UNIVERSITY OF ILLINOIS

3 MAY

THIS IS TO CERTIFY THAT THE THESIS PREPARED UNDER MY SUPERVISION BY

Leahanne M, Sarlo

Functional Anatomy and Allometry in the of ENTITLED......

Suspensory

IS APPROVED BY ME AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE

DEGREE OF. BACHELOR OF ARTS

LIBERAL ARTS AND SCIENCES

0-1M4 t « w or c o w n w

Introduction...... 1

Dafinibon*...... 2

fFunctional m ivw w iiw irwbaola am foriwi irviiiiHbimanual iw 9 jrvafwooaitiona! Vf aai irwbahavlort iiHviw i# in thaahoukter joint complex...... 3 nyvootiua pooioonaiaA^^AlttAAMAl oanainor...... i 4 *4 -TnoOV a awmang.a Ia MAMJI* nwoo—U^AtkA^AA i (snDBDfllluu)/AuMyAlftAlAnAH |At AkJA^AAtlA•wwacwai* |A ...... 4ia i -Tho lar : KM obate* ]|£...... 15

ANomatry...... 16

Material* and matooda...... 19

Meaeurementa...... 20

RaauN*...... 22

Dlacuaalon . .. .28

Conduaion...... 25

Graph 1 ...... 30

Graph 2 ...... 31

Graph 3...... 32

Graph 4 ...... 33

Literature cited...... 34 iQtUtUKtiQQ

The behavior and anatomy of the leaser ap«t it a topic which h it long interested scientists who study primates. Ths combination of bizarre elongated torsi!mbs (especially tha distal sagmant) with a startling and spectacular form of locomotion is indicativs of a highly apadaiizad adaptive pattsm. Howsvsr, it took an inordinata amount of tima batora scientists bagan to apply tha

anatomical studias ol tha gfcbon to thair bahavior; for a long tima, description took tha placa of any typo of functional analysis (Prsuschoft and Croat, 1984). Mora recently, howsvsr, many studias

havs boon conductsd which ara aimsd toward finding ths functional morphological basis for ths

locomotor and postural behaviors practiced most frequently by hylobatids: bimanual suspensory positions! behaviors. Different approaches havs been taken by different researchers. For example, an approach which utilizes analogies between functionally similar animals to identify anatomical features related to function has bean used by researchers such as Ashton and Oxnard, 1963,

1964b; Ashton, Oxnard and Spence, 1965; Fleagie, 1978b; Oxnard. 1963,1967; Roberts, 1974; and

Takahashi, 1990. A more reductionist approach has been used by researchers such as Fleagie at al., 1961; Jenkins et al., 1978; Jungers and Stem, 1980,1981,1984; Larson, 1968; Larson and

Stem, 1986; Larson et al., 1991; Stem at al., 19 77,1980a,b; Tuttle and Basmajian, 1977; and Tuttle et al., 1979. These researchers focus their studies on the experimental examination of positional behaviors using analyses of kk«tics, stress patterns, or muscle activity (Fleagie, 1976a). The anthropological community has yet to bring together a compete and agreeable synthesis of the work that has been done. This paper will examine some of the various conclusions which have been reached concerning the shoulder region of suspensory primates, as wen as some of the problems which have arisen, and will look at skeletal data from two closely related suspensory primates which use different amounts of suspensory behaviors in an effort to Identify functional and atometric characteristics of the shoulder girdle in these highly specialized primates. 2 Deflnfflnna

First and foremost, a definition of the word locomotion* is essential. A useful definition has been suggested by Prost (1965), who sttlss that "a prfmats performs a set of locomotor actions, or a locomotor pattern, when its body mass (as opposed to limb mass) is grossly displaced rsiatlvs to ns physical surroundings* (pg. 1200). This definition effectively distinguishes locomotion from posture, which Prost describes as a state in which "the positional relations between a 's body mass and environmental surroundings remain relatively stable* (pg. 1200). It is important to note that posture is by no means neceaaarily a static state of being, for in addition to sitting or standing it includes such activities as hanging beneath a branch or clinging to the trunk of a tree while feeding.

It is entirely possible that a given postural activity may require greater energy expenditure than many locomotor activities (Prost, 1965), and the postural repertoire of an animal must not be ignored when studying relationships between behavior and anatomy. Positional behavior can be defined as the totality of the locomotor and postural repertoire.

The classification of the different types of positional behaviors is a complicated matter, for there are nearly as many dassificatory schemes as there are reseerchers studying the phenomena being classified. The history of the study of brachiation provides an example of this confusion.

The term "brachiation' has long been associated with , including the proposed hominkf ancestor (HoHihn, 1984). This was primarily baaed on anatomical studies, which pointed to many similarities between the skeletal and muscular morphology of lesser apes, great apes, and humane

(HoHihn. 1964). It was proposed by Avis (1962) that our ancestors brachiated, and that brachiation was preadaptive for , due to its requirement of an upright orientation ollhe body (Clark,

1969; Keith, 1899; Gregory, 1928).

Although brachiation, or "-swinging", has been addressed in the literature for quite some time, a precise definition of the term which can be agreed upon is difficult to come by. it has been described as 'all bimanual suspensory locomotion, with or without the fair (Takahashi, 1990, pg.71 3

), or a t "bimanual progression along or between overheed suuctures for a distance of antral

metres without th« intermittent use of other typw of positional behavior and without support by the

hindlimte or tail" (Hollihn, 1964, pg.66). The former definition could include the locomotor activities

of both A fla t and Alouitti. while the latter could only truly refer to and siamangs. For the

purpose of this study, brachiation will be defined as bimanual suspensory progression without the

support of the hindlimbe or tail.

In addition, a locomotor category known u "semibrachiation" was introduced into the

literature in the early 1960's. Napier (1963) describee the positional behavior of a semtorachiator as

"basically that of an arboreal quadruped, but, in addition, a variable amount of time is spent swinging

by the and leaping with the forelimbe outstretched to grab a handhold* (pg 166). New World semibrachiators included Atei—. and Old World semtarachiators Included Cotabus and t j m l i

(Mittermeier and Fleagte, 1976). Mtttermeier and Fleagle (1976) suggest that this category should probably not be retained, due in part to the great diversity of locomotor patterns which are Included within the group.

The shoulder performs different roles in different groups of primates, depending upon the primary positional adaptations present within the group. Primates which move primarily by quadrupedal and , for instance, subject their to mainly compressive forces, while bimanual suspensory primates subject their shoulder girdle to tensile forces (Oxnard,

1967). In initial studies of functional anatomy, associations between particular skeletal features and positional behavior was often done using analogies. For example, there are certain characteristics of the shoulder which are found In hominoida almost exclusively. These were (and are) often considered to be related to the utilization of overhead suspensory positional behaviors, because this type of behavior is found almost exclusively in hominoids (Larson, 1968). Support for this idea came 4 from the fact that there Is anatomical convergence in this region between hominoids and New World suspensory primates, such as M alta. This convergence is reflected in such anatomical traits as axial elongation of the and the presence of a cranially directed glenoid fossa (Andrews and

Groves, 1976; Erikson, 1963).

By dividing primates that use different patterns of positional behavior into different locomotor groups' (quadrupeds, brachiators, semibrachiators, bipeds and hangers), Ashton and

Oxnard (19 6 3,1964b) sought to identify elements of the shoulder related to specific locomotor function. The/ chose a number of scapular features which appeared to be related to known differences in the musculature of the different groups of animals, or to orientation of the . These features were compared among the four locomotor groups within the Anthropoidea

(quadrupeds, brachiators, semibrachiators, and bipeds), and between the two groups within the

Prosimii (quadrupeds and hangers). The table below illustrates which primate groups were given as

Anthropoid quadruped compressive forces

Brachiator tensile forces Hutahrtat Ponflo Semibrachiator subject intermittently to both of the PrfflhytW N m lir AJomttr above kinds of forces Prosimian Quadruped comoressivt forcts lnnir Puftteua ftdanft Hanger tensile forces

Their results (Ashton and Oxnard, I964ab; Oxnard, 1967) suggested that the functional muscular group that is most closely related to bony differences in the scapula is the group responsible for arm-raising, or abduction of the combined with scapular rotation. They found that muscles associated with scapular rotation, such as the cranial portion of the trapezius and the radiating digitations of the serratus magnus, form a more efficient mechanism for rotation in brachiators, because they are better developed in these than in other primates. They also claim that tals greater efficiency is due to the more lateral insertion of the trapezius in brachiators, which s

enables the muscle to act more effectively u a rotator. They add that tht scapular spins and tha

acromion process ars oriented mors obliquely In brachiating animals, in order for the upper fibers of

the trapezius to meet the spine at nearly right angles and therefore act more effectively as a force

couple. Another muscle-related bony feature discussed in their study, caudal prolongation of the

inferior angle of the scapula, increases the extent to which the sarratus magnus can radiate caudaDy

and is claimed to be an added advantage in the rotation of the shoulder girdle in brachiators. The

more distal insertion of the deltoid, the principle abductor of the humerus, was also suggested as providing an advantage in armraising in brachiating primates.

Differences related to the orientation of the shoulder joint between the locomotor groups

were also found. In brachiating primates, the glenoid foesa and the lateral end of the were

found to be more cranially oriented than in other groups (associated with the arms being habitually

raised above the head), and the entire joint is projected laterally from the body wall in order to

facilitate a greater range of arm movement In contrast to this condition, the glenoid and lateral clavicle in quadrupeds is oriented less cranially, and the joint does not project far laterally. In these features and in the features mentioned previously, Ashton and Oxnard (1964b) state that the semibrachiator group occupied "a position intermediate between brachiators and quadrupeds,

[which] correlates with the fact that these animals sometimes adopt one pattern of locomotion, and sometimes the other" (pg. 64). Within the Prosimil, they state that parallel contrasts could be seen between the hangers and the quadrupeds, which relate to the feet that "in all hangers (as in brachiators), the arm is raised more frequently and for longer periods than in quadrupeds, and that when in a raised or semi-raised position, it is frequently subject to tensile forces* (64). Oxnard

(1967) noted that the dimensions they found to be related to locomotion were mainly located in the lower half of the scapula, in the peris of the which make up the "heavy bony buttreesee" which are related to the bearing of major stresses. 6

Mora recent studies have utilized electromyography (EM G) in order to identify which

muscles are activra during partfeuriar positional bshaviors. Tho results using B U G have often conflicted with earlier studies. For example, in an EM G study of great spes conducted by Tuttle and

Basmajian (1977), the cranial trapezius and the caudal serratus anterior (serratus magnus) were found to be inactive during arm-raising behaviors. This finding was contradictory to the functions of these muscles proposed by Ashton and Oxnard (1969,1964). The explanation for this result which was proposed by Tuttle and Basmajian was that the cranial orientation of the glenoid in apes reduced the need for scapular rotation during armraising, and thereby limited the activities of the two muscles. This explanation, however, leaves an important question unanswered: why would the elaboration of these muscles and their bony attachments be favored by natural selection if they were no longer needed (Larson at al., 1991)7

Other studies have contradicted the results of Tuttle and Basmajian. In an EM G study of the caudal serratus anterior in ateiines and A to u H H functional differentiations were found within this muscle (Stem et al., 1960a). Most of the muscle, except the extreme caudal bundles, was active in scapular rotation during armraising, whereas the caudal bundles were active in the transmission of weight from the trunk to the limb during suspensory positional behaviors (Stem, 1980a). The caudal serratus anterior and the trapezius were both examined in an EM G study of lar gfebons (Jungers and Stem, 1964), and similar results were found. The caudal trapezius and the lowest digitations of the caudal serratus anterior were both active in controlling the rate of descent of the trunk past the scapula during the first half of the support phase in brachiation, as wall as during pendant suspension (Jungers and Stem, 1964). It was also suggested that the activity of the cranial trapezius during other locomotor phases was related to head turning rather than scapular movement. Larson etal. (1991), in an EM G siudy of chimpenzeee, also found the high and middle bunefles of the serratus anterior to be active during armraising In producing scapular rotation. This finding supports the contention that the degree of caudal elongation of the infarior angle of the 7 scapula it related to toe prevalence of arm elevation in an animal's positional repertoire (Ashton and

Oxnard, 1964; AaNon at a!.. 1966; Oxnard, 1963,1967). Larson at al. (1991) diaagraa, however, with tha idaa that tha crankwaudally oriantad bundles of tha caudal expansion of tha sarratus antarlor ara an adaptation lor more efficient arm-raising, sines tosy ahowad no EM Q activity during ajparimontal arm alavationa. They propose instead, in agmement with Siam at al. (1960b), that thasa lowar dgHatons add axtra propuMon during brachiation without pulling tha scapula vanlrally.

Thair rasutts agraa with thoaa of Tuttla and Basmajian (1977) conoamfng tha rola ol tha cranial trapezius, in which they found negligible activity during armraising. Larson at al. (1991) attsmpt to explain the enlargement of this muscle In brachiating primates by proposing that it is rsiatod to muscular demands for greater head control in apes, which would bs consistent with tha fact that it is during head-turning that this muscle is moat commonly racrultad. Tha low level of activity of this muscle during armraising, howevsr, calls into question tha association bstwesn the orisntatton of ths scapular spine and the role of the cranial trapezius in rotating the scapula during arm elevation

(proposed by Ashton and Oxnard, 1964; Ashton st al., 1965; Oxnard, 1963,1967). Since tha angulation of the scapular spina in brachiating primates would seam to remove tha ability of the cranial trapezius to rotate the scapula, Larson at al. (1991) suggest that this high (aval of angulation may be related Instead to tha activity of tha dorsal members of tha rotator cuff during bimanual positional behaviors.

Tha glenohumeral muscles, a group of short and powerful musdee which act acroas tha glenohumeral joint, consist of the tares major, the deltoid, and tha muscles of tha rotator cuff (tha supraspinatus, infraspinatus, tares minor and subscapularis) (Larson and Stam, 1996). Inman at al.

(1944) proposed than the muscles of tha shoulder work as a pair of force couples* while raising tha arm, one ol which rotates the glenoid cranially, and another which elevates the humerus. These authors propose that it is in tha elevation of tha humerus that tha scapulohumeral muscles play their major role. They divide the scapulohumeral muscles into an upper unit constating of tha deltoid and 8 supraspinatus, which provide power for elevation, and lower unit consisting ot the remaining cuff muscles, which depress the humeral head. The combination of these two muscle units therefore allows the deltoid to abduct the humerus, rather than just displace it superiorly (Inman et al„ 1944).

Roberts (1974) proposed a second function of the rotator cuff: that of the stabilization of the shoulder joint in primates whose forelimbe require greater mobility, especially those in which the forelimb is frequently used in overhead activities which produce a large amount of transarticuiar tensile stress. He proposed that the presence of relatively large scapular fossae would reflect an emphasis on rotator cuff function, for forelimb movement or as joint stabilizers. After examine t the scapulae of a number of primate genera, he found a high incidence of large supraspinous fossae in animals which habitually use the forelimb above the level of the shoulder. He associated the presence of a relatively large infraapinous fossa with a habitual lifting of the weight of the body during climbing. However, there are some notable exceptions to this rule: Alomtts (principally a slow quadruped) has a relatively larger supraspinous fossa than (a "brachiator*), fiflflili has the largest supraspinous fossa of any primate, and Pongo has a large infraspinous fossa compared to other climbing, large-bodied primates (Larson and Stem, 1966).

Larson and Stem (1966) conducted an EM G study of the glenohumeral muscles in , and found the posterior deltoid to differ in function from the middte and anterior regions of the muscle. They also found the function of the teres minor to be completely different than that of the other rotator cuff muscles, and similar in function to the posterior deltoid and the teres major. Since it was found to be inconsequential in armraising, the force couple’ couldn't rely on this muscle to counter the upward pull of the deltoid, instead, this muscle was active during behaviors like hoisting and the propulsive phase of arm-swinging. The supraspinatus, infraspinatus, and subscapularis were found to be completely different in their functions. The supraspinatus was active as a pure abductor during the initial phases of brachial elevation, and the infraspinatus acted as an abductor/lateral rotator during the swing phase of vertical climbing. The subscapularis was found to 9

be differentiated internally, with the highest portion acting as an abductor/medial rotator during late

stages of brachial elevation, and the lower two-thirds grading from pure medial rotator to

abductor/medial rotator while positioning the hand to make a grasp. The implication that these

functional differences between the muscles in the rotator cuff have a related effect on the infra- and

supraspinous fossae is that Roberts' (1974) conclusions are probably too simplistic to reflect the real

picture, since each of the muscles in the cuff will effect the scapula differently (Larson and Stem,

1966). Even the subscapuiaris, since it arises from the entire ventral fossa of the scapula, could

possibly effect the shape and size of the two ventral fossae (Larson and Stem, 1966).

Larson (1988) examlnsd the function of the subscapuiaris in gfebons and chimpanzees and found differences between the two animals. In both species, the muscle was quite active as a

medial rotator during the support phase of climbing. However, in other positional activities, such as the swing phase of vertical climbing, the support phase of brachiation, the re-eievationofthefree limb during the swing phase of brachiation, the medial rotation of the arm after release at the beginning of the swing phase, and a voluntary reach in the scapular plane, the subscapuiaris showed a greater amount of participation in the gibbon. This action of this muscle in the gibbon would begin earlier during a given movement, would occur at a higher intensity, and would involve more of the muscle mass. Why would the gibbon have such a greater use for a powerful medial rotator? Larson (1968) suggests that the need could be related to the degree of torsion found in Ihe gibbon humerus. When the hominoid scapula became situated on the dorsum of a transversely widened , the glenoid fossa became more medially oriented, as did the head of the humerus

(through torsion) (Larson, 1968; Le Gros Clark, 1969). Larson (1968) found that while gibbons (and siamangs) have a larger humeral torsion angle than most monkeys, they exhttt less torsion than chimps and all other hominokls. This causes the arms to be "set" laterally, and creates a need for the arms to be rotated medially in order to perform most positional behaviors. Larson (1988) also suggests that this limited torsion in gibbons may reflect "a compromise between the need to maintain 10 a transverse axis at the joint consequent to scapular reorientation and the demand for extreme positioning of the elbow during arm-swinging* (458). African apes may never have exhibited this specialized trait, or may have undergone reduced selection pressure for a large range of lateral arm rotation, since they are much less reliant upon brachiation than hytobaWs. In the same article, she offers an alternate explanation, suggesting that the limited degree of torsion in the gibbon may be related to the use of ricochetal brachiation, where the abducted humerus must take an extreme laterally rotated position at the termination of the support phase.

In the preceding discussion, It is apparent that many of the skeletal and muecular features which have been associated with bimanual positional behaviors are related to things such as armraising, the positioning of the arms above the shoulders, and general mobWty of the shoulder joint These have been described as *brachlatlng* adaptations by many researchers, for a variety of reasons. As was discussed earlier, the terms "brachiation* and Drachiakx* have been used in different ways by different researchers. Great apes have commonly been included among the brachiators in part because of their anatomical similarities to lessor apee (Andrews and Groves,

1976), but brachiation plays a much smaller role in their locomotor repertoires than was once believed (Fleagle et at, 1961). £ |Q p«n)geu« has been observed to use only 21% brachiation in their locomotor repertoire (Susman, 1984). Similarly, only 19.8% of locomotor bouts was represented by brachiation in Pnnannvnm iui (Sugardjito and van Hoof!, 1966), as was .8% of all locomotion observed in £iq troglodyte (Hunt, 1961). Even spider monkeys, which use a great deal of suspensory positional behaviors, are less dependent upon these patterns than gfcbons and siamangs (Cant, 1986; Fleagle, 1976a; Mktermoier, 1978). MHtermeier (1978) found the percentage of brachiation in locomotion during travel and feeding in AtflillfHflfiCQld to be 26% and 25.3%, respectively. For Atflifl&QiailfiUl. the percentages were slightly higher 38.6% and 27.7% . While these values are high, hylobatkts depend on brachiation to a still greater degree, as will be discussed later in the paper In the section titled "Hyiobstid positional behavior". The form of 11

brachiation used by hylabatkte and spider monkeys is also qualitatively different. Gtftone seem to

have refined the pendular motion of brachiation to its greatest advantage where speed and energy

efficiency are concerned, whereas spider monkeys have sacrificed some of these advantages for

the extra safety advantage of prehensile tail use (Fleagle, 1974; Jenkins et al., 1978; Jungers and

Stem, 1981,1964; Stem et al., 1980). Therefore, a strong case can be made for the pattern of

positional behavior which is practiced by the gtobons and siamangs being fairly unique among

primates, and it is these animals that should be used as the morphological basis for brachiation,

rather than spider monkeys or other primates which may brachiale infrequently (Andrews and

Groves, 1976; Takahashi, 1990).

Takahashi (1990) conducted a multivariate analysis of skeletal features in a number of different primates, in order to determine which features set lesser apes apart anatomically. These

included a number of linear dimensions of the scapula, as well as lengths and diameters of the

humerus, radius, ulna and clavicle. The only feature of the shoulder which she found to be unique to hylobatide was the presence of a well-developed scapular spine, which she related to a strong attachment for the trapezius and deltoid muscles for increased efficiency and force in the elevation of the humerus. She admits that strong muscles performing this function would also be necessary in primates which are habitual vertical climbers, but suggests that brachiation may require a more powerful and/or different type of forelimb elevation than does vertical climbing.

A large, bulbous humeral head is a feature which many researchers have associated with brachiation (e.g. Erikson, 1963; Fleagle, 1976b), largely because of the need for extensive joint mobility in brachiating primates. However, in TakahashPa (1990) study, this trait did not dtotinguiah the gibbon from either other apes or spider monkeys. Instead, this variable (and in particular the measurement of greater diameter of the humeral head) seemed to especially separate howler monkeys from the other primates. Gebo (1992) found climbing to make up a considerable percentage of the locomotor repertoire of AJouitti Qillisii (37% of bouts). If one includes bridging 12 behaviors in the climbing category (e.g. Reagie, 1976a), which made up 10% of this animals locomotor bouts, climbing appears to be quite a prevalent locomotor behavior. In addition to this evidence, Schon-Ybarra and Schon (1987), who define "climbing* in a way similar to Fleagle

(1976a), describe climbing as "the most frequently recorded of an forms of locomotion" in a study of

Alouatta aaniculua (pg. 75). Since movements associated with climbing require extensive joint mobility, and since howlers depend upon climbing a great deal in their positional repertoire (Cant,

1986; Fleagle and Mittermeier, 1980; Gebo, 1992; Schon-Ybarra and Schon, 1987), it seems likely that this enlargement of the humeral head may be related to climbing behaviors. Axial elongation of the scapula (revealed as the length of the vertebral border) has also been suggested as a characteristic of brachiators (Andrews and Groves, 1976; Erlkson, 1963; Roberts, 1974), but this measurement also failed to distinguish gibbons from other primates in Tnkahashi's study, as did the other variables related to scapular elongation. She notes that howlers, sloths, and slow-climbing lorisines all have a well-developed vertebral border of the scapula, and suggests that this feature may also be related to climbing behaviors, rather than brach' tkm.

Fleagle et al. (1981) propose that most of the features which have been traditionally associated with bimanual positional behaviors may in fact be related instead to climbing. Climbing, in these experiments, is defined as "vertical climbing up a single continuous support" (Fleagle et al.,

1981). Their evidence comes from different researchers' EM G studies of various shoulder muscles, in which the muscles are more active during climbing behaviors than they are during bimanual positional behaviors. For instance, Stem et al. (1977), in their examination of five shoulder muscles in Ataltt and Ugottuia, found them to be as active or more active during vertical climbing than they were during brachiation. Stem et. al (1980 a,b) found the same to be true for the pectoralis major and the serratus anterior in a variety of different primates, and Jungers and Stem (1980) found six shoulder muscles to be more acMvo in climbing and hoisting that in either slow or ricochotal brachiation. Furthermore, Tuttle and Basmajian found shoulder and forelimb musdee to be most 13

active during hoisting and largely silent during passive suspension (Tuttle, Basmajian and taMda,

1979). These results make a strong case fa the contention that many of the muscular features

which have been discussed above are related to vertical climbing, rather that bimanual suspensory

positional behaviors. Fleagle at at. (1981) note, however, the wide range on locomota behaviors

which are regarded as 'climbing’ by various authors. "Climbing*, as used by some authors, has

included behaviors such as three- a four-limbed suspensory behaviors, bridging, and hoisting

(Fleagle, 1976a; Fleagle and Mittermeier, 1980; Mittermeier, 1978). If these behaviors contribute

significantly to the locomota repertoires of climbing and brachiating animals, they could have a

different mechanical Importance than that of vertical climbing in the experimental setting, and these

differences should be considered before an entire suite of anatomical features is assessed as being

related to climbing behaviors.

As HoMhn (1984) observed, comparative anatomical studies (including most of the ones discussed above) have often focused on comparisons between genera, "probably because related species were commonly supposed to belong to the same locomota category". However, differences in locomota behavior and morphology can and do exist between closely related species,

'and such lumping may obscure differences related to behavior" (HoMhn, 1984). Comparisons of functional anatomy and positional behavior between closely related species are less common,

although notable exceptions include Chartes-Dominique, 1977; Crompton, 1984; Fleagle, 1976b,

1977; Garber, 1991; and Mittermeia, 1978. Two closely related species that use a high degree of himannal nnaitlnnal hahavlnia am tha tor fthtwi IHylnhrt— lari and tha aiaman/i /MuIfthataa

(Symphaianomi ayndaetvkjsi. They do differ, however, in the extent to which bimanual suspensory positional behaviors play a role in their positional repertoire. These differences wM be discussed below. 14

The siamang: HytabUgg (Svnmhi lingua) svnriactvlua

The most detailed data concerning the positional behavior of this animal comes from

Fleagle's field studies of the Malayan siamang, which consisted of 800 hours of observations on adult individuals in the Krau Game Reserve, Pahang, Malaysia, from 1972 to 1975 (Fieagle, 1976a).

The following information is taken from Fieagle (1976a).

There are four locomotor patterns which make up the locomotor repertoire of the siamang

Fieagle studied: brachiation, climbing, bipedalism, and leaping. Fieagle defines brachiation as bimanual, suspensory progression, and notes that siamang normally do this along, rather than between, supports. He describes the movements of the brachiating siamang as 'pendular', with the trunk rotating up to 160 degrees in each swing. As has been noted by other researchers through anecdotal observation (i.e. Tuttle, 1969,1972; CNvers, 1972), the siamang uses leas ricochetal brachiation (brachiation in which the animal "throws* Itself from branch to branch) than the smaler gibbons.

Fieagle defines climbing as 'any continuous progression involving three or more limbs" (pg.

248). This type of progression is performed on a variety of different sizes and types of supports at a variety of different orientations, and appears to be primarily forelimb dominated. This locomotor activity also includes 'bridging", or climbing across a discontinuity in the canopy.

The remaining two locomotor activities, bipedalism and leaping, play a relatively small role in the siamang's locomotor repertoire. Bipedalism is performed with the trunk inclined forward and the arms abducted at the shoulder for balance, and involves a good deal of rotation at the pelvis.

Leaping in siamangs is always from a higher point in the canopy to a lower one, and often requires the animal to "pump in place* before attempting the leap, apparently in order to gain momentum.

The horizontal distance covered by these leaps is usually not very large, rarely reaching 10 m, while the vertical distance covered may be twice as large. IS

Reagl# divides Ns locomotor observations into two categories: locomotion during travel and locomotion during feeding. Locomotion during travel consists primarily of movement between feeding sites or to sleeping trees, and usually covers relatively large distances along specific routes, or "arboreal highways'. Locomotion during feeding is defined as locomotor movements from feeding posture to feeding posture within a food source, during a feeding session. In locomotion during travel, brachiation accounted for 51% of the observed locomotor bouts, while climbing accounted for

37% . Brachiation accounted for approximately two thirds of each kilometer traveled, while climbing accounted for about one quarter. During feeding, however, climbing was the dominant locomotor pattern, accounting for 74% of observed bouts, whereas brachiation accounted for 23% of bouts. In both categories of locomotion, brachiation occurred more often on boughs (branches > 10 cm) than on twigs (< 2 cm), and climbing occurred more often on twigs and lest on boughs.

Postural activities during feeding were divided into suspensory (hanging) postures (moat common while feeding on fruit) and seated postures (most common while feeding on leaves). In

72% of the observations of suspensory postures, three Nmbe were used to support the animal, while bimanual support accounted for only 19% . Three Hmba were also used to support the animal during

64% of the observations of seated postures. Suspensory postures were found to occur more often than expected on smaller sized supports, and seated postures, as would be expected, occurred most often on large supports.

The lar gibbon: Uylgfailttltt

Much of the available information on the positional behavior of the white-handed, or lar, gibbon comes from early studies by Carpenter (1940) and EHefson (1967,1968). However, the information on positional behavior in these studies was neither systematic nor quantitative, and can therefore not be compared directly to the quantitative field data on the siamang. Fiesgle (1978a) therefore spent 50 hours observing lar gibbons using th« same methods he used in observing the 16 siamang. He found that in locomotion during travel, brachiation accounteu (Or 56% of locomotor bouts, while climbing accounted for only 21% (compared to the siamang's 37%). He also noted that more of the gibbon's brachiation is ricochets! than is found in the siamang. In locomotion during feeding, 45% of locomotor bouts were brachiation, while only 51 % were climbing (74% in the siamang). The use of different-sized supports and of different feeding postures appears to be fairly similar in the two species. It is in the use of climbing as a means of locomotion that the two species differ from one another, with the siamang climbing more often than the gibbon, possibly in some small branch settings where their large size wont allow them to brachiate as the smaller gfebons do

(FI eagle, 1976a). It must also be kept in mind that given the limited period of observation, Fleagle's

(1976a) data on ty flb n n ii£ may not accurately represent the complete positional repertoire of the species.

Ailomatry

There appear to be differences between the behavior of these two closely related animals, but they also exhibit a further difference which would have an effect on their anatomy: differences in body size. The body weight of the average siamang is around 11 kg, while that of the lar gibbon is about half that (Jungers, 1985). It hrs been argued that "body maos, more than any other descriptive feature, is the primary determinant of ecological opportunities, as well as of the physiological and morphological requirements of an animal* (Undstedt and Caider, 1961, pg.2). The study of this relationship between body size and adaptation is known as aliometry. The simplest and most common form of an allometric analysis is a bivariate plot, with some measure of body size on the abscissa and a biological variable on the ordinate (Fleagle, 1985).

Allometric studies in primates have frequently focused on various limb proportions. For example, Biegert and Maurer (1972) have argued that within catarrhines, both foreiimbe and hindlimbs exhibit positive allometric scaling. In other words, as body size increases, limb length 17 increases more quickly than would be expected, and the slope of a bivariate scatterplot would be greater than one. Jungers (1964), however, points out some faults in their study, including a criticism of their use of skeletal trunk length as a measure of body size. He found skeletal trunk length to be negatively ailometric with body size in catarrhines, which would cause any variable which scales faster to exhibit positive altometry when scaled against it. This illustrates the importance of choosing a body size variable which scales isometrically (slope ■ 1) with body weight, in order to ascertain accurate ailometric relationships. It must also be realized that the attribution of differences in limb lengths between the smallest and largest catarrhines to aUometry would have very little explanatory power, for differences in behavior are very likely to play a role, as well

(Fleagle, 1965).

Another area of study in primate skeletal allometry involves limb joint surface areas. Swartz

(1989) has argued that joint surface area scales positively against body size in order to offset the disproportionately large increase in joint stresses which occur with increases in body size. Godfrey et aU. (1991) disagree with this conclusion, finding instead that positive allometry of joint surface area is not a universal phenomenon among . They found that joint surface areas exhfaited isometry in groups of animals which are functionally similar, and that deviations from isometry occurred within groups that included animals which were functionally dissimilar. For example, a positive allometry was found in joints of anthropoids, which reflected the functional differences between the hominoids (which use a lot of forelimb suspension and climbing) and the cercopithecoids and ceboids (which rely more on quadrupedal walking and running) (Godfrey et at.,

1991). The positive allometry found in the shoulder joint (humeral head) was attributed by the author to the range of movement in the hominoid shoulder joint being greater in all directions than it is in other primate groups.

As RoNinson and Martin (1961) noted, "limb proportions (as wen as many other skeletal features] are influenced by an intricate combination of body size and locomotor specialization* (pp. 18

400-40?). Tharelore, in a consideritton of ttw functional analomy of llw shouktor of t«vo suaperwory primates of different body sizes, both behavior and aHometiy must be addressed. In the following study, numerous dimensions of the scapula and humerus which have been related to bimanual suspensory positional behaviors by various researchers are measured in the siamang and the lar gibbon. The results will be analyzed taking both the dHferences in body size and positional behavior between the two animals into consideration, in an effort to answer the following questions:

1. If differences in the shoulder joint complex do exist between the animals, are they a function of behavior or of body size?

2. If differences found are assumed to be related to function, how might they be interpreted?

3. If differences between the animals are attributed to allometry, how might they be interpreted?

The blowing example illustrates the approach that will be taken in the interpretation of the data. If it is the prevalence of bimanual suspensory behaviors that influences the degree of development of the scapular spine, for example, one might expect to find a greater degree of development relative to body size in the gibbon since it brachiates more frequently than the siamang. This result might manifest itself by having the gfcbon values on a scatterpiot fall consistently above the regression line, or forming a completely separate duster altogether.

However, if the differences in the degree of scapular spine development are due to allometry, one might find the scapular spine measurements for both gibbons and siamangs fitting on the same regression line, being very closely correlated with body size without either species consistently being located above or below the line.

The same might be said about variables measuring axial elongation of the scapula. If the presence of an elongated scapula can be considered to be c result of caudal elongation for the attachment of the high and middle bundles of the serratua antarior, and this muscle has been shown to be more active during climbing than during brachiatkm (Stem et al., 1960a), one might eepect to find a relatively greater degree of elongation in the siamang, which dimbe more frequently then the 19 gibbon. H no meaningful deviations from the best fit line in the plot of axial elongation against body size are found, however, the results would suggest that this variable is more closely related to body size than function.

The scapular index (width of the supraspinous fossa/ width of the infraspinous fossa * 100) will also be assessed, in order to examine the suggestion that a relatively large supraspinous fossa is functionally related to brachiation (Roberts, 1974). If this is the case, the scapular index might be expected to be greater in the gibbon than the siamang.

In addition to these linear measurements of the scapula, angular measurements will be taken in order to assess the degree of obliqueness present in the scapular spine and the degree to which the glenoid fossa is cranially directed. If a higher degree of obliqueness in the scapular spine is related to the activity of the dorsal members of the rotator cuff during bimanual positional behaviors (Larson, 1991), this feature could be expected to be more prevalent in the gibbon than in the siamang. Similarly, if the glenoid fossa is particularly cranially directed in animals which brachiate most frequently, one would expect to find it more cranially directed in the gibbon.

As discussed earlier, a large, bulbous humeral head has often been suggested to be related to the need for a greater range of forelimb excursion in brachiating animals (e.g. Erikson, 1963;

Fleagie, 1976c). This trait will be evaluated in a way similar to the way described for the linear scapular variables discussed above, in order to determine whether or not gibbons have a significantly larger humeral head than the siamang. If the reverse situation is found to be the case, it would support Takahashi's (1990) contention that a large humeral head is related to climbing rather than brachiation.

A total of eleven specimens were measured in this study. Due to the paucity of skeletons available, the only requirement for an animal to be considered mature enough to be measured was 20

that it had one humerus with the epiphyses fused. The sample consisted of both wild-caught and zoo specimens, but some of the zoo specimens may have been caught in the wild during some point

in their lives, and then brought to the zoo. The information available at the Field Museum was not specific enough to allow discrimination between these different categories. In addition to these problems, some of the specimens had various broken skeletal elements, which made it impossible to assess some of the measurements relevant to this study. A summary of the specimens can be found in the table below: Snaftlman Snariaa Origin ftnmmanta 95842 H.s.avndactvlua F" S .E. Asia zoo specimen: None Chicago ZOol. Society W 60340 F Zoo specimen: Chicago Broken left Z o o l.P a rk scaoula • 99366 M Zoo specimen: Lincoln Right humerus- Park Zo o epiphysis not fused w 60555 F Zoo specimen: S .E. Asia N one 99761 H lar antallnlriaa M Thailand: N o n e Kamphaengphet Province (wild shot) Q* u a m |u 47400 H .la r M Captive specimen: Ow vpipy Borneo (MfOrTTiM ngm humerus and glenoid fossa 129463 H J a t M ZOo specimen: S .E. Asia N one 12 72 8 2 H .la r F ZOo specimen: S.E.Asia N one 44740 H .la r F Captive specimen Epiphyses on right scapula not all fused 99740 H lar antallnlriaa M Thailand: Kanchanaburi None Province (wild shot) 60679 H .la r M ZOo specimen: Asia None

Measurements

The measurements taken on the scapulae of the eleven animals in the sample for which the measurements were possible consisted of the following: «•

21

1. Width of the supraspirwus fossa (W SF), measured as the distance between the base of the

scapular spine at the vertebral border and the cranial tip of the glenoid fossa.

2. Width of the infraspinous fossa (W IF), measured as the distance between the caudal tip of the

glenoid fossa and the most medial portion of the vertebral border of the infraspinous fossa. This

value, and that of the preceding measurement, will be used to calculate the scapular index, as

defined by Fleagle (1976c).

3. Length of the scapular spine (LSS), measured from the base of the scapular spine at the

vertebral border to the tip of the acromion process.

4. Maximum breadth of the scapular spine (BSS), measured as the widest portion of the acromion

process.

5. Distance between the superior angle of the scapula and the tip of the acromion process (SSA).

This measurement and measurements 3 and 4 will be relevant when assessing the degree of

development of the scapular spine.

6. Distance between the cranial tip of the glenoid foesa and the inferior angle (SOI).

7. Length of the vertebral border of the scapula (SVL), measured as the distance between the

superior angle and the inferior angle.

6. Length of the axillary border of the scapula (SAL), measured as the distance between the caudal

tip of the glenoid fossa and the inferior angle.

9. Distance between the tip of the acromion process and the inferior angle of the scapula (SAI).

This measurement and measurements 6 ,7 , and 8 will be used to assess the degree of axial

elongation present in the scapula.

10. The angle between the scapular spine and the vertebral margin of the scapula (Angl). This

measure will be used to determine the degree of obliqueness of the scapular spine.

12. The angle between the glenoid fossa and the lateral margin of the scapula (Ang2). This win be

used to assess how cranlally directed the glenoid foesa is. 22

The diameter of the humeral head, both minimum (HHDMIN) and maximum (HHDM IN), was also measured on all specimens for which it was posable.

The linear measurements were taken to the nearest .1 mm using sliding calipers, and the angles were measured to the nearest degree using two protractors.

Skeletal mass was measured as the combined weight of the skull, mandfele, pelvis, femora, scapulae and humeri of each individual (Takahashi, 1990). In some individuals, elements of the skeleton included in skeletal mass were attached to other skeletal elements. For instance, the pelvic were attached to the sacrum and/or one lumbar vertebra in three of the individuals, in which case the attached elements were weighed in other individuals and subtracted from the final skeletal mass. I don't coraider this to be particuiaity detrimental to the assessment of the skeletal masa variable, since the skeletal elements which had to be subtracted for calculatfora were small and made negligible contributions to the complete mass. Skeletal mass was chosen to represent body size because it was relatively easy to measure for each specimen, and is highly correlated with body weight (r > 0.99) as well as being nearly isometric with it (slope ■ 1.03) (Takahashi, 1990). The sample utilizes specimens from both sexes, but this is not expected to be a problem since sexual dimorphism is negligible in hylobatids.

In order to insure reliability, a sample of measurements was taken a second time, and the results were compared to the original measurement values. Since the difference between the two trials for each variable were almost never more than 4% , I assume the measurements to be reliable.

Results

Each variable was plotted against log skeletal maas in a bivariate scatterpiot format, using the M YSTAT statistical applications package. A bivariate regression analysis was run for each scatterpiot. The Pearson correlation between the log of each variable and lag skeletal weight, as 23 well as the slope of the best-fit line for each scatterptot and the p values are listed in the table below.

All values refer to the rig lit side of the body. V ariah b I _ __ f i WSF .699 .14 6 .024 WIF .891 .2 76 .000 LSS .76 8 .19 9 .006 BSS .659 .464 .028 SSA .249 .062 .4 8 7 SGI .926 2 9 1 .000 SVL .824 .4 73 .003 SAL .9 15 .3 4 7 .000 SAI .906 .293 .000 A n o l .5 7 7 .632 .081 A n a 2 .451 -.0 5 7 .16 4 HHDMAX .6 70 .294 .034 HHOMIN .73 0 .2 8 7 .0 16

As is illustrated in the table above, every so.apular measurement exhibited a negative allometric relationship with skeletal mass. In the bivariate plots of the scapular variables against skeletal mass, there seemed to be no dear separation between the gibbons and siamange. For example, the siamang variable residuals were not consistently positive and the gibbon variable residuals negative. The general trend was for the siamangs to be found toward the ifg er end of the scatterptot distribution, due to their larger body size. The only variable that showed a very low correlation with skeletal mass was S S A. This variable seemed to show very little change with respect to skeletal mass (slope-.082).

The other two variables measured in order to assess the degree of development of the scapular spine, LSS and BSS, both scaled negatively against skeletal mass, with LSS both scaling more negatively against and being more highly correlated with mass than BSS. The four variables used to assess the degree of axial elongation present in the scapula, SGI, S V L, S A L, and SAI, were all fairly highly correlated with skeletal mass, and scaled negatively against it.

The results concerning the angles were a bit unusual. When Ang1 (on the right scapula) was plotted against skeletal mass, a negative allometric relationship was apparent. After the 24 elimination of an extreme outlier (a gibbon), the r value increased to .894, and the slope was calculated as .796. There appeared to be a certain degree of separation between the gibbons and siamang in the sample (see graph 1). However, values for the left scapula, when plotted, formed two discrete clusters which did not divide evenly into gibbons and siamangs (See graph 2). The same was true for the plot of Ang2, except the general trend was a decrease in the size of the angle as skeletal mass increased, and there did not appear to be any separation between gibbons and siamangs in the right scapula, as in Angt.

In addition to these scatterplots, the scapular index, as defined by Beagle (1976c) was determined for each of the animals for which it was possible. It was calculated as (W SF/W IF) * 100.

The results are summarized in the table below:

Th e mean

values for the siamang 99366 84.542 • 60555 82.941 6 9 .412 scapular indices were 99761 b J a i 97.692 99 .219 47400 96.599 9 2 .6 17 84.084 (right) and 129463 90.370 9 1 .7 0 7 88.040 (left). The 12728 2 96.552 97.534 44740 88.667 9 1.78 1 gibbon mean values 99740 95.513 97.368 60679 90.909 94.156 were 93.757 (right) and

94.912 (left).

In the case of the humeral measurements, negative allometry was the general trend, which could be seen in all of the measurements taken. A very interesting phenomenon could be seen in the plots of HHDM AX, which is illustrated in the graphs 3 and 4. The "A"s represent the siamangs, and the "A"s represent the gibbons The siamangs form at,, iter separate from the gibbons, especially in the HHDM AX variable measurements. While there to a negative allometric relationship within the gibbons (r ■ .972, slope ■ 2), there seems to be no real change within the siamangs as 25 skeletal mass increases (r«.15 # , slope ■ .023). The functional significance of this data will be discussed in the next section.

QjSfiUSSifiQ

The pervasiveness of negative allometric scaling in almost all of the shoulder elements measured is the dominant theme in the analysis of these data. In order to interpret the significance of this trend, it is necessary to go back and look at the functions which have been proposed for these features.

The variables related to the degree of development of the scapular spine showed different patterns when plotted. Because of the lack of significant correlations and recognizable patterns in the BSS (r-.659) and SSA (r>249) variables, only the variable LSS will be discussed. This variable scales very negatively against skeletal mass, and there appear to be no meaningful deviations from the best-fit line. As was discussed earlier, this would suggest that differences in the length of the scapular spine in gibbons and siamangs is related to allometry. This allometric relationship might be interpreted as a result of an evolutionary scenario in which siamangs represent the largtbodied ancestral condition, and gibbons represent an animal that decreased in body weight without exhibiting an equal decrease in muscie mass. If the higher amount of brachiation used by the gibbon is considered meaningful, however, the negative allometric relationship might be interpreted as a net*j in this animal to retain a strong attachment for the deltoid and trapezius. Both the deltoid and the trapezius seem to be important in bimanual positional behavior, contributing to arm elevation and to controlling the rate of descent of the trunk past the scapula during brachiation (Jungers and

Stern, 1964; Takahashi, 1990). TakahasN (1990) found the degree of development present in the scapular spine to be a variable which set gibbons apart from other suspensory primates, which implies that this may indeed be uniquely related to brachiation. If this is the case, then the relatively 26 small amount of development of this feature in siamangs could conceivably reflect the animal's lesser dependence upon brachiation in its locomotor repertoire.

All four of the variables used to assess the degree of axial elongation of the scapula showed a negative allometric relationship with skeletal mass. The axial elongation of the scapula can be considered to be a result of caudal elongation for attachment of the high and middle bundles of the serratus anterior, an arm-raising muscle which has been found to be more active during climbing than in brachiation (Stern at al„ 1980a). Takahashi (1990) also suggested that this feature may hold more relevance as a characteristic of climbers. The relevance of the negative allometry in this feature, like that in the length of the scapular spine, may also be related to the gibbon's possible descent from a large-bodied siamang-like ancestor. It seems that if an axially elongated scapula is a behavior-related characteristic of climbers, it should scale positively, with siamangs having relatively more axial elongation or deviating from the best-fit line in the scatterplot. This was not found to be true of this sample, which suggests either that this feature is not a reliable indicator of the prevalence of climbing behavior in an animal's positional repertoire, or that the behavioral differences between the gibbon and siamang were not significant enough to warrant a meaningful separation between the two species on the scatterplot.

Angf, a measure of the obliqueness of the scapular spine, shows the angle becoming less oblique as skeletal mass increases, with the siamangs in this sample falling slightly above the best- fit line on the scatterplot. This is consistent with the idea that a highly oblique scapular spine is associated with the activity of the dorsal members of the rotator cuff during bimanual positional behaviors (Larson, 1991), since the smaller gibbon uses these behaviors more often than the larger siamang. This separation between the two species only occurred with the right scapula, however.

The scatterplot for the left scapula (see graph) seemed to exhibit no pattern whatsoever (r >.357), and this casts considerable doubt upon the validity of the results for the right scapula. 27

Ang2, a measurement of hew craniaKy directed the glenoid fossa is, shows the gieooid becoming more craniaNy directed aa size increases, although the correlation between this variable and skeletal mass is fairly low (r ■ .451). This result may cast some doubt over the idea that a eranMv di fe tid ctenoid it an adn—ton for Nfoaourt tuiDiniOfv bthivio fi. ftinoo (ho loroor siamang uses leas of these than the gfebon. This feature may instead bs related to climbing behaviors.

The mean scapular index values are lower for the siamangs than for fte gfebone, which would indicate a relatively larger supraspinous fossa in the gibbon. Roberts (1974) has auggeeted that mis is a feature typical of brachistors, but (as discuesed above), there are important exceptions to this rule, and the functional anatomy of the ventral toasae of fte scapula appears to be more complicated than was once believed (Larson, 1991).

The intriguing separation between gfcbons and siamangs in tie plot ol HHOMex and

HHDMin deserves some attention. A large, bulbous humeral head has often been suggaeted to be associated with brachlation (e.g. Erikson, 1963; Fleagle, 1976c). However Takahashi (1990) found that this feature did not set the hyiobatids apart from other primates in her analysis. H did, however, set Alouatta. an animal that uses a great deal of climbing in its positional repertoire, away from the others. She suggests that this feature may be related to the need for groat joint mobMty during climbing behaviors. If this were the case, then the finding that the slamang's humeral heed diameter is in a dess by itselt compared to the gibbon may be considered to possUy be related to the greater amount of climbing used by the siamang in its positional repertoire.

The limits of the research design of this study are numerous, and need to be taken into consideration. The sample size was incredtoly small, a problem that was worsened by the fact that some of the relevant skeletal elements were broken and therefore not usable. The criteria I used to determine maturity in the specimens (having at least one humerus with all epiphyses fused) may be somewhat less than satisfactory. Concerning most of the specimens, it was unclear whether or not 28 the animal was wild caught or bom in a c a p ** setting, and M i may haw had a considerate affect upon its positional behavior. It would be desirable in the long rut to measure a much great* number of specimens of knovm origin.

The shoulder joint has been studied extsnsfvaly by a number of researchers. Its functional anatomy has boon axaminad in groat detail, both by studying tha gross anatomy of tha rogion and by mors sophisticated taohniquas such as electromyography. By combining that# studfes vrfth fWd

Cats on tha positional behavior o( suspensory primatas in tha wild, researchers have coma up with various features that tiay have propoeed to ba functional adaptations tor a posWonel rapartoira in which bimanual suapanaory behaviors ptay a dominate rota. Difficuitias have arisen, however, in

(M M fia loadfic Dotitiooil ctrtaoorisi and mtontno dtftafaot animals to than). wHh In i m I M n s ihU na NI §jy0 yW iAw wftAtniAlB w iw § WliAiift r® wL a a9 am ssstgnvQ 410a kAllAi^AiAluMW OfH ciipponNftAiAftnflAa pnniiR| i. mIm i |p | m h o l yn y y v t op M HORiyy rathar than data on their behavior in tha wild (sat Ripley, 1967). Further Held studies need to ba

COnOUOww •*» OfOtr IQ OllpTvIniV W ^MPiVwV PPNw OP OICniHQni HnpOlOw PfnaiHMI vNony pOfuirM, SnO 6P)0 ng OfnflvIQri Oil fW mWHnif m nyOMIOi, in OfuOf 10 nmf UnOOWinO VIO

VIunloyo WWMw soodalizitionsOWWVIf IOONW WP foundIW FOI in" * thaaaOIWW WMPMWaanim ili VIIt wouldWWWW alaobsralavantW O¥ Wv iOWVOMI toVW oorwldarW l VWVWWV which W1HVM animal'smW flfiP w

OnVvWihouldof W * VwWjoint VI amonoW IIW MW thow W hviobatidt1 VW W W O W lOWraoroaonta fOOW IIO thoIV Vw ancaatralW W OOM VI conditionW U W iO M a Onow llW miohtVVfVMVVl ftontlrtifW IIW W W I tho OIW aspects of positional behavior which tha gtobon and siamang share to ba aspects of common descent. The aspects of these animals' positional repertoires which seam moat simlar are the high prevalence of brachiation in locomotion during travel, and of suspended feeding postures (Fleagle,

1976a). Where these animals differ Is in their locomotion during feeding, in vrtveh the siamang shows a much higher prevalence of climbing than does the gfebon. If the high level of usage of bimanual positional behaviors is considered to be the ancestral hytabaM condition, tha behavior of the siamang would appear to be derived. However, soma researchers have suggested that tha 29 siamang represents the primitive hytobabd condition, based upon a number of morphological characteristics (see Creel and Preuschoft, 1964). This would suggest that die smaller size and higher incidence of brachiation found in the gibbon would represent a derived state. Before a resolution can be reached, a more thorough comparison of anatomy, behavior, and ecology between ail of the different species of hylobatids must be conducted. A synthesis of detailed and quantitative field data concerning the positional behavior of the hylobatids, combined with all of the anatomical research, both simple and sophisticated, which is at the disposal of today's scientists will be necessary before this can occur. Focusing on studies of different species within the Hylobatidae may also help to elucidate the interactions between body size and behavior that shape the anatomy of these highly specialized animals. 30 Angle 1 (Right scapula) 1.78 1 .7 1.65 1.6 1.56 1.5 1.45 1.4 1.35 1.3 1.25 2 .2 ------2 .3 ------2 .4 Log skalatal m att (grams)

Graph 1 31 Angle 1 (Left scapula)

Log skslstal mass (grams) Graph 2 32 Right humeral head diamater (M a x ) 0 .3 8 0 .3 8 0 .3 7 0 .3 8 0 .3 5 0 .3 4 0 .3 3 0 .3 2 0 .3 1 0 .3 0 .2 8 0 .2 8 0 .2 7 0 .2 6 0 .2 5 0 .2 4 0 .2 3 0.22 T T 216 2.2 2 25 2.3 2 36 M 2 45 2:5 2 56 Log skeletal m as* (grama) Graph 3 33 Left humeral head diameter ( M a x ) 0 .3 7 0 .3 6 0 .3 6 0 .3 4 0 .3 3 0 .3 2 0 .3 1 0 .3 0 .2 9 0 .2 6 0 .2 7 0 .2 6 0 .2 5 0 .2 4 0 .2 3 0.22 2.15 T 2 .2 5 2 .3 5 2 .4 5 2 .2 ------2 .3 ------2 .4 Log skeletal maee (grams) Graph 4 34 UfttiflUtM ti

Andrews P and Groves CP (1976) Gibbons and brachiation. In Gfobon and Siamang (ed. DM

Rumbaugh), voi. 4:167-210. Kaiger, Basil.

Ashton EH and Oxnard CE(1963) The musculature of the primate shoulder. Trans. Zool. Soc.

Lond. 29353-650.

------(1964a) Functional adaptations in the primate shoulder girdle. Proc. od. Soc. Lond. 142:49-

66.

— (1964b) Locomotor patterns in primates. Proc. Zool. Soc. Lond. 142:1-24.

Ashton EH , Oxnard C E, Spence T F (1966) Scapular shape and primate classification. Proc. Zool.

Soc. Lond. 145:125-42.

Avis V (1962) BncNatkm^he crucial issue for man's ancestry. Southw. J. Anthropoi. 15:119-46.

Cint JOH HOfffl Locomotion and finding nooturM of jniriar and hovriino montoys: fMd

study and evolutionary interpretation. Folia Primatol. 46:1-14.

Carpenter C R (1940) A field study in Siam of the behavior and social reiattons of the gfobon

/Hfitahi— tot comp. Psychol. Monogr. 16:1-212.

Chlvers D J (1972) The siamang and the gfobon in ths Malay Peninsula. In Gfobon and Siamang

(od, DM Rumbaugh) voi. 1:103-35. Kaiger, Baal.

Clark W E La Qros (1971) The antecedent* ot man, 3rd ad. Edinburgh University Press, Edinburgh.

Crtil^i^V tN V Wm d PrtuichoH t ^tPii t h* m(19B4) m BPSvMuvisdCB V i tViVW^P iPt of it tho W Moor i d m : iNIi iiifl^iV ounnktitivi W llNiv V tViWBlIW tsxonomic tiPttt^iP Vtt anttait

of oiraomovic inu ow r w n o N i in MWiicnoii Hi v i WM b iMt ofocniWiin w * mo vfN i

n in§L0S90r npm, cvomvonsry mo MniviofBi Dtoiogy. cuNVXjfyn umvofiny

m m , eorourgn.

WCromoton W lW IlPf 111RH I(19M I Fonoini MBHPVMI*habitat VWstructure IPIIP wy W tndtocomotonintwosoiciiiofGilatt tif WIPIPVVIlrWM Wt IffV VfrW W i VI Init I

Rodman and Cant (ads.): 73-109. 35

Blefson JO (1966) Territorial behavior in the common white-handed gibbon, HyinhatM lar Unn. In

PC Jay (ad.): Primates: studies in adaptation and variability:189-99. Holt, Rinehart, and

Winston, New York.

Erikson G E (1963) Brachiation in New World monkeys and in anthropoid apes. Symp. Zoo!. Soc.

Lond. 10:135*64.

Fleagle JG (1976a) Locomotion and posture of the Malayan siamang and implications for hominoid

evolutfon. Folia Primatol. 26245*269.

— -(1976b) Locomotor behavior and skeletal anatomy of sympatric Malaysian leaf-monkeys

rPfw«hyti« abtsua and PfaahwtM mauinnhnt' Yrbk. Phys. Anthropol. 20:440-453.

— (1977) Locomotion, posture and habitat utilization in two sympatric, Malaysian leaf-monkeys

tPraahytte and P -**? *1* maiainphfla) in: The Ecology of Arboreal FoUvores. G.G.

Montgomery, ad. Proc. Nat. Zool. Park, 1. Smithsonian Press (Random House, New York).

------(1985) Size and adaptation in primates. In Jungers W L (ad.) Size and scaling in primate

biology, pp 1*20. Planum Presa, New York.

Fleagle JG and Mittermeier RA (1980) Locomotor behavior, body size, and comparative ecology of

seven Surinam monkeys. Am. J. Phys. Anthropol. 52:301-14.

Fleagle JG , Stem JT , Jungers W L, Sueman R L, VangorAK, Wets JP (1911) Climbing: a

biomechanical link with brachiation and bipadafiatn. Symp. Zool. Soc. Lond. 46:359-375.

Garber PA (in press, 1991) Verticel dinging, ama« body sie and the evolution of feeding

adaptations in the CaUitrichinae.

Gabo 0(1992) Locomotor and poatural behavior in K o n tfifliN iiH and CMuifilBVGiOUl- Am. J.

Primatol. 26277-290.

Godfrey L , Sutherland M, Boy D and Gomberg N (1991) Scaling of Hmb joint surface areas in

anthropoid primates and other mammals. J. Zo d ., Lond. 223:603-25. 36

Gregory WK (1928) Were the ancestors of man primitive brachiators? Proc. Am. Philosoph. Soc.

6 7:129 -5 0 .

Hollihn U (1984) Bimanual suspensory behavior: morphology, selective advantages and

phylogeny. In Preuschoft H, Chivers D J, Brockelman W Y and Creel N (ads.): The Lesser

Apes, pp 85-96. Edinburgh Univ. Press, Edinburgh.

Hunt KO (1991) Positional behavior in the hominoidea. Inf. J. Primatot. 12:95-118.

Inman VT, Saunders JB deCM and Abbott LC (1944) Observations on the function of the shoulder

joint. J. Bone Joint Surg. 26:1 -30.

Jenkins FA , Dombrowski P J, and Gordon EP (1978) Analysis of the shoulder in brachiating spider

monkeys. Am. J. Phys. Anthropol. 48:65-76.

Jungers W L (1984) Scaling of the hominoid locomotor skeleton with special reference to lesser

apes. In Preuschoft H , Chivers O J, Brockelman W Y and Creel N (eds.): The Lesser Apes,

pp. 146-69. Edinburgh Univ. Press, Edinburgh.

— (1985) Body size and scaling of limb proportions in primates. lnJungersW L(ed.) Size and

scaling in primate biology, pp 345-82.

Jungers W L and Stem JT (1980) Telemetered electromyography of forelimb muscle chains in

gibbons (Hytobttaa Ila/). Science 208617-619.

— (1981) Preliminary eiectromyographica! analysis of brachiation in gfcbon and spider .

Ini J. Primatol. 2:19-33.

— (1984) Nnesiotogical aspects of brachiation in largtobons. In Preuschoft H , Chivers D J,

Brockelman W Y and Creel N (eds.): The Lesser Apes, pp 119-34.

Keith A (1899) On the chimpanzees and their relationship to the goria. Proc. Zool. Soc. Lond.

1899:236-312.

Larson S G (1988) Subecapularis function in gtobons and chimpanzees: implications tor

interpretation of humeral head torsion in Hominoids. Am . J . Phys. Anthropol. 76:449-62. 37

Larson SG and Stam JT (1986) EM G of scapuiohumaral musdas in ths chimpanzaa during

reaching and "arboraaT locomotion. Am. J. Anal 178:171-190.

Larson S G , Stam JT , and Jungsrs W L (1961) EM G of stratus anterior and trapezius in the

CnimpBnZSi.SCApUIAr rOIAIOvSrAVIARAQ.Ml. J, rnyS.M vwOpOI.oD./i-04.

Lindatedt S L and Caldsr W A, III (1981) Body size, physiological tims, and longevity of

homeofharmic animals. Quart. Rav. Biol. 56:1*16.

Msndsl FC (1979) The wrist joint of taro-toed sloths and its relevance to brseNating adaptations in

tha hominoidM. J Morphol. 162:413-424.

Mittormeior RA (1978) Locomotion andposture in AttilBflMCIaad«nd AttiifiialaGU I- FoHa

PrimaW. 30:161*93.

Mittermeier RA and Flaagia JG (1976) Ths locomotor and postural rapartoim of Atd M flK ttQ Xi

Ana uiODUIflUlIlftli inQ ■ rttviiuAiiOn or v ia momoiof csivgory SAmixicniiDon. m i . j .

Phys. Anthropol. 45235*256.

Napiar JR (1963) BracNafon and brachiatort . Symp. ZOol. Soc. Lond. 10:183-96.

Oxnard C E (1963) Locomotor adaptations in Ihsprimats forelimb. Symp. Zool. Soc. Lond. 10:166*

62.

— —(19871* 1 PPAP * M Th w * w^p i functional M A M N R A H P "™wAPwmofDhokxiy H^A WPWAPABW APPof IhA Ai W^P B^PwPPWAHA^PorimAtA Aihouldtfm PAPA^^AP^P* WBwa s PrAvoalAd APwA^^P^^^^AP bv APW comoATAtivA A^A^AP^p^AAPwP^PA^P

ftntiOiTiHJHiOsHofiiMiCi traQMGnnwisniluncponwonni^jM* a i v k ti« iijjf#*A mV iiu p q i* a *

219-240.

Preuschoft H and Creel N (1964) Anatomy and function. In Prsuschoft H , C ig a rs D J, Broefcaiman

WY and Creel N (eds.): Tha Lasaar Apes, pp. 81-4. Edinburgh Univ. Prase, Edinburgh.

Prod JH (1965) A dsfinitional systam for tha daaaification o( primala looomotion. Amar. Anthropol.

67:1196-214. nptayS (1967) Tha looping ol langurs: a problom in tha study ol locomotor adaptation. Am. J.

Phys. Arthropd. 26:149*70. 38

Robarls 0 (1974) Structur* and function In tha primal* scapula. In FA Jenkins (sd.): Primate

Locomotion, pp 171-200. Academic Prats, New York.

RoUinson J and Martin RD (1961) Comparative upsets o( primate locomotion with apscial

reference to arboreal cercopltheclnee. Symp. Zoo!. Soc. Lond. 46:377-427.

Schon-YbarraMA and Schon MAUI (1967) Positional behavior and Rmb bone adaptations in rad

howriirtg monkeys (AW m w Mnfaifcit) Fo8a Primato!. 49:70-69.

Stem JT , WeUs JP , Jtngers W L, Vangor AK, and Ftoagto JG (1960a) An etockomyograpNc study

of tho oodoralii maior in AfloKnoa ind hMohsttM with n c U rofafonco to tho avokjton of

a para davfculeria. Am . J . Phys. Arrihropd. 52:13-25.

Stem JT , WeUa JP , Junpera W L, and Vangor AK (1980b) An electromyographic study of asrratus

antonor m Aiaanoa ano a k h m o w . Rnpacaaoni for nontmoKi avoaioon* a m j . rn y a .

Anthropoi. 52:323-334.

Stem JT , Welle JP.'Vangor AK , Fleagle JG (1977) Electromyography of some muadee In the upper

limb in t m and t-fs h ri* Yb. Phys. Anthropoi. 20:496-507.

Sugardjito J and van Hooff JARAM (1966) Age-sex dass differences in tha positional behavior of

the Sumatran orang-utan foogpotgauiui M all In the QunungLeueer National Pmk,

Indonesia. FolaPrimatoL 47:14-25.

SuamanRL (1964) The looomotor behavior of JEiQBiallBUI in the LomakotoraeL inSuem anRL

(ad.): The Pygmy Ghlmpanee. Plenum Prase, Nay York, pp. 366-394. owanz 9Mf t a a \iw / a A Av) A \ t inaiuncaonii h * l a « mOtyjnoiQQp ------—t ov wtognin)oonng. L a — — HNvnDjovam K U 2 « i aunocooroo------, ORornopy U in

anthropoid primates. J. ZboL, litnd. 218:441-460.

TakahaaNP WWW w W LK rail (1990) ePwraPI VWraW Morohoiooicol IW P^P^^aPbaaia raP» of raP aitMwInoinoi * P H^P PmPP muNvariata IraPP WWW W i^P raPanahioi W PT^^^W raWof OPtho ^P IraWfotoMinha

of u n b U M and A f t L Fbtia PrimatoL 54:71X6.

Tuttle RH (1969) Knuckle welting and the problem of origins. Science 166:963-61. 39

——(19721...... ^ ®FLreHonai ABi OIWand WbI tvolulonifv ^ wiy B^Po| Ph*h i^ ^ n lM «iii B P R p Ptmufa WPB B B P n ^ j P in> Win * Bp^^Bn y PoB^PPmiNIHB^^Bnumteiiah 1

(•d.)Gtfcon and Starting, vol 1:136-206. Kargar, B m L

Tuttle RH end Batm ijitn JV (1877) Bectrofflyognphy ot ponjM ihoddif rnupdei end hofntnoid

•nMIQQn. I. nvuayw i 01 me nUmfnlB M O TOm B I 01 W m ICipill. TfOL m yi. /VWwOpOI.

2 0 :4 9 1-7.

Tuttla RH, Bumajian JV and IsNda HI (1979). AcMtoaofpongMMghmuadaa during bipadal

bahavior. Am . J. Phys. Anthrapol. 50:123-136.