt to attribute dissimilarities in EMG patterns of their shoulders to differences between arboreal and terrestrial quadrupedalism versus other factors. Functional and evolutionary implications The shoulder muscles of the three great apes exhibited remarkably few major differences in overall activity pattern during knuckling behavior - the entire spectrum of static and locomotor hand postures in which the dorsa of middle segments of digits 11-V are the chief contacts with the substrate (Tuttle, '75) - versus other modes of quadrupedalism (except perhaps crutch-walking about which we have limited data). This is strikingly revealed during locomotor bouts wherein the chimpanzee exhibited similar EMG patterns during facultative fist-walking and knucklewalking steps, often interspersed with one another. Further, we found relatively few consistent explainable differences between the muscle activities of fist-walking Pongo and knucklewalking Pan gorilla and Pan troglodytes. Certain aspects of positional behavior, like the supportive phases of descents onto the fist and the stance phases of some bouts of fist-walking, seemed to elicit somewhat higher potentials in certain muscles of Pongo than their counterparts in Pan. But the caveat r e too fine-grained comparisons between EMG experiments must be recalled here. More controlled future studies, employing horizontal and inclined treadmills, might reveal clearcut differences between the capacities of orangutans and the African apes to sustain terrestrial quadrupedal gaits. That an orangutan can engage in a variety of spontaneous short-term quadrupedal behaviors without continuous inordinately high EMG potentials in its shoulder muscles is not too surprising since orangutans are practiced arboreal quadrupeds and adult males sometimes progress quadrupedally on the ground in their natural habitats (see Tuttle, '77: pp. 284-287, for a review of current knowledge about their naturalistic positional behavior). Further, our experimental captive had had ample opportunity to accommodate to terrestrial quadrupedalism. The eye-catching structures of their shoulder girdles, which are commonly linked to versatile arboreal climbing, suspensory behavior and overhead feeding, apparently are not extreme or uncompromising enough to require extraordinary muscular effort during run-of-the-mill quadrupedal behavior. Ligaments and tendonized components of muscles might be important factors here. But whatever specific features are operating, we may conclude that the evolutionary transformation from a shoulder complex similar to that of Pongo to ones like Pan or vice versa probably would not be problematic vis-a-vis myological features. Such transformations could probably occur rather rapidly, especially in response to alterations in the selective complex as would accompany an increase in terrestrial or arboreal quadrupedalism. ACKNOWLEDGMENTS This study was supported mainly by NSF grants GS-3209 and SOC75-02478 and by a Public Health Service Research Career Development Award (1-K04-GM16347-01)from the National Institutes of Health. Supplementary support was provided by NIH Grant RR00165 to the Yerkes Regional Primate Research Center and the Marian and Adolph Lichtstern Fund of the University of Chicago. We are especially thankful for the assistance of Doctor G. H. Bourne, J. Malone, E. Regenos, J. Perry, R. Pollard, S. Lee, R. Mathis, J. Roberts, Doctor M. Keeling, Doctor M. Vitti, J. Hudson, K. Barnes and C. Lin-Bodien. LITERATURE CITED Basmajian, J. V. 1974 Muscles Alive. Their Functions Revealed by Electromyography. Third ed.The Williams & Wilkins Co.. Baltimore. Leach, W. J. 1961 Functional Anatomy, Mammalian and Comparative. Third ed. McGraw-Hill Book Co., Inc., New York. Stern, J. T., J r . , J. P. Wells, A. K. Vangor and J . G. Fleagle 1977 Electromyography of some muscles of the upper limb in Ateles and Lagot.hrix. Yearbook of Physical Anthropology - 1976. Vol. 20. pp. 498-507. Tuttle, R. H. 1967 Knuckle-walking and the evolution of hominoid hands. Am. J. Phys. Anthrop., 26: 171-206. 1975 Knuckle-walking and knuckle-walkers: a commentary on some recent perspectives on hominoid evolution. In: Primate Functional Morphology and Evolution. R. H. Tuttle, ed. Mouton, The Hague, pp. 203-212. 1977 Naturalistic positional behavior of apes and models of hominid evolution, 1929-1976. In: Progress in Ape Research. G. H. Bourne, ed. Academic Press, New York, pp. 277-296. EMG OF PONGID SHOULDER MUSCLES AND QUADRUPEDALISM Tuttle, R. H., and J. V. Basmajian 1974a Electromyography of brachial muscles in Pun gorilla and hominoid evolution. Am. J. Phys. Anthrop., 41: 71-90. 197413 Electromyography of forearm musculature in gorilla and problems related to knuckle-walking. In: Primate Locomotion. F. A. Jenkins, Jr., ed. Academic Press, New York, pp. 293-347. 1977 Electromyography of pongid shoulder rnus- 69 cles and hominoid evolution. I. Retractory of the humerus and “rotators” of the scapula. Yearbook of Physical Anthropology . 1976. Vol. 20, pp. 491-497. 1978 Electromyography of pongid shoulder muscles. 11. Deltoid, rhomboid and “rotator cuff.” Am. J. Phys. Anthrop., 49: 47-66. Young, J. 2. 1957 The Life of Mammals. Oxford University Press, London.