Electromyography of back muscles during quadrupedal and bipedal walking in primates.код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 93:491-504 (1994) Electromyography of Back Muscles During Quadrupedal and Bipedal Walking in Primates LIZA J. SHAF'IRO AND WILLIAM L. JUNGERS Department of Anthropology, University of Texas at Austin, Austin, Texas 78712 (L.J.S.) and Department of Anatomical Sciences, SUNY at Stony Brook, Stony Brook, New York 11794 (W.L.J.) KEY WORDS EMG, Quadrupedalism, Bipedalism, Hominoids, Baboon, Erector spinae, Multifidus ABSTRACT Despite the extensive electromyographic research that has addressed limb muscle function during primate quadrupedalism, the role of the back muscles in this locomotor behavior has remained undocumented. We report here the results of an electromyographic (EMG) analysis of three intrinsic back muscles (multifidus, longissimus, and iliocostalis) in the baboon (Pupio unubis), chimpanzee (Pun troglodytes), and orangutan (Pongo pygrnueus) during quadrupedal walking. The recruitment patterns of these three back muscles are compared to those reported for the same muscles during nonprimate quadrupedalism. In addition, the function of the back muscles during quadrupedalism and bipedalism in the two hominoids is compared. Results indicate that the back muscles restrict trunk movements during quadrupedalism by contracting with the touchdown of one or both feet, with more consistent activity associated with touchdown of the contralateral foot. Moreover, despite reported differences in their gait preferences and forelimb muscle EMG patterns, primates and nonprimate mammals recruit their back muscles in an essentially similar fashion during quadrupedal walking. These quadrupedal EMG patterns also resemble those reported for chimpanzees, gibbons and humans (but not orangutans) walking bipedally. The fundamental similarity in back muscle function across species and locomotor behaviors is consistent with other data pointing to conservatism in the evolution of the neural control of tetrapod limb movement, but does not preclude the suggestion (based on forelimb muscle EMG and spinal lesion studies) that some aspects of primate neural circuitry are unique. 0 1994 Wiley-Liss, Inc. Much study has been devoted to investigating primate quadrupedalism in comparison to that of nonprimate mammals (e.g., Prost, 1965; Hildebrand, 1966, 1967; Kimura et al., 1979; Rollinson and Martin, 1981; Vangor and Wells, 1983; Reynolds 1985a)b;Vilensky and Larson, 1989; Larson and Stern, 1987, 1989; Demes et al., 1994; see Vilensky, 1987, 1989 for reviews). As a result, numerous aspects of primate quadrupedalism have been proposed to be unique among mammals, including their preferred gait pattern (Muybridge, 1899; Prost, 1965; Hildebrand, 1966, 1967; Rollinson and Martin, 1981), mechanisms of propulsion 0 1994 WILEY-LISS, INC. (Kimura et al., 1979), and muscle activity patterns (Larson and Stern, 1987,1989). The difference between primates and nonprimates with respect to footfall patterns in symmetrical gaits is well documented. Primates preferentially (though not exclusively) use a diagonal-sequence gait (LhRfRhLf, Lh = left hind limb, Rf = right forelimb, etc.) rather than the lateral-sequence gait (LhLfRhRf) utilized by nonpri- Received March 18,1993; accepted October 21,1993 Address reprint requests to Liza Shapiro, Dept. of Anthropology, University of Texas at Austin, Austin, TX 78712-1086. 492 L.J. SHAPIRO AND W.L. JUNGERS mate mammals (Muybridge, 1899; Prost, 1965;Hildebrand, 1966,1967;Rollinson and Martin, 1981). Primates also have been singled out among mammals for their “frontsteering, rear-driving” rather than “frontsteering, front-driving’’ propulsive mechanism (Kimura et al., 1979). Both of these presumably unique aspects of primate quadrupedalism traditionally have been explained in terms of morphology. For example, it has been suggested that a diagonalsequence gait is more stable for primates than is a lateral-sequence gait if the center of gravity is more posteriorly located in primates than in nonprimate mammals (Rollinson and Martin, 1981). Morphological explanations also hold that the posterior propulsive mechanism is associated with the fact that primates support most of their weight on their hind limbs (Kimura et al., 1979; see also Reynolds, 1985a,b; Vilensky 1987, 1989; Vilensky and Larson, 1989; Demes et al., 19941, and thereby “free” the forelimbs for other behaviors. More recently, however, doubt has been cast on the notion that primates are unique or even consistent in employing a hind-limb propulsive mechanism. For instance, propulsive forces on primate hind limbs may be asymmetrical (e.g., on leading vs. trailing limbs in a gallop; Demes et al., 1992, 1994), and it has been demonstrated that both primates and nonprimates propel themselves predominantly with their hind limbs (Demes et al., 1992, 1994; see also Pandy et al., 1988; Vilensky, 1989). Moreover, Vilensky (1987, 1989) and Vilensky and Larson (1989)present convincing evidence that neither a more posteriorly located center of gravity nor increased hind-limb weight can adequately explain the primate preference for diagonal sequence gaits. These authors point out that the location of the center of gravity should not affect stability at most speeds and gaits chosen by primates, and even so, “the actual difference in CG locations between for example, cats and primates appears to be slight” (Vilensky, 1989: 358). Vilensky and Larson (1989) also note that although diagonal-sequence gaits are preferred, lateral-sequence gaits are not uncommon among primates. Accordingly, these authors offer a neurological rather than morphological explanation for primates’ overall preference for diagonal-sequence gaits, as discussed below (see also Larson and Stern, 1989). Electromyographic analyses have contributed substantially to characterizing the nature of primate quadrupedalism, as well as permitting detailed comparisons between primate and nonprimate quadrupedalism with respect to muscle function. These studies have revealed overall similarity in hindlimb muscle function between primates and nonprimate mammals (particularly domestic cats) when they walk quadrupedally (e.g.,Jungers et al., 1983;Vangor and Wells, 1983), notwithstanding some exceptions (e.g., Jungers et al., 1980; Jungers and Anapol, 1985; Vilensky, 1987). The continuity of hind-limb muscle activity patterns across mammalian groups (and between mammals and other tetrapods) is suggestive of an evolutionary conservatism in locomotor neural circuitry (Peters and Goslow, 1983). However, EMG studies also have revealed dissimilarity between quadrupedal primates and nonprimates with respect to the recruitment patterns of forelimb muscles (e.g., Larson and Stern, 1987, 1989). It has been suggested that both the unique forelimb EMG patterns exhibited by primates and their preference for diagonal-sequence gaits are not simply due to biomechanical factors, but are likely to be the result of changes in primate neural control mechanisms. That is, neural control of quadrupedalism may depend more on supraspinal input in primates than in nonprimate mammals (Eidelberg et al., 1981).This evolutionary change in primate neural control is thought to be associated with the superior manipulative abilities of primate forelimbs and their tendency to be released from locomotor function (Vilensky, 1987, 1989; Vilensky and Larson, 1989; Larson and Stern, 1989). Although primates do not appear to be unique in hind-limb muscle function during quadrupedalism, the contrasting results for forelimb muscles suggest that further investigation of muscle function during primate quadrupedalism in a comparative context is warranted. A thorough comparison of mus- BACK MUSCLE EMG IN QUADRUPEDALISM AND BIPEDALISM cle function during primate vs. nonprimate quadrupedalism should include a consideration of all muscle groups important to locomotion, i.e., axial as well as appendicular muscles. Research on back function in the cat has demonstrated that during quadrupedal walking, multifidus, longissimus, and iliocostalis contract bilaterally with touchdown of each foot; i.e., there are two “bursts” of muscle activity during each step cycle (Carlson et al., 1979; English, 1980; Zomlefer et al., 1984).Tokuriki’s (1973a)data on longissimus in a dog (at the level of the fourth lumbar vertebra) also indicate that this muscle contracts in association with touchdown of each foot during walking. Taylor (1978),referring to unpublished observations, noted that the lumbar iliocostalis of dogs is inactive during walking. Back-muscle function during primate quadrupedalism remains undocumented. The purpose of the study reported here is to examine the recruitment patterns of multifidus, longissimus, and iliocostalis in primates walking quadrupedally and to compare the results to those reported for nonprimate mammals. In a previous study, we collected EMG data on the same back muscles for a chimpanzee and two gibbons walking bipedally, and compared the results to human backmuscle activity during bipedalism (Shapiro and Jungers, 1988). Results indicated a basic similarity in back-muscle recruitment patterns between the nonhuman hominoids and humans, reminiscent of the “conservative” nature of hind-limb muscle function characterizing primates and nonprimates walking quadrupedally. This “conservation” of back-muscle function across species raises the question of whether or not backmuscle function is also maintained across locomotor behaviors. Therefore, in the current study, the recruitment patterns of selected axial muscles in primates (chimpanzee, orangutan, and baboon) walking quadrupedally are not only compared to those characterizing nonprimate quadrupedalism, but are also compared to our previous data on bipedalism. In addition, new data on back-muscle function in bipedalism are presented here for the orangutan. The aims of the study can be summarized as follows: 493 1. To describe and compare the recruitment patterns of several back muscles during quadrupedalism in three primate species. This aspect of the study allows us to test whether or not the function of primate back muscles during quadrupedalism is consistent across species, as was reported for bipedalism (Shapiro and Jungers, 1988). This represents the first attempt known to use to investigate primate back-muscle function in quadrupedalism. 2. To test whether or not there are any differences in back-muscle function between primates and nonprimate mammals walking quadrupedally, given their reported differences in forelimb muscle activity as well as their differences in footfall sequence preferences. Nonprimate back-muscle EMG data were taken from the literature. 3. To compare back-muscle function in primates during quadrupedalism to backmuscle function during bipedalism in nonhuman hominoids and humans. Are the muscle activity patterns utilized in quadrupedalism called upon in bipedalism, or does the role of the back muscles change as the trunk is subjected to new biomechanical demands? MATERIALS AND METHODS Subjects The experiments were performed in a large enclosure (7.3 m x 3.7 m x 2.7 m). Details of the experimental environment can be found elsewhere (Stern et al., 1980; Jungers et al., 1983). Each subject was encouraged to walk quadrupedally andlor bipedally for food rewards. During quadrupedalism, all subjects walked with a diagonalsequence gait. The subjects included one male chimpanzee (Pan troglodytes), two orangutans (Pongo pygmaeus), and one male baboon (Papio anubis). We also collected quadrupedal and bipedal data for a second chimpanzee. The results for the two individuals were consistent, but data from only one individual are presented here. This chimpanzee is the same individual whose bipedalism was investigated in Shapiro and Jungers (1988). For one of the orangutans (“Solok), most of the quadrupedal step cycles were irregular, leaving only one or two 494 L.J. SHAPIRO AND W.L. JUNGERS digitizable step cycles. Even after training, the other orangutan (“Tombak) was rarely able to walk bipedally without hand support from the trainer. Therefore, the quadrupedal data presented here are from one of the orangutans (Tombak), and the bipedal data are from the other (Solok). cle revealed no tail contraction, verifying that the electrode was not placed in the extensor caudae lateralis. Longissirnus In the hominoids, the electrode was placed in longissimus at the lower thoracic level, about 4-5 cm lateral to the spinous proExperimental technique cesses. In the baboon, the electrode was The technique employed was telemetered placed about 1cm lateral to the spinous proelectromyography with simultaneous video cesses at the level of the last rib. In all subrecording as described in Stern et al. (1977, jects, electrode placement was verified by 1980). Each animal was placed under stimulation, revealing the contraction of the halothanelnitrous oxide anesthesia, after craniocaudally oriented fibers characterizwhich fine-wire bipolar electrodes were ining longissimus. serted into multifidus, longissimus, and iliocostalis. Dissections of cadavers of compara- Iliocos tal is ble size to each of the experimental subjects In the hominoids, the electrode was placed were performed in order to determine relevant musculoskeletal landmarks and the lo- in iliocostalis at a point along a vertical line cation at which each muscle is most accessi- passing through the midpoint of the iliac ble for electrode insertion (see Figs. 1and 2 crest, at the level of the last two ribs. In the baboon, the electrode was placed about 3.5 in Shapiro and Jungers, 1988). cm lateral to the spinous processes at a level Multifidus approximately midway between the ribs and the iliac crest. Again, electrode placement In the hominoids, the electrode was inserted just medial t o the most posteromedial was verified for all subjects by stimulation, aspect of the iliac crest (i.e., the “posterior revealing the contraction of the craniolatersuperior iliac spine”).Multifidus is the most ally oriented fibers characterizing iliocostawell developed here, and although it is a lis. relatively deep muscle, at this level only Data analysis skin, fascia and erector spinae aponeurosis lie superficially. For this reason, confidence Details of the method used here can be in the accuracy of electrode placement was found in Stern et al. (19801, Jungers et al. high. Electrode placement was verified by (19831, and Larson and Stern (1986); only a sending a small stimulating current back brief summary follows. In order to quantify through the electrode, revealing the contrac- the results, the videotape recording of each tion of the craniomedially oriented fibers animal with the EMG signals superimposed characterizing multifidus. was played back frame by frame. The exact The existence of the tail in the baboon timing of behavioral events with respect to made electrode placement a bit more diffi- the EMG signals was noted and transposed cult. Just medial to the iliac crest and lateral onto a hard copy of the video image. This to multifidus lies the lumbar portion of the was possible because at the end of each twoextensor caudae lateralis (i.e., the “lumbo- second sweep of the oscilloscope beams, the coccygeus”; see Bogduk, 1973, 1980), which image of the animal is deleted, leaving only is relatively large in the baboon. Therefore, an image of the EMG signals. placement of the electrode just medial to the Behavioral “events” refer to the onset or iliac crest would most likely be located in cessation of step cycles. Although the hands this tail extensor, rather than in multifidus. as well as the feet contact the substrate durTo avoid this problem, in the baboon the ing quadrupedal walking, the feet were choelectrode was placed deeply at a level just sen as the point of reference for consistency above the iliac crest, and just lateral to the with both the bipedal data and with the litspinous processes. Stimulation of the mus- erature on nonprimate back muscle function BACK MUSCLE EMG IN QUADRUPEDALISM AND BIPEDALISM during quadrupedalism (e.g., Tokuriki, 1973a; CarIson et al., 1979; English, 1980; Zomlefer et al., 1984). That is, “touchdown77 of the foot (TD) marks the beginning of support phase, and “toe-off of the foot (TO) marks the beginning of swing phase. We also identified the midphases (M) of support and swing (the points at which the hip and ankle are aligned vertically). Since touchdown of the opposite or contralateral foot was important in the interpretation of back muscle function in bipedalism, we also included this reference point ( C ) in the analysis of quadrupedalism (e.g., Fig. 1). In order to quantify the data, the duration and relative amplitude of the raw EMG signals were digitized in conjunction with the behavioral events, and transformed into a matrix representation. The matrix indicates the percent frequency of muscle activity at a given amplitude and at a given point in the locomotor cycle (Stern et al., 1980; Jungers et al., 1983; Larson and Stern, 1986). The relative intensity of EMG activity was estimated by comparing signal amplitude to a designated maximum burst for each muscle during each experiment. For all results except iliocostalis in Papio, activity referred to as “consistent” occurred at least 67% of the time; activity occurring between 33% and 67% of the time was considered to be frequent but less predictable (e.g., Fig. 1). The EMG data presented for the baboon’s multifidus and longissimus were collected in a single experiment, during which the signal for iliocostalis was lost. A second experiment on the same three muscles in the same individual was only successful in providing data for multifidus and iliocostalis. In this second experiment, the frequency of muscle activity was lower overall than in the first baboon experiment and when compared to the other primate subjects. Despite the differences in frequency, the muscle activity patterns of multifidus were similar in the two experiments, and results are presented here from the first experiment. For Papio’s iliocostalis, however, activity from the second experiment is depicted which occurred between 30% and 40% of the time and between 5% and 30% of the time. Although walking speed was not measured directly, the baboon clearly exhibited 495 Multifidus: Quadrupedalism Papio Pan TO M C TO Pongo D Fig. 1. Quantified muscle activity patterns of multifidus (right side) during quadrupedal walking in Papio, Pan, and Pongo. TD = touchdown of the ipsilateral (right) foot, marking the beginning of support phase. M = midphase of support or swing, when the hip and ankle form a line perpendicular to the substrate. C = touchdown of the contralateral (left)foot. TO = toeoff of the ipsilateral (right) foot, marking the beginning of swing phase. Blackened areas indicate consistent activity; that is, activity occurring at least 67% of the time. Striped areas indicate frequent, but less predictable activity; that is, activity occurring between 33%and 67% of the time. The numbers in each box represent the number of step cycles analyzed for each support or swing phase. Amplitude is scaled to the maximum activity observed in each muscle. L.J. SHAPIRO AND W.L. JUNGERS 496 a variety of speeds. Nevertheless, the baboon's slow and moderately paced step cycles were combined after separate examination of results revealed similar patterns of muscle activity at each speed (this pertains to the results for multifidus and longissimus only; speed differences were less noticable in the separate experiment from which the ilio- Papio costalis data were generated). tongissimus: Quadrupedalism RESULTS Primate quadrupedalism Multifidus Multifidus activity in all three primates generally resembles that reported for the cat: there are two bursts of multifidus activity during the step cycle, one associated with touchdown of the ipsilateral foot (TD), and the other associated with touchdown of the Pan contralateral foot (C; Fig. 1). The latter burst is clearly centered around contralateral touchdown and is more consistent in pattern among the three primates, while the ipsilateral burst is somewhat more variable in pattern among the three. Longissimus 19 In all three primates, longissimus (like multifidus) is consistently active as the contralateral foot touches down (Fig. a), a pat- Pongo tern resembling that reported for cats and dogs, However, the ipsilateral burst displayed by the nonprimates is more variable in pattern among the primates. For examTD M C TO M ple, in the chimpanzee, longissimus shows Fig. 2. Quantified muscle activity patterns of longissivariable activity throughout swing phase, with a slight increase in amplitude just be- mus (right side) during quadrupedal walking in Papio, Pun, and Pongo. See Figure 1for abbreviations and confore ipsilateral touchdown, and brief but ventions. consistent activity at touchdown. There is some activity surrounding ipsilateral touchdown in the orangutan, but it is not consistent. Overall, the results for the baboon tivity is associated with ipsilateral and espemore closely resemble those reported for cially contralateral touchdown (Fig. 3). This nonprimates in that there are two consis- pattern resembles that of the other two mustent bursts of activity, one at contralateral cles (Figs. 1,2) and also resembles results touchdown and the other just after ipsilat- reported for cats. If anything, it is likely that eral touchdown. we have underestimated the frequency of iliocostalis activity at touchdown in the baIliocostalis boon. Iliocostalis is recruited differently in the Despite the lower frequency and relative inconsistency of iliocostalis activity recorded chimpanzee (Fig. 3) than are multifidus and for Papio, when iliocostalis is active, its ac- longissimus (Figs. 1,2). Unlike the latter two BACK MUSCLE EMG IN QUADRUPEDALISM AND BIPEDALISM Iliocostalis: Quadrupedalism wing Phase rppoti Phase 11 13 497 the only consistent activity is a burst centered around touchdown of the contralateral foot, and there is variable activity just before and at ipsilateral touchdown. Its “double burst” pattern is also quite similar to the activity exhibited by multifidus (Fig. 1). Primate us. nonprimate quadrupedalism Papio <D a k M 17 16 Pan TD M C TO M TD As the results reported above make clear, there is a general (but not identical) resemblance between primates and nonprimates in terms of back-muscle recruitment patterns during quadrupedalism. The baboon’s back-muscle activity more closely resembles that of cats and dogs than does that of the two hominoids. Importantly, the chimpanzee’s deviations from the nonprimate quadrupedal pattern (e.g., variable activity of longissimus throughout swing phase, Fig. 2; absence of ipsilateral activity of iliocostalis, Fig. 3) are the same patterns that produce the resemblances between this hominoid’s quadrupedal and bipedal EMG patterns (see below). Quadrupedulism us. bipedalism Pongo TD M C TO M TD Fig. 3. Quantified muscle activity patterns of iliocostalis (right side) during quadrupedal walking in Papio, Pun, and Pongo. For Pupio only, blackened areas indicate activity occurring between 30%and 40% of the time and striped areas indicate activity occurring between 5%and 30%of the time (see text for explanation). Otherwise, abbreviations and conventions are as in Figure 1. muscles, iliocostalis exhibits no ipsilateral touchdown activity at all. However, like multifidus and longissimus, iliocostalis contracts consistently a t contralateral touchdown, followed by variable swing phase activity. The recruitment pattern for the orangutan’s iliocostalis (Fig. 3) is essentially identical to that for its longissimus (Fig. 2): Figure 4 reveals that the chimpanzee uses the three muscles quite similarly during both quadrupedalism and bipedalism, despite the different mechanical demands placed on the back in these two locomotor behaviors. During both behaviors, multifidus and longissimus respond to touchdown of each foot (with an emphasis on the contralateral burst), and iliocostalis responds almost solely to contralateral touchdown in both behaviors. In other words, control of the trunk in bipedalism is accomplished with the same basic patterns that are utilized during quadrupedalism. In fact, the recruitment of longissimus and especially iliocostalis when a chimpanzee walks quadrupedally is more similar to that of a chimpanzee walking bipedally than it is to that of a cat walking quadrupedally. In the orangutan, iliocostalis exhibits the same recruitment pattern during bipedalism that it does in the chimpanzee, gibbon, and human (Shapiro and Jungers, 1988);its activity is associated with touchdown of the contralateral foot and restriction of excessive movement of the trunk in that direction L.J. SHAPIRO AND W.L. JUNGERS 498 C HIMPANZE E F I Quadrupedalism Bipedalism Mult. 18 17 35 Long. I I cio d rb ’I ”I Ilio. TD M C TO M TD Fig. 4. Comparison of EMG activity during quadrupedal and bipedal walking in the chimpanzee for all three back muscles. Mult. = Multifidus; Long. = Longissimus; Ilio. = Iliocostalis. See Figure 1 for other abbreviations and conventions. (Fig. 5). However, whereas multifidus and longissimus contract bilaterally with touchdown of each foot in the chimpanzee, gibbon, and human, these two muscles in the orangutan respond to contralateral touchdown only (resembling the pattern for iliocostalis). DISCUSSION The three primates included in this study exhibit an overall similarity to each other in the recruitment patterns of each of the three muscles during quadrupedalism. In addition to the consistency among the primate species, the results for the primates reveal a basic similarity to those reported for nonprimate mammals. That is, in both primates and nonprimates, each of these muscles (with the exception of iliocostalis in the chimpanzee) contracts bilaterally with or just before touchdown of each foot. Zomlefer et al. (1984) reported that when “high decerebrate” cats (whose back muscle activity closely resembles that reported for “intact” cats) walk, back muscle contractions associated with the contralateral foot are of longer duration and higher amplitude BACK MUSCLE EMG IN QUADRUPEDALISM AND BIPEDALISM 499 ORANGUTAN Quadrupedalism Bipedalism Mult. D Long. TD M C TO M 14 TD TO M C TO M 17 19 TD 17 Ilio. TD M C TO M TD TD M C TO M TD Fig. 5. Comparison of EMG activity during quadrupedal and bipedal walking in the orangutan for all three back muscles. See Figures 1 and 4 for abbreviations and conventions. than those associated with ipsilateral touchdown. The authors concluded that the primary role of the back muscles during walking is unlikely to be one of increasing step length through lateral flexion of the trunk (as suggested for dogs by Tokuriki, 1973a). Rather, during quadrupedal walking, the back muscles act not to initiate movements, but to “compensate for andor minimize the movements of the trunk at each step” (Zomlefer et al., 1984:259; see also Carlson et al., 1979).This, in turn, stabilizes the pelvis and provides a rigid origin for muscles acting on the hind limb (English, 1980; see also Carrier, 1987). Zomlefer et al.’s (1984) con- clusion is consistent with other evidence demonstrating that during quadrupedal walking in cats, the range of angular movements of the lumbar spine in the sagittal and frontal planes is small (Goslow et al., 1973; Carlson et al., 1979; English, 1980). Despite the overall similarity between primate and nonprimate back-muscle EMG during quadrupedalism, a t least two differences are notable. First, the chimpanzee does not recruit iliocostalis at ipsilateral touchdown. This difference becomes important when quadrupedalism is compared to bipedalism (see below). Second, when there is both an ipsilateral and contralateral burst 500 L.J. SHAPIRO AND W.L. JUNGERS of muscle activity during primate quadrupedalism, the bursts differ little in amplitude or duration (in contrast to the cat data). Nevertheless, the back-muscle activity of both primates and cats exhibits an emphasis on the response to contralateral touchdown. In cats, this emphasis is expressed by higher amplitude and longer duration (Zomlefer et al., 1984). Among the primates, it is expressed by the fact that more consistent activity occurs at contralateral than at ipsilateral touchdown, especially in the hominoids. In other words, although the primatenonprimate EMG is not identical in pattern, the results are similar enough to suggest that like cats and dogs, it is likely that backmuscle activity during quadrupedal walking in primates is associated primarily with the restriction of movements of the trunk. Similarity in recruitment patterns between primates (Ateles, Lagothrix, and Erythrocebus) and nonprimates during quadrupedal walking has also been reported with respect to hind-limb muscle function (Vangor and Wells, 1983). Thus, despite certain functional specializations that have been associated with primate quadrupedalism (e.g., distribution of weight between forelimbs and hind limbs, footfall pattern; EMG of forelimb muscles), neither hind limb nor backmuscle function during quadrupedalism appears to clearly or consistently distinguish primates from other mammals (at least those few mammals for which we have data). The reason why the chimpanzee recruits iliocostalis solely at contralateral touchdown in quadrupedalism remains to be explained. While speed was not recorded in this study, this type of muscle recruitment pattern does not appear to be clearly related to walking speed. In nonprimate mammals in which speed has been measured, back muscles are inactive at very slow walking speeds. Otherwise, burst duration is inversely correlated with speed during walking, but the biphasic pattern of back-muscle activity remains intact even at a trot. It is only higher speeds accompanied by a change in gait to a gallop that produce a markedly different muscle activity pattern from the walkhot pattern (Tokuriki, 1973a,b, 1974; Taylor, 1978; Carlson et al., 1979; English, 1980; Zomlefer et al., 1984). Therefore, it is unlikely that the chimpanzee’s absence of iliocostalis activity at one foot touchdown but not the other is attributable to its walking speed. Alternatively, the chimpanzee’s distinctive pattern may be related to the fact that in hominoids, the iliac blade is mediolaterally expanded, and accordingly, iliocostalis (which originates on the iliac crest) is situated more laterally than in monkeys or in cats (Reynolds, 1931). Its EMG pattern in chimpanzees could then be viewed as a reflection of its efficient leverage for lateral flexion, which is emphasized when counteracting movements of the trunk brought about by contralateral touchdown (in quadrupedalism as well as bipedalism). This hypothesis is supported by the orangutan’s iliocostalis activity in quadrupedalism. In the orangutan (whose iliocostalis is also laterally situated), ipsilateral activity is present but not a t all consistent, while there is a consistent burst centered around contralateral touchdown. Finally, the difference between chimpanzees (and to some extent orangutans) on the one hand, and cats and baboons on the other, might be related to overall body shape differences (e.g., thoracic widtwdepth proportions) or t o currently undocumented kinematic differences in walking pattern between these two groups. The primate results become even more illuminating when quadrupedalism is compared to bipedalism. Unlike quadrupedalism, bipedalism requires balance of the trunk in the sagittal and frontal planes (Thorstensson et al., 1982,1984).During bipedalism in chimpanzees, gibbons and humans, the more medially placed multifidus and longissimus contract bilaterally with each step, controlling movement in the sagittal plane. The more laterally placed iliocostalis contracts unilaterally in response to contralateral touchdown only, controlling movement of the trunk in the oblique and frontal planes (Shapiro and Jungers, 1988). In the chimpanzee, this “bipedal” pattern is maintained from the same basic pattern characterizing quadrupedalism. In fact, although the chimpanzee differs from the cat with respect to a lack of ipsilateral activity in iliocostalis during quadrupedalism, this same lack of ipsilateral activity occurs when BACK MUSCLE EMG IN QUADRUPEDALISM AND BIPEDALISM the chimpanzee walks bipedally. In other words, the recruitment patterns used in quadrupedalism for control of the trunk or stabilization of the pelvis appear to serve equally well for control of the fore-aft and lateral movements of the trunk associated with bipedalism (Thorstensson et al., 1982, 1984). Unlike the bipedal patterns of the other hominoids, all three back muscles in the orangutan respond solely to contralateral touchdown. The results for the orangutan are consistent with its awkward and “unnatural” form of bipedal walking (orangutans have not been observed walking bipedally in the wild; Sugardjito, 19821, and indicate that for this hominoid, the maintenance of balance requires almost exclusively side to side control of the trunk (i.e., in the frontal plane). The more human-like bipedalism employed by the chimpanzee and gibbon is characterized by control of movements in both the sagittal and frontal planes. In chimpanzees, gibbons, and humans, however, back-muscle activity associated with contralateral touchdown is of higher amplitude than that responding to ipsilateral touchdown (for those muscles with adequate lateral leverage; longissimus and iliocostalis). This suggests that one of the most demanding mechanical requirements of bipedalism in hominoids is the necessity to prevent the trunk from swaying in the frontal plane (Keith, 1923)- the response to which is most exaggerated in the least natural biped, the orangutan. The uniqueness of the orangutan’s back-muscle activity in bipedalism also accounts for the fact that back-muscle recruitment patterns differ between quadrupedal and bipedal walking in this highly specialized/suspensory hominoid (unlike the results for the chimpanzee; cf. Figs. 4,5). Vangor and Wells (1983) found that for several hind-limb muscles in Ateles, Lagothrix, and Erythrocebus, the phasic activity patterns characterizing quadrupedalism were very similar to those associated with bipedalism in the same animals. In other words, the functions of those particular hind-limb muscles in bipedal gait were “conserved” from quadrupedal walking. In addition, for those muscles, the patterns for qua- 501 drupedal and bipedal walking were also used by Ateles and Lagothrix during vertical climbing. For other hind-limb muscles, recruitment patterns used in bipedal walking were found to be different from those observed during quadrupedal walking, but similar to those used during vertical climbing. Vangor and Wells (1983) concluded from their results that “vertical climbing, therefore, rather than quadrupedal walking, may establish a phasic activity pattern for a particular muscle which is conserved when the primate undertakes bipedal walking” (pp. 133-134). In other words, they considered climbing to be “preadaptive” for bipedalism (see also Prost, 1980; Fleagle et al., 1981). Finally, Vangor and Wells (1983) found no specific or consistent resemblance to human bipedalism in the phasic activity patterns of various hind-limb muscles during nonhuman primate quadrupedalism, bipedalism, or vertical climbing. The implication of their results is that “the ancestor of bipedal humans. . . had to undergo much change in muscle function with the evolution of bipedal gait, whether or not it was originally human-like in gross structure or gross locomotor behavior” (Vangor and Wells, 1983:134). The overall similarity in back-muscle activity reported here for quadrupedalism, nonhuman bipedalism and human bipedalism is noteworthy. It suggests that, contrary to Vangor and Wells’ (1983) statement regarding hind-limb muscle function, the ancestor of bipedal humans would not have had to undergo major change in back-muscle function with the evolution of bipedal gait. Notwithstanding the similarity in back function during quadrupedalism and bipedalism, similar back-muscle EMG patterns in vertical climbing and bipedalism would lend further support to the notion that climbing is preadaptive for bipedalism. We are currently investigating back-muscle function in vertical climbing and other suspensory behaviors in nonhuman primates (see also Hurov, 1982). The suggestion that primate “spinal circuitry” differs from that characterizing nonprimate mammals (as a result of the evolution of primates’ emphasis on grasping and manipulative abilities of the forelimb) 502 L.J. SHAPIRO AND W.L. JUNGERS has been based largely on the observed differences in forelimb muscle EMG patterns between primates and other mammals (Larson and Stern, 1987; Vilensky, 1987, 1989; Vilensky and Larson, 1989; Larson and Stern, 1989). If primate neural circuitries diverged from that of other mammals, it is possible that this evolutionary specialization may be reflected only in forelimb muscle function, since this is the part of the body on which “neural evolution”has presumably directly acted. Therefore, the lack of pronounced differences between primates and nonprimate mammals in back (and hindlimb) muscle function neither adds support nor weakens this neurological hypothesis. CONCLUSIONS AND SUMMARY This study represents the first time the EMG of back muscles during quadrupedalism in primates has been documented. The results can be summarized as follows: 1.In Pan, Pongo, and Papio, back muscles restrict movements of the trunk by responding to touchdown of one or both feet during quadrupedalism, with an emphasis especially in the hominoids on the response to contralateral touchdown. Recruitment patterns during quadrupedalism are generally consistent across species, as was reported for bipedalism (Shapiro and Jungers, 1988). 2. Nonprimate mammals (i.e., cats and dogs) and primates recruit their back muscles in a generally similar fashion (but not identically) during quadrupedal walking despite the differences between these two groups in footfall sequence and forelimb muscle recruitment patterns. The discrepancies in back-muscle EMG patterns that were observed between hominoids on the one hand and cats and monkeys on the other most likely reflect differences in muscular arrangement, general body shape andlor kinematic differences in walking between these two groups. 3. There is a fundamental similarity in back-muscle activity patterns during quadrupedalism (nonprimate and primate) and bipedalism (including humans), despite the differences in mechanical demands placed on the back in these two behaviors. In conclusion, back-muscle activity patterns demonstrate a basic conservatism across animals and locomotor behaviors. This finding contrasts with EMG studies on the primate forelimb (Larson and Stern, 1987, 19891, but does not preclude the suggestion that aspects of primate neural circuitry for locomotion are unique. Regardless, the results reported here are consistent with other electromyographic data pointing to considerable conservatism in the evolution of the neural control of tetrapod limb movement (Peters and Goslow, 1983 and references therein). ACKNOWLEDGMENTS The EMG data could not have been collected without the help of Susan Larson, Jack Stern, Hillary Johnston, Malcolm McClinton, Aaron Blaisdell, and Marianne Crisci. Daniel Schmitt deserves special thanks for his help in data analysis, and thanks also go to George Boykin and Charles Garrison for their patience and generosity concerning dissection materials. The paper was improved by the suggestions of two anonymous reviewers, and John Fleagle, Jack Stern, and Farish Jenkins provided useful comments and discussion on an earlier version of the manuscript. This research was supported by NSF grants to Jack Stern: BNS 8519747, BNS 8606781, BNS 8819621, BNS 8823083, and DBS 9209004, and by the University of Texas at Austin Research Institute. LITERATURE CITED Bogduk N (1973)Aspects of comparative anatomy of the dorsal lumbosacral region. Thesis, University of Sydney, Australia. Bogduk N (1980) The dorsal lumbar muscles of the cat. Anat. Anz. 148r55-67. Carlson H , Halbertsma J, and Zomlefer M (1979) Control of the trunk during walking in the cat. Acta Physiol. Scand. 205t251-253. Carrier D (1987) The evolution of locomotor stamina in tetrapods: Circumventing a mechanical constraint. Paleobiology 23:32&341. Demes B, Larson S, Stern JT, Jr., and Jungers W (1992) The hind limb drive of primates-theoretical reconsideration and empirical examination of a widely held concept. Am. J. Phys. Anthropol. (Supp. 14):69. Demes B, Larson S, Stern JT, Jr., Jungers W, and Biknevicius A (1994) The kinetics of primate quadrupedalism: “Hind limb drive” reconsidered. J. Hum. Evol. (in press). Eidelberg E, Walder J , and Nguyen L (1981) Locomotor control in macaque monkeys. Brain 204t647-663. BACK MUSCLE EMG IN QUADRUPEDALISM AND BIPEDALISM English A (1980) The functions of the lumbar spine during stepping in the cat. J . Morphol. 165:5546. Fleagle JG, Stern J T Jr., Jungers WL, Susman RL, Vangor AK, and Wells JP (1981) Climbing: A biomechanical link with brachiation and with bipedalism. In MH Day (ed.): Vertebrate Locomotion. Symposia of the Zoological Society of London 48:35!%375. London: Academic Press. Goslow G J r , Reinking R, and Stuart D (1973) The cat step cycle: Hind limb joint angles and muscle lengths during unrestrained locomotion. J. Morphol. 141: 142. Hildebrand M (1966) Analysis of the symmetrical gaits of tetrapods. Folia Biotheor. 6:9-22. Hildebrand M (1967) Symmetrical gaits of primates. Am. J. Phys. Anthropol. 26:119-130. Hurov J (1982) Surface electromyography of superficial back muscles in human children: Functions during vertical climbing and suspension and implications for the evolution of human bipedalism. J . Hum. Evol. 11: 117-130. Jungers W, and Anapol F (1985) Interlimb coordination and gait in the brown lemur (Lemur fuluus) and the talapoin monkey (Miopithecus tulupoin). Am. J . Phys. Anthropol. 67t89-97. Jungers W, Jouffroy F, and Stern JT, J r . (1980) Gross structure and function of the quadriceps femoris in Lemur fuluus: An analysis based on telemetered electromyography. J. Morphol. 164:287-299. Jungers W, Stern JT, Jr., and Jouffroy F (1983) Functional morphology of the quadriceps femoris in primates: A comparative anatomical and experimental analysis. Ann. Sci. Natl. Zool. 5:lOl-116. Keith A (1923) Man’s posture: Its evolution and disorders 11. The evolution of the orthograde spine. Brit. Med. J . 1:499-502. Kimura T, Okada M, and Ishida H (1979) Kinesiological characteristics of primate walking: Its significance in human walking. In M Morbeck, H Preuschoft, and N Gomberg (eds.): Environment, Behavior, and Morphology: Dynamic Interactions in Primates. New York Gustav Fischer, pp. 297-311. Larson S, and Stern JT, J r . (1986) EMG of scapulohumeral muscles in the chimpanzee during reach-q ing and “arboreal” locomotion. Am. J . Anat. 176:171190. Larson S, and Stern JT, Jr. (1987) EMG of chimpanzee shoulder muscles during knuckle-walking: Problems of terrestrial locomotion in a suspensory adapted primate. J . Zool. 212r629-655. Larson S, and Stern JT, Jr. (1989) The role of propulsive muscles of the shoulder during quadrupedalism in vervet monkeys (Cercopithecus aethiops): Implications for neural control of locomotion. J . Motor Behav. 21:457472. Muybridge E (1899) Animals in Motion. London: Chapman and Hall. Pandy M, Kumar V, Berme N, and Waldron K (1988) The dynamics of quadrupedal locomotion. J. kiomech. Eng. 110:230-237. Peters S, and Goslow GJ (1983) From salamanders to mammals: Continuity in musculoskeletal function during locomotion. Brain Behav. Evol. 22t191197. 503 Prost J (1965) The methodology of gait analysis and gaits of monkeys. Am. J. Phys. Anthropol. 23:203208. Prost J (1980) Origin of bipedalism. Am. J . Phys. Anthropol. 52:175-190. Reynolds E (1931) The evolution of the human pelvis in relation to the mechanics of the erect posture. Papers of the Peabody Museum of American Archaeology and Ethnology 11r255-334. Harvard University. Reynolds T (1985a) Mechanics of increased support of weight by the hindlimbs in primates. Am. J . Phys. Anthropol. 67:335-349. Reynolds T (198513) Stresses on the limbs of quadrupedal primates. Am. J . Phys. Anthropol. 67:351362. Rollinson J, and Martin R (1981) Comparative aspects of primate locomotion, with special reference to arboreal cercopithecines. Vertebrate Locomotion, Symposia of the Zoological Society of London 48:377427. London: Academic Press. Shapiro LJ, and Jungers WL (1988) Back muscle function during bipedal walking in chimpanzee and gibbon: Implications for the evolution of human locomotion. Am. J. Phys. Anthropol. 77:201-212. Stern JT, Jr., Wells J, Vangor A, and Fleagle J (1977) Electromyography of some muscles of the upper limb in Ateles and Lugothrix. Yrbk. Phys. Anthropol. 20: 498-507. Stern JT, Jr., Wells J , Jungers W, Vangor A, and Fleagle J (1980)An electromyographic study of pectoralis major in atelines and Hylobates, with special reference to the evolution of a pars clavicularis. Am. J . Phys. Anthropol. 52:13-25. Sugardjito J (1982) Locomotor behavior of the Sumatran Orang Utan IPongo pygmueus abelii) a t Ketambe, Gunung Leuser National Park. Malay Nat. J . 35: 57-64. Taylor CR (1978) Why change gaits? Recruitment of muscles and muscle fibers as a function of speed and gait. Am. Zool. 18:153-161. Thorstensson A, Carlson H, Zomlefer R. and Nilsson J (1982) Lumbar back muscle activity in relation to trunk movements during locomotion in man. Acta Physiol. Scand. 116:13-20. Thorstensson A, Nilsson J , Carlson H, and Zomlefer R (1984) Trunk movements in human locomotion. Acta Physiol. Scand. 121:9-22. Tokuriki M (1973a) Electromyographic and joint-mechanical studies in quadrupedal locomotion. I. Walk. Jap. J . Vet. Sci. 35r433-446. Tokuriki M (1973b) Electromyographic and joint-mechanical studies in quadrupedal locomotion. I. Trot. Jap. J . Vet. Sci. 35:525-533. Tokuriki M (1974) Electrornyographic and joint-rnechanical studies in quadrupedal locomotion. I. Gallop. Jap. J. Vet. Sci. 36:121-132. Vangor A, and Wells J (1983) Muscle recruitment and the evolution of bipedality: Evidence from telemetered electromyography of spider, woolly and patas monkeys. Ann. Sci. Natl. Zool. 5:125-135. Vilensky J (1987) Locomotor behavior and control in human and nonhuman primates: Comparisons with cats and dogs. Neurosci. Biobehav. Rev. 11:263274. 504 L.J. SHAPIRO AND W.L. JUNGERS Vilensky J (1989) Primate quadrupedalism: How and why does it differ from that of typical quadrupeds? Brain Behav. Evol. 34t357-364. Vilensky J, and Larson S (1989) Primate locomotion: Utilization and control of symmetrical gaits. Annu. Rev. Anthropol. 18:17-35. Zomlefer M, Provencher J, Blanchette G , and Rossignol S (1984) Electromyographic study of lumbar back muscles during locomotion in acute high decerebrate and low spinal cats. Brain Res. 290:249-260.